[DISCUSS] [FLINK-16824] Creating Temporal Table Function via DDL

[Kurt Young]

I agree with Timo, semantic about primary key needs more thought and
discussion, especially after FLIP-95 and FLIP-105.

Best,
Kurt


On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]> wrote:

> Hi Leonard,
>
> thanks for the summary.
>
> After reading all of the previous arguments and working on FLIP-95. I
> would also lean towards the conclusion of not adding the TEMPORAL keyword.
>
> After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can be
> represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The FOR
> SYSTEM_TIME AS OF t would trigger the internal materialization and
> "temporal" logic.
>
> However, we should discuss the meaning of PRIMARY KEY again in this
> case. In a TEMPORAL TABLE scenario, the source would emit duplicate
> primary keys with INSERT changeflag but at different point in time.
> Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> changelog semantics of FLIP-95 and FLIP-105 don't work well with a
> primary key declaration.
>
> Regards,
> Timo
>
>
> On 20.06.20 17:08, Leonard Xu wrote:
> > Hi everyone,
> >
> > Thanks for the nice discussion. I’d like to move forward the work,
> please let me simply summarize the main opinion and current divergences.
> >
> > 1. The agreements have been achieved:
> >
> > 1.1 The motivation we're discussing temporal table DDL is just for
> creating temporal table in pure SQL to replace pre-process temporal table
> in YAML/Table API for usability.
> > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD FOR
> SYSTEM_TIME” is to make user understand easily.
> > 1.3 For append-only table, it can convert to changelog table which has
> been discussed in FLIP-105, we assume the following temporal table is comes
> from changelog (Jark, fabian, Timo).
> > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x" instead of
> the current `LATERAL TABLE(rates(x))`  has come to an agreement(Fabian,
> Timo, Seth, Konstantin, Kurt).
> >
> > 2. The small divergence :
> >
> > About the definition syntax of the temporal table,
> >
> > CREATE [TEMPORAL] TABLE rates (
> >     currency CHAR(3) NOT NULL PRIMARY KEY,
> >     rate DOUBLE,
> >     rowtime TIMESTAMP,
> >     WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > WITH (...);
> >
> > there is small divergence whether add "TEMPORAL" keyword or not.
> >
> > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo, Fabian, Seth),
> the main advantages are:
> > （1）"TEMPORAL" keyword is intuitive to indicate the history tracking
> semantics.
> > （2）"TEMPORAL" keyword illustrates that queries can visit the previous
> versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME" keyword.
> >
> > 2.2 the other is using "CREATE TABLE"(Kurt), the main advantages are:
> > （1）Just primary key and time attribute can track previous versions of a
> table well.
> > （2）The temporal behavior is triggered by temporal join syntax rather
> than in DDL, all Flink DDL table are dynamic table logically including
> temporal table. If we decide to use "TEMPORAL" keyword and treats changelog
> as temporal table, other tables backed queue like Kafka should also use
> "TEMPORAL" keyword.
> >
> >
> > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows with 2.1
> may confuse users much. If we take a second to think about, for source/sink
> table which may backed queue (like kafka) or DB (like MySQL), we did not
> add any keyword in DDL to specify they are source or sinks, it works well.
> > I think temporal table is the third one,  kafka data source and DB data
> source can play as a source/sink/temporal table depends on the
> position/syntax that user put them in the query. The above rates table
> >      - can be a source table if user put it at `SELECT * FROM rates;`
> >      - can be a temporal table if user put it at `SELECT * FROM orders
> JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> >               ON orders.currency = rates.currency;`
> >      - can be sink table if user put is at `INSERT INTO rates SELECT *
> FROM …; `
> >  From these cases, we found all tables defined in Flink should be
> dynamic table logically, the source/sink/temporal role depends on the
> position/syntax in user’s query.
> >        In fact we have used similar syntax for current lookup table, we
> didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and trigger the
> temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x") in query.
> >
> > So, I prefer to resolve the small divergence with “CREATE TABLE” which
> > (1) is more unified with our source/sink/temporal dynamic table
> conceptually,
> > (2) is aligned with current lookup table,
> > (3) also make users learn less keyword.
> >
> > WDYT？
> >
> > Best,
> > Leonard Xu
> >
> >
>
>

[Konstantin Knauf-4]

Hi everyone,

I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.

Best,

Konstantin

On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]> wrote:

> I agree with Timo, semantic about primary key needs more thought and
> discussion, especially after FLIP-95 and FLIP-105.
>
> Best,
> Kurt
>
>
> On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]> wrote:
>
> > Hi Leonard,
> >
> > thanks for the summary.
> >
> > After reading all of the previous arguments and working on FLIP-95. I
> > would also lean towards the conclusion of not adding the TEMPORAL
> keyword.
> >
> > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can be
> > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The FOR
> > SYSTEM_TIME AS OF t would trigger the internal materialization and
> > "temporal" logic.
> >
> > However, we should discuss the meaning of PRIMARY KEY again in this
> > case. In a TEMPORAL TABLE scenario, the source would emit duplicate
> > primary keys with INSERT changeflag but at different point in time.
> > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > changelog semantics of FLIP-95 and FLIP-105 don't work well with a
> > primary key declaration.
> >
> > Regards,
> > Timo
> >
> >
> > On 20.06.20 17:08, Leonard Xu wrote:
> > > Hi everyone,
> > >
> > > Thanks for the nice discussion. I’d like to move forward the work,
> > please let me simply summarize the main opinion and current divergences.
> > >
> > > 1. The agreements have been achieved:
> > >
> > > 1.1 The motivation we're discussing temporal table DDL is just for
> > creating temporal table in pure SQL to replace pre-process temporal table
> > in YAML/Table API for usability.
> > > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD FOR
> > SYSTEM_TIME” is to make user understand easily.
> > > 1.3 For append-only table, it can convert to changelog table which has
> > been discussed in FLIP-105, we assume the following temporal table is
> comes
> > from changelog (Jark, fabian, Timo).
> > > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x" instead
> of
> > the current `LATERAL TABLE(rates(x))`  has come to an agreement(Fabian,
> > Timo, Seth, Konstantin, Kurt).
> > >
> > > 2. The small divergence :
> > >
> > > About the definition syntax of the temporal table,
> > >
> > > CREATE [TEMPORAL] TABLE rates (
> > >     currency CHAR(3) NOT NULL PRIMARY KEY,
> > >     rate DOUBLE,
> > >     rowtime TIMESTAMP,
> > >     WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > > WITH (...);
> > >
> > > there is small divergence whether add "TEMPORAL" keyword or not.
> > >
> > > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo, Fabian, Seth),
> > the main advantages are:
> > > （1）"TEMPORAL" keyword is intuitive to indicate the history tracking
> > semantics.
> > > （2）"TEMPORAL" keyword illustrates that queries can visit the previous
> > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME" keyword.
> > >
> > > 2.2 the other is using "CREATE TABLE"(Kurt), the main advantages are:
> > > （1）Just primary key and time attribute can track previous versions of a
> > table well.
> > > （2）The temporal behavior is triggered by temporal join syntax rather
> > than in DDL, all Flink DDL table are dynamic table logically including
> > temporal table. If we decide to use "TEMPORAL" keyword and treats
> changelog
> > as temporal table, other tables backed queue like Kafka should also use
> > "TEMPORAL" keyword.
> > >
> > >
> > > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows with
> 2.1
> > may confuse users much. If we take a second to think about, for
> source/sink
> > table which may backed queue (like kafka) or DB (like MySQL), we did not
> > add any keyword in DDL to specify they are source or sinks, it works
> well.
> > > I think temporal table is the third one,  kafka data source and DB data
> > source can play as a source/sink/temporal table depends on the
> > position/syntax that user put them in the query. The above rates table
> > >      - can be a source table if user put it at `SELECT * FROM rates;`
> > >      - can be a temporal table if user put it at `SELECT * FROM orders
> > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> > >               ON orders.currency = rates.currency;`
> > >      - can be sink table if user put is at `INSERT INTO rates SELECT *
> > FROM …; `
> > >  From these cases, we found all tables defined in Flink should be
> > dynamic table logically, the source/sink/temporal role depends on the
> > position/syntax in user’s query.
> > >        In fact we have used similar syntax for current lookup table, we
> > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and trigger
> the
> > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x") in
> query.
> > >
> > > So, I prefer to resolve the small divergence with “CREATE TABLE” which
> > > (1) is more unified with our source/sink/temporal dynamic table
> > conceptually,
> > > (2) is aligned with current lookup table,
> > > (3) also make users learn less keyword.
> > >
> > > WDYT？
> > >
> > > Best,
> > > Leonard Xu
> > >
> > >
> >
> >
>

--

Konstantin Knauf

https://twitter.com/snntrable

https://github.com/knaufk

[Jark Wu-2]

I'm also +1 for not adding the TEMPORAL keyword.

+1 to make the PRIMARY KEY semantic clear for sources.
From my point of view:

1) PRIMARY KEY on changelog souruce:
It means that when the changelogs (INSERT/UPDATE/DELETE) are materialized,
the materialized table should be unique on the primary key columns.
Flink assumes messages are in order on the primary key. Flink doesn't
validate/enforces the key integrity, but simply trust it (thus NOT
ENFORCED).
Flink will use the PRIMARY KEY for some optimization, e.g. use the PRIMARY
KEY to update the materilized state by key in temporal join operator.

2) PRIMARY KEY on insert-only source:
I prefer to have the same semantic to the batch source and changelog
source, that it implies that records are not duplicate on the primary key.
Flink just simply trust the primary key constraint, and doesn't valid it.
If there is duplicate primary keys with INSERT changeflag, then result of
Flink query might be wrong.

If this is a TEMPORAL TABLE FUNCTION scenario, that source emits duplicate
primary keys with INSERT changeflag, when we migrate this case to temporal
table DDL,
I think this source should emit INSERT/UPDATE (UPSERT) messages instead of
INSERT-only messages,  e.g. a Kafka compacted topic source?

Best,
Jark


On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]> wrote:

> Hi everyone,
>
> I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
>
> Best,
>
> Konstantin
>
> On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]> wrote:
>
> > I agree with Timo, semantic about primary key needs more thought and
> > discussion, especially after FLIP-95 and FLIP-105.
> >
> > Best,
> > Kurt
> >
> >
> > On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]> wrote:
> >
> > > Hi Leonard,
> > >
> > > thanks for the summary.
> > >
> > > After reading all of the previous arguments and working on FLIP-95. I
> > > would also lean towards the conclusion of not adding the TEMPORAL
> > keyword.
> > >
> > > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can be
> > > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The FOR
> > > SYSTEM_TIME AS OF t would trigger the internal materialization and
> > > "temporal" logic.
> > >
> > > However, we should discuss the meaning of PRIMARY KEY again in this
> > > case. In a TEMPORAL TABLE scenario, the source would emit duplicate
> > > primary keys with INSERT changeflag but at different point in time.
> > > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > > changelog semantics of FLIP-95 and FLIP-105 don't work well with a
> > > primary key declaration.
> > >
> > > Regards,
> > > Timo
> > >
> > >
> > > On 20.06.20 17:08, Leonard Xu wrote:
> > > > Hi everyone,
> > > >
> > > > Thanks for the nice discussion. I’d like to move forward the work,
> > > please let me simply summarize the main opinion and current
> divergences.
> > > >
> > > > 1. The agreements have been achieved:
> > > >
> > > > 1.1 The motivation we're discussing temporal table DDL is just for
> > > creating temporal table in pure SQL to replace pre-process temporal
> table
> > > in YAML/Table API for usability.
> > > > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD FOR
> > > SYSTEM_TIME” is to make user understand easily.
> > > > 1.3 For append-only table, it can convert to changelog table which
> has
> > > been discussed in FLIP-105, we assume the following temporal table is
> > comes
> > > from changelog (Jark, fabian, Timo).
> > > > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x" instead
> > of
> > > the current `LATERAL TABLE(rates(x))`  has come to an agreement(Fabian,
> > > Timo, Seth, Konstantin, Kurt).
> > > >
> > > > 2. The small divergence :
> > > >
> > > > About the definition syntax of the temporal table,
> > > >
> > > > CREATE [TEMPORAL] TABLE rates (
> > > >     currency CHAR(3) NOT NULL PRIMARY KEY,
> > > >     rate DOUBLE,
> > > >     rowtime TIMESTAMP,
> > > >     WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > > > WITH (...);
> > > >
> > > > there is small divergence whether add "TEMPORAL" keyword or not.
> > > >
> > > > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo, Fabian,
> Seth),
> > > the main advantages are:
> > > > （1）"TEMPORAL" keyword is intuitive to indicate the history tracking
> > > semantics.
> > > > （2）"TEMPORAL" keyword illustrates that queries can visit the previous
> > > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
> keyword.
> > > >
> > > > 2.2 the other is using "CREATE TABLE"(Kurt), the main advantages are:
> > > > （1）Just primary key and time attribute can track previous versions
> of a
> > > table well.
> > > > （2）The temporal behavior is triggered by temporal join syntax rather
> > > than in DDL, all Flink DDL table are dynamic table logically including
> > > temporal table. If we decide to use "TEMPORAL" keyword and treats
> > changelog
> > > as temporal table, other tables backed queue like Kafka should also use
> > > "TEMPORAL" keyword.
> > > >
> > > >
> > > > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows with
> > 2.1
> > > may confuse users much. If we take a second to think about, for
> > source/sink
> > > table which may backed queue (like kafka) or DB (like MySQL), we did
> not
> > > add any keyword in DDL to specify they are source or sinks, it works
> > well.
> > > > I think temporal table is the third one,  kafka data source and DB
> data
> > > source can play as a source/sink/temporal table depends on the
> > > position/syntax that user put them in the query. The above rates table
> > > >      - can be a source table if user put it at `SELECT * FROM rates;`
> > > >      - can be a temporal table if user put it at `SELECT * FROM
> orders
> > > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> > > >               ON orders.currency = rates.currency;`
> > > >      - can be sink table if user put is at `INSERT INTO rates SELECT
> *
> > > FROM …; `
> > > >  From these cases, we found all tables defined in Flink should be
> > > dynamic table logically, the source/sink/temporal role depends on the
> > > position/syntax in user’s query.
> > > >        In fact we have used similar syntax for current lookup table,
> we
> > > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and trigger
> > the
> > > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x") in
> > query.
> > > >
> > > > So, I prefer to resolve the small divergence with “CREATE TABLE”
> which
> > > > (1) is more unified with our source/sink/temporal dynamic table
> > > conceptually,
> > > > (2) is aligned with current lookup table,
> > > > (3) also make users learn less keyword.
> > > >
> > > > WDYT？
> > > >
> > > > Best,
> > > > Leonard Xu
> > > >
> > > >
> > >
> > >
> >
>
>
> --
>
> Konstantin Knauf
>
> https://twitter.com/snntrable
>
> https://github.com/knaufk
>

[Fabian Hueske-2]

Hi everyone,

Every table with a primary key and an event-time attribute provides what is
needed for an event-time temporal table join.
I agree that, from a technical point of view, the TEMPORAL keyword is not
required.

I'm more sceptical about implicitly deriving the versioning information of
a (temporal) table as the table's only event-time attribute.
In the query

SELECT *
FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
WHERE o.currency = r.currency

the syntax of the temporal table join does not explicitly reference the
version of the temporal rates table.
Hence, the system needs a way to derive the version of temporal table.

Implicitly using the (only) event-time attribute of a temporal table (rates
in the example above) to identify the right version works in most cases,
but probably not in all.
* What if a table has more than one event-time attribute? (TableSchema is
designed to support multiple watermarks; queries with interval joins
produce tables with multiple event-time attributes, ...)
* What if the table does not have an event-time attribute in its schema but
the version should only be provided as meta data?

We could add a clause to define the version of a table, such as:

CREATE TABLE rates (
   currency CHAR(3) NOT NULL PRIMARY KEY,
   rate DOUBLE,
   rowtime TIMESTAMP,
   WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
VERSION (rowtime)
WITH (...);

The presence of a the VERSION clause (or whatever syntax) would explicitly
define the version of a (temporal) table.
It would also render the need for the TEMPORAL keyword superfluous because
there would be another indicator that a table can be used in a temporal
table join.

I'm OK with not adding the TEMPORAL keyword, but I recommend that we think
again about the proposed implicit definition of a table's version and how
it might limit use in the future.

Cheers,
Fabian

Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:

> I'm also +1 for not adding the TEMPORAL keyword.
>
> +1 to make the PRIMARY KEY semantic clear for sources.
> From my point of view:
>
> 1) PRIMARY KEY on changelog souruce:
> It means that when the changelogs (INSERT/UPDATE/DELETE) are materialized,
> the materialized table should be unique on the primary key columns.
> Flink assumes messages are in order on the primary key. Flink doesn't
> validate/enforces the key integrity, but simply trust it (thus NOT
> ENFORCED).
> Flink will use the PRIMARY KEY for some optimization, e.g. use the PRIMARY
> KEY to update the materilized state by key in temporal join operator.
>
> 2) PRIMARY KEY on insert-only source:
> I prefer to have the same semantic to the batch source and changelog
> source, that it implies that records are not duplicate on the primary key.
> Flink just simply trust the primary key constraint, and doesn't valid it.
> If there is duplicate primary keys with INSERT changeflag, then result of
> Flink query might be wrong.
>
> If this is a TEMPORAL TABLE FUNCTION scenario, that source emits duplicate
> primary keys with INSERT changeflag, when we migrate this case to temporal
> table DDL,
> I think this source should emit INSERT/UPDATE (UPSERT) messages instead of
> INSERT-only messages,  e.g. a Kafka compacted topic source?
>
> Best,
> Jark
>
>
> On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]> wrote:
>
> > Hi everyone,
> >
> > I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
> >
> > Best,
> >
> > Konstantin
> >
> > On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]> wrote:
> >
> > > I agree with Timo, semantic about primary key needs more thought and
> > > discussion, especially after FLIP-95 and FLIP-105.
> > >
> > > Best,
> > > Kurt
> > >
> > >
> > > On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
> wrote:
> > >
> > > > Hi Leonard,
> > > >
> > > > thanks for the summary.
> > > >
> > > > After reading all of the previous arguments and working on FLIP-95. I
> > > > would also lean towards the conclusion of not adding the TEMPORAL
> > > keyword.
> > > >
> > > > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can be
> > > > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The FOR
> > > > SYSTEM_TIME AS OF t would trigger the internal materialization and
> > > > "temporal" logic.
> > > >
> > > > However, we should discuss the meaning of PRIMARY KEY again in this
> > > > case. In a TEMPORAL TABLE scenario, the source would emit duplicate
> > > > primary keys with INSERT changeflag but at different point in time.
> > > > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > > > changelog semantics of FLIP-95 and FLIP-105 don't work well with a
> > > > primary key declaration.
> > > >
> > > > Regards,
> > > > Timo
> > > >
> > > >
> > > > On 20.06.20 17:08, Leonard Xu wrote:
> > > > > Hi everyone,
> > > > >
> > > > > Thanks for the nice discussion. I’d like to move forward the work,
> > > > please let me simply summarize the main opinion and current
> > divergences.
> > > > >
> > > > > 1. The agreements have been achieved:
> > > > >
> > > > > 1.1 The motivation we're discussing temporal table DDL is just for
> > > > creating temporal table in pure SQL to replace pre-process temporal
> > table
> > > > in YAML/Table API for usability.
> > > > > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD FOR
> > > > SYSTEM_TIME” is to make user understand easily.
> > > > > 1.3 For append-only table, it can convert to changelog table which
> > has
> > > > been discussed in FLIP-105, we assume the following temporal table is
> > > comes
> > > > from changelog (Jark, fabian, Timo).
> > > > > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
> instead
> > > of
> > > > the current `LATERAL TABLE(rates(x))`  has come to an
> agreement(Fabian,
> > > > Timo, Seth, Konstantin, Kurt).
> > > > >
> > > > > 2. The small divergence :
> > > > >
> > > > > About the definition syntax of the temporal table,
> > > > >
> > > > > CREATE [TEMPORAL] TABLE rates (
> > > > >     currency CHAR(3) NOT NULL PRIMARY KEY,
> > > > >     rate DOUBLE,
> > > > >     rowtime TIMESTAMP,
> > > > >     WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > > > > WITH (...);
> > > > >
> > > > > there is small divergence whether add "TEMPORAL" keyword or not.
> > > > >
> > > > > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo, Fabian,
> > Seth),
> > > > the main advantages are:
> > > > > （1）"TEMPORAL" keyword is intuitive to indicate the history tracking
> > > > semantics.
> > > > > （2）"TEMPORAL" keyword illustrates that queries can visit the
> previous
> > > > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
> > keyword.
> > > > >
> > > > > 2.2 the other is using "CREATE TABLE"(Kurt), the main advantages
> are:
> > > > > （1）Just primary key and time attribute can track previous versions
> > of a
> > > > table well.
> > > > > （2）The temporal behavior is triggered by temporal join syntax
> rather
> > > > than in DDL, all Flink DDL table are dynamic table logically
> including
> > > > temporal table. If we decide to use "TEMPORAL" keyword and treats
> > > changelog
> > > > as temporal table, other tables backed queue like Kafka should also
> use
> > > > "TEMPORAL" keyword.
> > > > >
> > > > >
> > > > > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
> with
> > > 2.1
> > > > may confuse users much. If we take a second to think about, for
> > > source/sink
> > > > table which may backed queue (like kafka) or DB (like MySQL), we did
> > not
> > > > add any keyword in DDL to specify they are source or sinks, it works
> > > well.
> > > > > I think temporal table is the third one,  kafka data source and DB
> > data
> > > > source can play as a source/sink/temporal table depends on the
> > > > position/syntax that user put them in the query. The above rates
> table
> > > > >      - can be a source table if user put it at `SELECT * FROM
> rates;`
> > > > >      - can be a temporal table if user put it at `SELECT * FROM
> > orders
> > > > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> > > > >               ON orders.currency = rates.currency;`
> > > > >      - can be sink table if user put is at `INSERT INTO rates
> SELECT
> > *
> > > > FROM …; `
> > > > >  From these cases, we found all tables defined in Flink should be
> > > > dynamic table logically, the source/sink/temporal role depends on the
> > > > position/syntax in user’s query.
> > > > >        In fact we have used similar syntax for current lookup
> table,
> > we
> > > > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
> trigger
> > > the
> > > > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x") in
> > > query.
> > > > >
> > > > > So, I prefer to resolve the small divergence with “CREATE TABLE”
> > which
> > > > > (1) is more unified with our source/sink/temporal dynamic table
> > > > conceptually,
> > > > > (2) is aligned with current lookup table,
> > > > > (3) also make users learn less keyword.
> > > > >
> > > > > WDYT？
> > > > >
> > > > > Best,
> > > > > Leonard Xu
> > > > >
> > > > >
> > > >
> > > >
> > >
> >
> >
> > --
> >
> > Konstantin Knauf
> >
> > https://twitter.com/snntrable
> >
> > https://github.com/knaufk
> >
>

[Kurt Young]

Hi Fabian,

I agree with you that implicitly letting event time to be the version of
the table will
work in most cases, but not for all. That's the reason I mentioned `PERIOD
FOR` [1]
syntax in my first email, which is already in sql standard to represent the
validity of
each row in the table.

If the event time can't be used, or multiple event time are defined, we
could still add
this syntax in the future.

What do you think?

[1]
https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
Best,
Kurt


On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]> wrote:

> Hi everyone,
>
> Every table with a primary key and an event-time attribute provides what is
> needed for an event-time temporal table join.
> I agree that, from a technical point of view, the TEMPORAL keyword is not
> required.
>
> I'm more sceptical about implicitly deriving the versioning information of
> a (temporal) table as the table's only event-time attribute.
> In the query
>
> SELECT *
> FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
> WHERE o.currency = r.currency
>
> the syntax of the temporal table join does not explicitly reference the
> version of the temporal rates table.
> Hence, the system needs a way to derive the version of temporal table.
>
> Implicitly using the (only) event-time attribute of a temporal table (rates
> in the example above) to identify the right version works in most cases,
> but probably not in all.
> * What if a table has more than one event-time attribute? (TableSchema is
> designed to support multiple watermarks; queries with interval joins
> produce tables with multiple event-time attributes, ...)
> * What if the table does not have an event-time attribute in its schema but
> the version should only be provided as meta data?
>
> We could add a clause to define the version of a table, such as:
>
> CREATE TABLE rates (
>    currency CHAR(3) NOT NULL PRIMARY KEY,
>    rate DOUBLE,
>    rowtime TIMESTAMP,
>    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
> VERSION (rowtime)
> WITH (...);
>
> The presence of a the VERSION clause (or whatever syntax) would explicitly
> define the version of a (temporal) table.
> It would also render the need for the TEMPORAL keyword superfluous because
> there would be another indicator that a table can be used in a temporal
> table join.
>
> I'm OK with not adding the TEMPORAL keyword, but I recommend that we think
> again about the proposed implicit definition of a table's version and how
> it might limit use in the future.
>
> Cheers,
> Fabian
>
> Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:
>
> > I'm also +1 for not adding the TEMPORAL keyword.
> >
> > +1 to make the PRIMARY KEY semantic clear for sources.
> > From my point of view:
> >
> > 1) PRIMARY KEY on changelog souruce:
> > It means that when the changelogs (INSERT/UPDATE/DELETE) are
> materialized,
> > the materialized table should be unique on the primary key columns.
> > Flink assumes messages are in order on the primary key. Flink doesn't
> > validate/enforces the key integrity, but simply trust it (thus NOT
> > ENFORCED).
> > Flink will use the PRIMARY KEY for some optimization, e.g. use the
> PRIMARY
> > KEY to update the materilized state by key in temporal join operator.
> >
> > 2) PRIMARY KEY on insert-only source:
> > I prefer to have the same semantic to the batch source and changelog
> > source, that it implies that records are not duplicate on the primary
> key.
> > Flink just simply trust the primary key constraint, and doesn't valid it.
> > If there is duplicate primary keys with INSERT changeflag, then result of
> > Flink query might be wrong.
> >
> > If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
> duplicate
> > primary keys with INSERT changeflag, when we migrate this case to
> temporal
> > table DDL,
> > I think this source should emit INSERT/UPDATE (UPSERT) messages instead
> of
> > INSERT-only messages,  e.g. a Kafka compacted topic source?
> >
> > Best,
> > Jark
> >
> >
> > On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
> wrote:
> >
> > > Hi everyone,
> > >
> > > I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
> > >
> > > Best,
> > >
> > > Konstantin
> > >
> > > On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]> wrote:
> > >
> > > > I agree with Timo, semantic about primary key needs more thought and
> > > > discussion, especially after FLIP-95 and FLIP-105.
> > > >
> > > > Best,
> > > > Kurt
> > > >
> > > >
> > > > On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
> > wrote:
> > > >
> > > > > Hi Leonard,
> > > > >
> > > > > thanks for the summary.
> > > > >
> > > > > After reading all of the previous arguments and working on
> FLIP-95. I
> > > > > would also lean towards the conclusion of not adding the TEMPORAL
> > > > keyword.
> > > > >
> > > > > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can be
> > > > > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
> FOR
> > > > > SYSTEM_TIME AS OF t would trigger the internal materialization and
> > > > > "temporal" logic.
> > > > >
> > > > > However, we should discuss the meaning of PRIMARY KEY again in this
> > > > > case. In a TEMPORAL TABLE scenario, the source would emit duplicate
> > > > > primary keys with INSERT changeflag but at different point in time.
> > > > > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > > > > changelog semantics of FLIP-95 and FLIP-105 don't work well with a
> > > > > primary key declaration.
> > > > >
> > > > > Regards,
> > > > > Timo
> > > > >
> > > > >
> > > > > On 20.06.20 17:08, Leonard Xu wrote:
> > > > > > Hi everyone,
> > > > > >
> > > > > > Thanks for the nice discussion. I’d like to move forward the
> work,
> > > > > please let me simply summarize the main opinion and current
> > > divergences.
> > > > > >
> > > > > > 1. The agreements have been achieved:
> > > > > >
> > > > > > 1.1 The motivation we're discussing temporal table DDL is just
> for
> > > > > creating temporal table in pure SQL to replace pre-process temporal
> > > table
> > > > > in YAML/Table API for usability.
> > > > > > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD FOR
> > > > > SYSTEM_TIME” is to make user understand easily.
> > > > > > 1.3 For append-only table, it can convert to changelog table
> which
> > > has
> > > > > been discussed in FLIP-105, we assume the following temporal table
> is
> > > > comes
> > > > > from changelog (Jark, fabian, Timo).
> > > > > > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
> > instead
> > > > of
> > > > > the current `LATERAL TABLE(rates(x))`  has come to an
> > agreement(Fabian,
> > > > > Timo, Seth, Konstantin, Kurt).
> > > > > >
> > > > > > 2. The small divergence :
> > > > > >
> > > > > > About the definition syntax of the temporal table,
> > > > > >
> > > > > > CREATE [TEMPORAL] TABLE rates (
> > > > > >     currency CHAR(3) NOT NULL PRIMARY KEY,
> > > > > >     rate DOUBLE,
> > > > > >     rowtime TIMESTAMP,
> > > > > >     WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > > > > > WITH (...);
> > > > > >
> > > > > > there is small divergence whether add "TEMPORAL" keyword or not.
> > > > > >
> > > > > > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo, Fabian,
> > > Seth),
> > > > > the main advantages are:
> > > > > > （1）"TEMPORAL" keyword is intuitive to indicate the history
> tracking
> > > > > semantics.
> > > > > > （2）"TEMPORAL" keyword illustrates that queries can visit the
> > previous
> > > > > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
> > > keyword.
> > > > > >
> > > > > > 2.2 the other is using "CREATE TABLE"(Kurt), the main advantages
> > are:
> > > > > > （1）Just primary key and time attribute can track previous
> versions
> > > of a
> > > > > table well.
> > > > > > （2）The temporal behavior is triggered by temporal join syntax
> > rather
> > > > > than in DDL, all Flink DDL table are dynamic table logically
> > including
> > > > > temporal table. If we decide to use "TEMPORAL" keyword and treats
> > > > changelog
> > > > > as temporal table, other tables backed queue like Kafka should also
> > use
> > > > > "TEMPORAL" keyword.
> > > > > >
> > > > > >
> > > > > > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
> > with
> > > > 2.1
> > > > > may confuse users much. If we take a second to think about, for
> > > > source/sink
> > > > > table which may backed queue (like kafka) or DB (like MySQL), we
> did
> > > not
> > > > > add any keyword in DDL to specify they are source or sinks, it
> works
> > > > well.
> > > > > > I think temporal table is the third one,  kafka data source and
> DB
> > > data
> > > > > source can play as a source/sink/temporal table depends on the
> > > > > position/syntax that user put them in the query. The above rates
> > table
> > > > > >      - can be a source table if user put it at `SELECT * FROM
> > rates;`
> > > > > >      - can be a temporal table if user put it at `SELECT * FROM
> > > orders
> > > > > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> > > > > >               ON orders.currency = rates.currency;`
> > > > > >      - can be sink table if user put is at `INSERT INTO rates
> > SELECT
> > > *
> > > > > FROM …; `
> > > > > >  From these cases, we found all tables defined in Flink should be
> > > > > dynamic table logically, the source/sink/temporal role depends on
> the
> > > > > position/syntax in user’s query.
> > > > > >        In fact we have used similar syntax for current lookup
> > table,
> > > we
> > > > > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
> > trigger
> > > > the
> > > > > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
> in
> > > > query.
> > > > > >
> > > > > > So, I prefer to resolve the small divergence with “CREATE TABLE”
> > > which
> > > > > > (1) is more unified with our source/sink/temporal dynamic table
> > > > > conceptually,
> > > > > > (2) is aligned with current lookup table,
> > > > > > (3) also make users learn less keyword.
> > > > > >
> > > > > > WDYT？
> > > > > >
> > > > > > Best,
> > > > > > Leonard Xu
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> > > --
> > >
> > > Konstantin Knauf
> > >
> > > https://twitter.com/snntrable
> > >
> > > https://github.com/knaufk
> > >
> >
>

[Fabian Hueske-2]

Thanks Kurt,

Yes, you are right.
The `PERIOD FOR SYSTEM_TIME` that you linked before corresponds to the
VERSION clause that I used and would explicitly define the versioning of a
table.
I didn't know that the `PERIOD FOR SYSTEM_TIME` cause is already defined by
the SQL standard.
I think we would need a slightly different syntax though because (so far)
the validity of a row is determined by its own timestamp and the timestamp
of the next row.

Adding a clause later solves the ambiguity issue for tables with multiple
event-time attributes.
However, I'd feel more comfortable having such a cause and an explicit
definition of the temporal property from the beginning.
I guess this is a matter of personal preference so I'll go with the
majority if we decide that every table that has a primary key and an
event-time attribute should be usable in an event-time temporal table join.

Thanks, Fabian


Am Di., 23. Juni 2020 um 16:58 Uhr schrieb Kurt Young <[hidden email]>:

> Hi Fabian,
>
> I agree with you that implicitly letting event time to be the version of
> the table will
> work in most cases, but not for all. That's the reason I mentioned `PERIOD
> FOR` [1]
> syntax in my first email, which is already in sql standard to represent the
> validity of
> each row in the table.
>
> If the event time can't be used, or multiple event time are defined, we
> could still add
> this syntax in the future.
>
> What do you think?
>
> [1]
>
> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
> Best,
> Kurt
>
>
> On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]> wrote:
>
> > Hi everyone,
> >
> > Every table with a primary key and an event-time attribute provides what
> is
> > needed for an event-time temporal table join.
> > I agree that, from a technical point of view, the TEMPORAL keyword is not
> > required.
> >
> > I'm more sceptical about implicitly deriving the versioning information
> of
> > a (temporal) table as the table's only event-time attribute.
> > In the query
> >
> > SELECT *
> > FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
> > WHERE o.currency = r.currency
> >
> > the syntax of the temporal table join does not explicitly reference the
> > version of the temporal rates table.
> > Hence, the system needs a way to derive the version of temporal table.
> >
> > Implicitly using the (only) event-time attribute of a temporal table
> (rates
> > in the example above) to identify the right version works in most cases,
> > but probably not in all.
> > * What if a table has more than one event-time attribute? (TableSchema is
> > designed to support multiple watermarks; queries with interval joins
> > produce tables with multiple event-time attributes, ...)
> > * What if the table does not have an event-time attribute in its schema
> but
> > the version should only be provided as meta data?
> >
> > We could add a clause to define the version of a table, such as:
> >
> > CREATE TABLE rates (
> >    currency CHAR(3) NOT NULL PRIMARY KEY,
> >    rate DOUBLE,
> >    rowtime TIMESTAMP,
> >    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
> > VERSION (rowtime)
> > WITH (...);
> >
> > The presence of a the VERSION clause (or whatever syntax) would
> explicitly
> > define the version of a (temporal) table.
> > It would also render the need for the TEMPORAL keyword superfluous
> because
> > there would be another indicator that a table can be used in a temporal
> > table join.
> >
> > I'm OK with not adding the TEMPORAL keyword, but I recommend that we
> think
> > again about the proposed implicit definition of a table's version and how
> > it might limit use in the future.
> >
> > Cheers,
> > Fabian
> >
> > Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:
> >
> > > I'm also +1 for not adding the TEMPORAL keyword.
> > >
> > > +1 to make the PRIMARY KEY semantic clear for sources.
> > > From my point of view:
> > >
> > > 1) PRIMARY KEY on changelog souruce:
> > > It means that when the changelogs (INSERT/UPDATE/DELETE) are
> > materialized,
> > > the materialized table should be unique on the primary key columns.
> > > Flink assumes messages are in order on the primary key. Flink doesn't
> > > validate/enforces the key integrity, but simply trust it (thus NOT
> > > ENFORCED).
> > > Flink will use the PRIMARY KEY for some optimization, e.g. use the
> > PRIMARY
> > > KEY to update the materilized state by key in temporal join operator.
> > >
> > > 2) PRIMARY KEY on insert-only source:
> > > I prefer to have the same semantic to the batch source and changelog
> > > source, that it implies that records are not duplicate on the primary
> > key.
> > > Flink just simply trust the primary key constraint, and doesn't valid
> it.
> > > If there is duplicate primary keys with INSERT changeflag, then result
> of
> > > Flink query might be wrong.
> > >
> > > If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
> > duplicate
> > > primary keys with INSERT changeflag, when we migrate this case to
> > temporal
> > > table DDL,
> > > I think this source should emit INSERT/UPDATE (UPSERT) messages instead
> > of
> > > INSERT-only messages,  e.g. a Kafka compacted topic source?
> > >
> > > Best,
> > > Jark
> > >
> > >
> > > On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
> > wrote:
> > >
> > > > Hi everyone,
> > > >
> > > > I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
> > > >
> > > > Best,
> > > >
> > > > Konstantin
> > > >
> > > > On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]>
> wrote:
> > > >
> > > > > I agree with Timo, semantic about primary key needs more thought
> and
> > > > > discussion, especially after FLIP-95 and FLIP-105.
> > > > >
> > > > > Best,
> > > > > Kurt
> > > > >
> > > > >
> > > > > On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
> > > wrote:
> > > > >
> > > > > > Hi Leonard,
> > > > > >
> > > > > > thanks for the summary.
> > > > > >
> > > > > > After reading all of the previous arguments and working on
> > FLIP-95. I
> > > > > > would also lean towards the conclusion of not adding the TEMPORAL
> > > > > keyword.
> > > > > >
> > > > > > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can
> be
> > > > > > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
> > FOR
> > > > > > SYSTEM_TIME AS OF t would trigger the internal materialization
> and
> > > > > > "temporal" logic.
> > > > > >
> > > > > > However, we should discuss the meaning of PRIMARY KEY again in
> this
> > > > > > case. In a TEMPORAL TABLE scenario, the source would emit
> duplicate
> > > > > > primary keys with INSERT changeflag but at different point in
> time.
> > > > > > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > > > > > changelog semantics of FLIP-95 and FLIP-105 don't work well with
> a
> > > > > > primary key declaration.
> > > > > >
> > > > > > Regards,
> > > > > > Timo
> > > > > >
> > > > > >
> > > > > > On 20.06.20 17:08, Leonard Xu wrote:
> > > > > > > Hi everyone,
> > > > > > >
> > > > > > > Thanks for the nice discussion. I’d like to move forward the
> > work,
> > > > > > please let me simply summarize the main opinion and current
> > > > divergences.
> > > > > > >
> > > > > > > 1. The agreements have been achieved:
> > > > > > >
> > > > > > > 1.1 The motivation we're discussing temporal table DDL is just
> > for
> > > > > > creating temporal table in pure SQL to replace pre-process
> temporal
> > > > table
> > > > > > in YAML/Table API for usability.
> > > > > > > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD
> FOR
> > > > > > SYSTEM_TIME” is to make user understand easily.
> > > > > > > 1.3 For append-only table, it can convert to changelog table
> > which
> > > > has
> > > > > > been discussed in FLIP-105, we assume the following temporal
> table
> > is
> > > > > comes
> > > > > > from changelog (Jark, fabian, Timo).
> > > > > > > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
> > > instead
> > > > > of
> > > > > > the current `LATERAL TABLE(rates(x))`  has come to an
> > > agreement(Fabian,
> > > > > > Timo, Seth, Konstantin, Kurt).
> > > > > > >
> > > > > > > 2. The small divergence :
> > > > > > >
> > > > > > > About the definition syntax of the temporal table,
> > > > > > >
> > > > > > > CREATE [TEMPORAL] TABLE rates (
> > > > > > >     currency CHAR(3) NOT NULL PRIMARY KEY,
> > > > > > >     rate DOUBLE,
> > > > > > >     rowtime TIMESTAMP,
> > > > > > >     WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > > > > > > WITH (...);
> > > > > > >
> > > > > > > there is small divergence whether add "TEMPORAL" keyword or
> not.
> > > > > > >
> > > > > > > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo,
> Fabian,
> > > > Seth),
> > > > > > the main advantages are:
> > > > > > > （1）"TEMPORAL" keyword is intuitive to indicate the history
> > tracking
> > > > > > semantics.
> > > > > > > （2）"TEMPORAL" keyword illustrates that queries can visit the
> > > previous
> > > > > > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
> > > > keyword.
> > > > > > >
> > > > > > > 2.2 the other is using "CREATE TABLE"(Kurt), the main
> advantages
> > > are:
> > > > > > > （1）Just primary key and time attribute can track previous
> > versions
> > > > of a
> > > > > > table well.
> > > > > > > （2）The temporal behavior is triggered by temporal join syntax
> > > rather
> > > > > > than in DDL, all Flink DDL table are dynamic table logically
> > > including
> > > > > > temporal table. If we decide to use "TEMPORAL" keyword and treats
> > > > > changelog
> > > > > > as temporal table, other tables backed queue like Kafka should
> also
> > > use
> > > > > > "TEMPORAL" keyword.
> > > > > > >
> > > > > > >
> > > > > > > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
> > > with
> > > > > 2.1
> > > > > > may confuse users much. If we take a second to think about, for
> > > > > source/sink
> > > > > > table which may backed queue (like kafka) or DB (like MySQL), we
> > did
> > > > not
> > > > > > add any keyword in DDL to specify they are source or sinks, it
> > works
> > > > > well.
> > > > > > > I think temporal table is the third one,  kafka data source and
> > DB
> > > > data
> > > > > > source can play as a source/sink/temporal table depends on the
> > > > > > position/syntax that user put them in the query. The above rates
> > > table
> > > > > > >      - can be a source table if user put it at `SELECT * FROM
> > > rates;`
> > > > > > >      - can be a temporal table if user put it at `SELECT * FROM
> > > > orders
> > > > > > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> > > > > > >               ON orders.currency = rates.currency;`
> > > > > > >      - can be sink table if user put is at `INSERT INTO rates
> > > SELECT
> > > > *
> > > > > > FROM …; `
> > > > > > >  From these cases, we found all tables defined in Flink should
> be
> > > > > > dynamic table logically, the source/sink/temporal role depends on
> > the
> > > > > > position/syntax in user’s query.
> > > > > > >        In fact we have used similar syntax for current lookup
> > > table,
> > > > we
> > > > > > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
> > > trigger
> > > > > the
> > > > > > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
> > in
> > > > > query.
> > > > > > >
> > > > > > > So, I prefer to resolve the small divergence with “CREATE
> TABLE”
> > > > which
> > > > > > > (1) is more unified with our source/sink/temporal dynamic table
> > > > > > conceptually,
> > > > > > > (2) is aligned with current lookup table,
> > > > > > > (3) also make users learn less keyword.
> > > > > > >
> > > > > > > WDYT？
> > > > > > >
> > > > > > > Best,
> > > > > > > Leonard Xu
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > >
> > > > Konstantin Knauf
> > > >
> > > > https://twitter.com/snntrable
> > > >
> > > > https://github.com/knaufk
> > > >
> > >
> >
>

[Leonard Xu]

Hi, everyone

Thanks Fabian,Kurt for making the multiple version(event time) clear, I also like the 'PERIOD FOR SYSTEM' syntax which supported in SQL standard. I think we can add some explanation of the multiple version support in the future section of FLIP.

For the PRIMARY KEY semantic, I agree with Jark's point that the semantic should unify both on changelog source and insert-only source.

Currently, Flink supports PRIMARY KEY after FLIP-87, Flink uses PRIMARY KEY NOT ENFORCED because Flink does not own the data like other DBMS therefore Flink won't validate/enforce the key integrity and only trusts the external systems. It is  expected user and external system/application should make sure no deduplicated records happened when using NOT ENFORCED.

(a) For PRIMARY KEY NOT ENFORCED semantic on changelog source:
It means the materialized changelogs (INSERT/UPDATE/DELETE) should be unique on the primary key constraints.Flink assumes messages are in order on the primary key. Flink will use the PRIMARY KEY for some optimization, e.g. use the PRIMARY KEY to update the materialized state by key in temporal join operator.
 
(b) For PRIMARY KEY NOT ENFORCED semantic on insert-only source:
It means records should be unique on the primary key constraints. If there are INSERT records with duplicate primary key columns, the result of SQL query might be nondeterministic because it broken the PRIMARY KEY constraints.

Cheers,
Leonard

> 在 2020年6月23日，23:35，Fabian Hueske <[hidden email]> 写道：
>
> Thanks Kurt,
>
> Yes, you are right.
> The `PERIOD FOR SYSTEM_TIME` that you linked before corresponds to the
> VERSION clause that I used and would explicitly define the versioning of a
> table.
> I didn't know that the `PERIOD FOR SYSTEM_TIME` cause is already defined by
> the SQL standard.
> I think we would need a slightly different syntax though because (so far)
> the validity of a row is determined by its own timestamp and the timestamp
> of the next row.
>
> Adding a clause later solves the ambiguity issue for tables with multiple
> event-time attributes.
> However, I'd feel more comfortable having such a cause and an explicit
> definition of the temporal property from the beginning.
> I guess this is a matter of personal preference so I'll go with the
> majority if we decide that every table that has a primary key and an
> event-time attribute should be usable in an event-time temporal table join.
>
> Thanks, Fabian
>
>
> Am Di., 23. Juni 2020 um 16:58 Uhr schrieb Kurt Young <[hidden email]>:
>
>> Hi Fabian,
>>
>> I agree with you that implicitly letting event time to be the version of
>> the table will
>> work in most cases, but not for all. That's the reason I mentioned `PERIOD
>> FOR` [1]
>> syntax in my first email, which is already in sql standard to represent the
>> validity of
>> each row in the table.
>>
>> If the event time can't be used, or multiple event time are defined, we
>> could still add
>> this syntax in the future.
>>
>> What do you think?
>>
>> [1]
>>
>> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
>> Best,
>> Kurt
>>
>>
>> On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]> wrote:
>>
>>> Hi everyone,
>>>
>>> Every table with a primary key and an event-time attribute provides what
>> is
>>> needed for an event-time temporal table join.
>>> I agree that, from a technical point of view, the TEMPORAL keyword is not
>>> required.
>>>
>>> I'm more sceptical about implicitly deriving the versioning information
>> of
>>> a (temporal) table as the table's only event-time attribute.
>>> In the query
>>>
>>> SELECT *
>>> FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
>>> WHERE o.currency = r.currency
>>>
>>> the syntax of the temporal table join does not explicitly reference the
>>> version of the temporal rates table.
>>> Hence, the system needs a way to derive the version of temporal table.
>>>
>>> Implicitly using the (only) event-time attribute of a temporal table
>> (rates
>>> in the example above) to identify the right version works in most cases,
>>> but probably not in all.
>>> * What if a table has more than one event-time attribute? (TableSchema is
>>> designed to support multiple watermarks; queries with interval joins
>>> produce tables with multiple event-time attributes, ...)
>>> * What if the table does not have an event-time attribute in its schema
>> but
>>> the version should only be provided as meta data?
>>>
>>> We could add a clause to define the version of a table, such as:
>>>
>>> CREATE TABLE rates (
>>>   currency CHAR(3) NOT NULL PRIMARY KEY,
>>>   rate DOUBLE,
>>>   rowtime TIMESTAMP,
>>>   WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
>>> VERSION (rowtime)
>>> WITH (...);
>>>
>>> The presence of a the VERSION clause (or whatever syntax) would
>> explicitly
>>> define the version of a (temporal) table.
>>> It would also render the need for the TEMPORAL keyword superfluous
>> because
>>> there would be another indicator that a table can be used in a temporal
>>> table join.
>>>
>>> I'm OK with not adding the TEMPORAL keyword, but I recommend that we
>> think
>>> again about the proposed implicit definition of a table's version and how
>>> it might limit use in the future.
>>>
>>> Cheers,
>>> Fabian
>>>
>>> Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:
>>>
>>>> I'm also +1 for not adding the TEMPORAL keyword.
>>>>
>>>> +1 to make the PRIMARY KEY semantic clear for sources.
>>>> From my point of view:
>>>>
>>>> 1) PRIMARY KEY on changelog souruce:
>>>> It means that when the changelogs (INSERT/UPDATE/DELETE) are
>>> materialized,
>>>> the materialized table should be unique on the primary key columns.
>>>> Flink assumes messages are in order on the primary key. Flink doesn't
>>>> validate/enforces the key integrity, but simply trust it (thus NOT
>>>> ENFORCED).
>>>> Flink will use the PRIMARY KEY for some optimization, e.g. use the
>>> PRIMARY
>>>> KEY to update the materilized state by key in temporal join operator.
>>>>
>>>> 2) PRIMARY KEY on insert-only source:
>>>> I prefer to have the same semantic to the batch source and changelog
>>>> source, that it implies that records are not duplicate on the primary
>>> key.
>>>> Flink just simply trust the primary key constraint, and doesn't valid
>> it.
>>>> If there is duplicate primary keys with INSERT changeflag, then result
>> of
>>>> Flink query might be wrong.
>>>>
>>>> If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
>>> duplicate
>>>> primary keys with INSERT changeflag, when we migrate this case to
>>> temporal
>>>> table DDL,
>>>> I think this source should emit INSERT/UPDATE (UPSERT) messages instead
>>> of
>>>> INSERT-only messages,  e.g. a Kafka compacted topic source?
>>>>
>>>> Best,
>>>> Jark
>>>>
>>>>
>>>> On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
>>> wrote:
>>>>
>>>>> Hi everyone,
>>>>>
>>>>> I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
>>>>>
>>>>> Best,
>>>>>
>>>>> Konstantin
>>>>>
>>>>> On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]>
>> wrote:
>>>>>
>>>>>> I agree with Timo, semantic about primary key needs more thought
>> and
>>>>>> discussion, especially after FLIP-95 and FLIP-105.
>>>>>>
>>>>>> Best,
>>>>>> Kurt
>>>>>>
>>>>>>
>>>>>> On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
>>>> wrote:
>>>>>>
>>>>>>> Hi Leonard,
>>>>>>>
>>>>>>> thanks for the summary.
>>>>>>>
>>>>>>> After reading all of the previous arguments and working on
>>> FLIP-95. I
>>>>>>> would also lean towards the conclusion of not adding the TEMPORAL
>>>>>> keyword.
>>>>>>>
>>>>>>> After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can
>> be
>>>>>>> represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
>>> FOR
>>>>>>> SYSTEM_TIME AS OF t would trigger the internal materialization
>> and
>>>>>>> "temporal" logic.
>>>>>>>
>>>>>>> However, we should discuss the meaning of PRIMARY KEY again in
>> this
>>>>>>> case. In a TEMPORAL TABLE scenario, the source would emit
>> duplicate
>>>>>>> primary keys with INSERT changeflag but at different point in
>> time.
>>>>>>> Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
>>>>>>> changelog semantics of FLIP-95 and FLIP-105 don't work well with
>> a
>>>>>>> primary key declaration.
>>>>>>>
>>>>>>> Regards,
>>>>>>> Timo
>>>>>>>
>>>>>>>
>>>>>>> On 20.06.20 17:08, Leonard Xu wrote:
>>>>>>>> Hi everyone,
>>>>>>>>
>>>>>>>> Thanks for the nice discussion. I’d like to move forward the
>>> work,
>>>>>>> please let me simply summarize the main opinion and current
>>>>> divergences.
>>>>>>>>
>>>>>>>> 1. The agreements have been achieved:
>>>>>>>>
>>>>>>>> 1.1 The motivation we're discussing temporal table DDL is just
>>> for
>>>>>>> creating temporal table in pure SQL to replace pre-process
>> temporal
>>>>> table
>>>>>>> in YAML/Table API for usability.
>>>>>>>> 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD
>> FOR
>>>>>>> SYSTEM_TIME” is to make user understand easily.
>>>>>>>> 1.3 For append-only table, it can convert to changelog table
>>> which
>>>>> has
>>>>>>> been discussed in FLIP-105, we assume the following temporal
>> table
>>> is
>>>>>> comes
>>>>>>> from changelog (Jark, fabian, Timo).
>>>>>>>> 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
>>>> instead
>>>>>> of
>>>>>>> the current `LATERAL TABLE(rates(x))`  has come to an
>>>> agreement(Fabian,
>>>>>>> Timo, Seth, Konstantin, Kurt).
>>>>>>>>
>>>>>>>> 2. The small divergence :
>>>>>>>>
>>>>>>>> About the definition syntax of the temporal table,
>>>>>>>>
>>>>>>>> CREATE [TEMPORAL] TABLE rates (
>>>>>>>>    currency CHAR(3) NOT NULL PRIMARY KEY,
>>>>>>>>    rate DOUBLE,
>>>>>>>>    rowtime TIMESTAMP,
>>>>>>>>    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
>>>>>>>> WITH (...);
>>>>>>>>
>>>>>>>> there is small divergence whether add "TEMPORAL" keyword or
>> not.
>>>>>>>>
>>>>>>>> 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo,
>> Fabian,
>>>>> Seth),
>>>>>>> the main advantages are:
>>>>>>>> （1）"TEMPORAL" keyword is intuitive to indicate the history
>>> tracking
>>>>>>> semantics.
>>>>>>>> （2）"TEMPORAL" keyword illustrates that queries can visit the
>>>> previous
>>>>>>> versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
>>>>> keyword.
>>>>>>>>
>>>>>>>> 2.2 the other is using "CREATE TABLE"(Kurt), the main
>> advantages
>>>> are:
>>>>>>>> （1）Just primary key and time attribute can track previous
>>> versions
>>>>> of a
>>>>>>> table well.
>>>>>>>> （2）The temporal behavior is triggered by temporal join syntax
>>>> rather
>>>>>>> than in DDL, all Flink DDL table are dynamic table logically
>>>> including
>>>>>>> temporal table. If we decide to use "TEMPORAL" keyword and treats
>>>>>> changelog
>>>>>>> as temporal table, other tables backed queue like Kafka should
>> also
>>>> use
>>>>>>> "TEMPORAL" keyword.
>>>>>>>>
>>>>>>>>
>>>>>>>> IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
>>>> with
>>>>>> 2.1
>>>>>>> may confuse users much. If we take a second to think about, for
>>>>>> source/sink
>>>>>>> table which may backed queue (like kafka) or DB (like MySQL), we
>>> did
>>>>> not
>>>>>>> add any keyword in DDL to specify they are source or sinks, it
>>> works
>>>>>> well.
>>>>>>>> I think temporal table is the third one,  kafka data source and
>>> DB
>>>>> data
>>>>>>> source can play as a source/sink/temporal table depends on the
>>>>>>> position/syntax that user put them in the query. The above rates
>>>> table
>>>>>>>>     - can be a source table if user put it at `SELECT * FROM
>>>> rates;`
>>>>>>>>     - can be a temporal table if user put it at `SELECT * FROM
>>>>> orders
>>>>>>> JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
>>>>>>>>              ON orders.currency = rates.currency;`
>>>>>>>>     - can be sink table if user put is at `INSERT INTO rates
>>>> SELECT
>>>>> *
>>>>>>> FROM …; `
>>>>>>>> From these cases, we found all tables defined in Flink should
>> be
>>>>>>> dynamic table logically, the source/sink/temporal role depends on
>>> the
>>>>>>> position/syntax in user’s query.
>>>>>>>>       In fact we have used similar syntax for current lookup
>>>> table,
>>>>> we
>>>>>>> didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
>>>> trigger
>>>>>> the
>>>>>>> temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
>>> in
>>>>>> query.
>>>>>>>>
>>>>>>>> So, I prefer to resolve the small divergence with “CREATE
>> TABLE”
>>>>> which
>>>>>>>> (1) is more unified with our source/sink/temporal dynamic table
>>>>>>> conceptually,
>>>>>>>> (2) is aligned with current lookup table,
>>>>>>>> (3) also make users learn less keyword.
>>>>>>>>
>>>>>>>> WDYT？
>>>>>>>>
>>>>>>>> Best,
>>>>>>>> Leonard Xu
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>>
>>>>> Konstantin Knauf
>>>>>
>>>>> https://twitter.com/snntrable
>>>>>
>>>>> https://github.com/knaufk
>>>>>
>>>>
>>>
>>

[Leonard Xu]

Hi, kurt, Fabian

After an offline discussion with Jark, We think that the  'PERIOD FOR SYSTEM_TIME(operation_time)' statement might be needed now. Changelog table is superset of insert-only table, use PRIMARY KEY and rowtime may work well in insert-only or upsert source but has some problem in changelog table.

'PERIOD FOR SYSTEM_TIME(operation_time)' in a temporal table defines/maintains  the valid time of each row, the rowtime can not play the history tracking function well.

# 1.operation time (version time) vs rowtime (watermark)

I will take an example to explain. The following changelog records came from database table using debezium tool:
{ "before":  null
  "after":    {"currency": "Euro", "rate": 118, "gmt_modified": "12:00:01"},
  "op":       "c",  //INSERT
  "ts_ms": 1592971201000 // 2020-06-24 12:00:02
}
{ "before": {"currency": "Euro", "rate": 114, "gmt_modified": "12:00:05"},
  "after":    {"currency": "Euro", "rate": 118, "gmt_modified": "12:00:05"},
  "op":       "u",  //UPDATE
  "ts_ms": 1592971206000 // 2020-06-24 12:00:06
}

{ "before": {"currency": "Euro", "rate": 118, "gmt_modified": "12:00:05"},
  "after":     null,
  "op":        "d",  //DELETE
  "ts_ms":  1593000011000  // 2020-06-24 20:00:11
}

The rowtime should be the "gmt_modified" field that belongs to the original record，the "ts_ms" is the the operation time when the DML statement happen in the DB. For DELETE changelog record, its "gmt_modified" field (12:00:05) can not reflect the real operation time (20:00:11).

In temporal join case, we should maintain the valid time of each row. For a DELETE event, we should use the operation time of DELETE as the “end time” of the row. That says, the record {"currency": "Euro", "rate": 118} is not exist anymore after “20:00:11”, not “12:00:05”.

we would not access the record {"currency": "Euro", "rate": 118, "gmt_modified": "12:00:05"} when rowtime is bigger than (12:00:05) if we use rowtime to track the history version, because the DELETE changelog record also has rowtime (12:00:05) and will clear the record in state. In fact, the expected result is that the record expires until (20:00:11) when the record is deleted rather than the last update time(20:00:11) in materialized state.

From this case, I found rowtime and operation time should be orthogonal in temporal table scenario. The operation time should be strictly monotonically increasing  (no out of order) and only be used for tracking a history version of a changelog table, every history version of changelog table equals a database table snapshot due to the stream-table duality.

# 2.The semantic of rowtime and watermark on changelog

The rowtime and watermark can also be defined on a changelog table just like other source backed queue, Flink supports cascaded window aggregation (with ) in SQL like:
SELECT
     TUMBLE_ROWTIME(rowtime, INTERVAL '60' SECOND),
     MAX(rate) AS rate
FROM (
       SELECT
          MAX(rate) AS rate,
          TUMBLE_ROWTIME(rowtime, INTERVAL '5' SECOND) AS `rowtime`
       FROM currency
            GROUP BY TUMBLE(rowtime, INTERVAL '5' SECOND)
    )
    GROUP BY TUMBLE(rowtime, INTERVAL '60' SECOND

We can think of the output of the first window aggregation as a changelog source of the second window aggregation. There are INSERT/UPDATE/DELETE messages and also watermarks in the changelog stream. And the rowtime in the changelog stream is the `TUMBLE_ROWTIME` value (just like the `gmt_modified` column in DB).

#  summary
we should use ‘PERIOD FOR SYSTEM_TIME(operation_time) syntax to track history version by operation time rather than rowtime in temporal table scenario.
we also support define a rowtime(watermark) on changelog table, but the rowtime will not be used to track the history of changelog stream.  


WDYT? please correct me if I am wrong.


Best,
Leonard

> 在 2020年6月24日，11:31，Leonard Xu <[hidden email]> 写道：
>
> Hi, everyone
>
> Thanks Fabian,Kurt for making the multiple version(event time) clear, I also like the 'PERIOD FOR SYSTEM' syntax which supported in SQL standard. I think we can add some explanation of the multiple version support in the future section of FLIP.
>
> For the PRIMARY KEY semantic, I agree with Jark's point that the semantic should unify both on changelog source and insert-only source.
>
> Currently, Flink supports PRIMARY KEY after FLIP-87, Flink uses PRIMARY KEY NOT ENFORCED because Flink does not own the data like other DBMS therefore Flink won't validate/enforce the key integrity and only trusts the external systems. It is  expected user and external system/application should make sure no deduplicated records happened when using NOT ENFORCED.
>
> (a) For PRIMARY KEY NOT ENFORCED semantic on changelog source:
> It means the materialized changelogs (INSERT/UPDATE/DELETE) should be unique on the primary key constraints.Flink assumes messages are in order on the primary key. Flink will use the PRIMARY KEY for some optimization, e.g. use the PRIMARY KEY to update the materialized state by key in temporal join operator.
>  
> (b) For PRIMARY KEY NOT ENFORCED semantic on insert-only source:
> It means records should be unique on the primary key constraints. If there are INSERT records with duplicate primary key columns, the result of SQL query might be nondeterministic because it broken the PRIMARY KEY constraints.
>
> Cheers,
> Leonard
>
>
>> 在 2020年6月23日，23:35，Fabian Hueske <[hidden email] <mailto:[hidden email]>> 写道：
>>
>> Thanks Kurt,
>>
>> Yes, you are right.
>> The `PERIOD FOR SYSTEM_TIME` that you linked before corresponds to the
>> VERSION clause that I used and would explicitly define the versioning of a
>> table.
>> I didn't know that the `PERIOD FOR SYSTEM_TIME` cause is already defined by
>> the SQL standard.
>> I think we would need a slightly different syntax though because (so far)
>> the validity of a row is determined by its own timestamp and the timestamp
>> of the next row.
>>
>> Adding a clause later solves the ambiguity issue for tables with multiple
>> event-time attributes.
>> However, I'd feel more comfortable having such a cause and an explicit
>> definition of the temporal property from the beginning.
>> I guess this is a matter of personal preference so I'll go with the
>> majority if we decide that every table that has a primary key and an
>> event-time attribute should be usable in an event-time temporal table join.
>>
>> Thanks, Fabian
>>
>>
>> Am Di., 23. Juni 2020 um 16:58 Uhr schrieb Kurt Young <[hidden email] <mailto:[hidden email]>>:
>>
>>> Hi Fabian,
>>>
>>> I agree with you that implicitly letting event time to be the version of
>>> the table will
>>> work in most cases, but not for all. That's the reason I mentioned `PERIOD
>>> FOR` [1]
>>> syntax in my first email, which is already in sql standard to represent the
>>> validity of
>>> each row in the table.
>>>
>>> If the event time can't be used, or multiple event time are defined, we
>>> could still add
>>> this syntax in the future.
>>>
>>> What do you think?
>>>
>>> [1]
>>>
>>> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15 <https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15>
>>> Best,
>>> Kurt
>>>
>>>
>>> On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]> wrote:
>>>
>>>> Hi everyone,
>>>>
>>>> Every table with a primary key and an event-time attribute provides what
>>> is
>>>> needed for an event-time temporal table join.
>>>> I agree that, from a technical point of view, the TEMPORAL keyword is not
>>>> required.
>>>>
>>>> I'm more sceptical about implicitly deriving the versioning information
>>> of
>>>> a (temporal) table as the table's only event-time attribute.
>>>> In the query
>>>>
>>>> SELECT *
>>>> FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
>>>> WHERE o.currency = r.currency
>>>>
>>>> the syntax of the temporal table join does not explicitly reference the
>>>> version of the temporal rates table.
>>>> Hence, the system needs a way to derive the version of temporal table.
>>>>
>>>> Implicitly using the (only) event-time attribute of a temporal table
>>> (rates
>>>> in the example above) to identify the right version works in most cases,
>>>> but probably not in all.
>>>> * What if a table has more than one event-time attribute? (TableSchema is
>>>> designed to support multiple watermarks; queries with interval joins
>>>> produce tables with multiple event-time attributes, ...)
>>>> * What if the table does not have an event-time attribute in its schema
>>> but
>>>> the version should only be provided as meta data?
>>>>
>>>> We could add a clause to define the version of a table, such as:
>>>>
>>>> CREATE TABLE rates (
>>>>   currency CHAR(3) NOT NULL PRIMARY KEY,
>>>>   rate DOUBLE,
>>>>   rowtime TIMESTAMP,
>>>>   WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
>>>> VERSION (rowtime)
>>>> WITH (...);
>>>>
>>>> The presence of a the VERSION clause (or whatever syntax) would
>>> explicitly
>>>> define the version of a (temporal) table.
>>>> It would also render the need for the TEMPORAL keyword superfluous
>>> because
>>>> there would be another indicator that a table can be used in a temporal
>>>> table join.
>>>>
>>>> I'm OK with not adding the TEMPORAL keyword, but I recommend that we
>>> think
>>>> again about the proposed implicit definition of a table's version and how
>>>> it might limit use in the future.
>>>>
>>>> Cheers,
>>>> Fabian
>>>>
>>>> Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:
>>>>
>>>>> I'm also +1 for not adding the TEMPORAL keyword.
>>>>>
>>>>> +1 to make the PRIMARY KEY semantic clear for sources.
>>>>> From my point of view:
>>>>>
>>>>> 1) PRIMARY KEY on changelog souruce:
>>>>> It means that when the changelogs (INSERT/UPDATE/DELETE) are
>>>> materialized,
>>>>> the materialized table should be unique on the primary key columns.
>>>>> Flink assumes messages are in order on the primary key. Flink doesn't
>>>>> validate/enforces the key integrity, but simply trust it (thus NOT
>>>>> ENFORCED).
>>>>> Flink will use the PRIMARY KEY for some optimization, e.g. use the
>>>> PRIMARY
>>>>> KEY to update the materilized state by key in temporal join operator.
>>>>>
>>>>> 2) PRIMARY KEY on insert-only source:
>>>>> I prefer to have the same semantic to the batch source and changelog
>>>>> source, that it implies that records are not duplicate on the primary
>>>> key.
>>>>> Flink just simply trust the primary key constraint, and doesn't valid
>>> it.
>>>>> If there is duplicate primary keys with INSERT changeflag, then result
>>> of
>>>>> Flink query might be wrong.
>>>>>
>>>>> If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
>>>> duplicate
>>>>> primary keys with INSERT changeflag, when we migrate this case to
>>>> temporal
>>>>> table DDL,
>>>>> I think this source should emit INSERT/UPDATE (UPSERT) messages instead
>>>> of
>>>>> INSERT-only messages,  e.g. a Kafka compacted topic source?
>>>>>
>>>>> Best,
>>>>> Jark
>>>>>
>>>>>
>>>>> On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
>>>> wrote:
>>>>>
>>>>>> Hi everyone,
>>>>>>
>>>>>> I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
>>>>>>
>>>>>> Best,
>>>>>>
>>>>>> Konstantin
>>>>>>
>>>>>> On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]>
>>> wrote:
>>>>>>
>>>>>>> I agree with Timo, semantic about primary key needs more thought
>>> and
>>>>>>> discussion, especially after FLIP-95 and FLIP-105.
>>>>>>>
>>>>>>> Best,
>>>>>>> Kurt
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
>>>>> wrote:
>>>>>>>
>>>>>>>> Hi Leonard,
>>>>>>>>
>>>>>>>> thanks for the summary.
>>>>>>>>
>>>>>>>> After reading all of the previous arguments and working on
>>>> FLIP-95. I
>>>>>>>> would also lean towards the conclusion of not adding the TEMPORAL
>>>>>>> keyword.
>>>>>>>>
>>>>>>>> After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can
>>> be
>>>>>>>> represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
>>>> FOR
>>>>>>>> SYSTEM_TIME AS OF t would trigger the internal materialization
>>> and
>>>>>>>> "temporal" logic.
>>>>>>>>
>>>>>>>> However, we should discuss the meaning of PRIMARY KEY again in
>>> this
>>>>>>>> case. In a TEMPORAL TABLE scenario, the source would emit
>>> duplicate
>>>>>>>> primary keys with INSERT changeflag but at different point in
>>> time.
>>>>>>>> Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
>>>>>>>> changelog semantics of FLIP-95 and FLIP-105 don't work well with
>>> a
>>>>>>>> primary key declaration.
>>>>>>>>
>>>>>>>> Regards,
>>>>>>>> Timo
>>>>>>>>
>>>>>>>>
>>>>>>>> On 20.06.20 17:08, Leonard Xu wrote:
>>>>>>>>> Hi everyone,
>>>>>>>>>
>>>>>>>>> Thanks for the nice discussion. I’d like to move forward the
>>>> work,
>>>>>>>> please let me simply summarize the main opinion and current
>>>>>> divergences.
>>>>>>>>>
>>>>>>>>> 1. The agreements have been achieved:
>>>>>>>>>
>>>>>>>>> 1.1 The motivation we're discussing temporal table DDL is just
>>>> for
>>>>>>>> creating temporal table in pure SQL to replace pre-process
>>> temporal
>>>>>> table
>>>>>>>> in YAML/Table API for usability.
>>>>>>>>> 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD
>>> FOR
>>>>>>>> SYSTEM_TIME” is to make user understand easily.
>>>>>>>>> 1.3 For append-only table, it can convert to changelog table
>>>> which
>>>>>> has
>>>>>>>> been discussed in FLIP-105, we assume the following temporal
>>> table
>>>> is
>>>>>>> comes
>>>>>>>> from changelog (Jark, fabian, Timo).
>>>>>>>>> 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
>>>>> instead
>>>>>>> of
>>>>>>>> the current `LATERAL TABLE(rates(x))`  has come to an
>>>>> agreement(Fabian,
>>>>>>>> Timo, Seth, Konstantin, Kurt).
>>>>>>>>>
>>>>>>>>> 2. The small divergence :
>>>>>>>>>
>>>>>>>>> About the definition syntax of the temporal table,
>>>>>>>>>
>>>>>>>>> CREATE [TEMPORAL] TABLE rates (
>>>>>>>>>    currency CHAR(3) NOT NULL PRIMARY KEY,
>>>>>>>>>    rate DOUBLE,
>>>>>>>>>    rowtime TIMESTAMP,
>>>>>>>>>    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
>>>>>>>>> WITH (...);
>>>>>>>>>
>>>>>>>>> there is small divergence whether add "TEMPORAL" keyword or
>>> not.
>>>>>>>>>
>>>>>>>>> 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo,
>>> Fabian,
>>>>>> Seth),
>>>>>>>> the main advantages are:
>>>>>>>>> （1）"TEMPORAL" keyword is intuitive to indicate the history
>>>> tracking
>>>>>>>> semantics.
>>>>>>>>> （2）"TEMPORAL" keyword illustrates that queries can visit the
>>>>> previous
>>>>>>>> versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
>>>>>> keyword.
>>>>>>>>>
>>>>>>>>> 2.2 the other is using "CREATE TABLE"(Kurt), the main
>>> advantages
>>>>> are:
>>>>>>>>> （1）Just primary key and time attribute can track previous
>>>> versions
>>>>>> of a
>>>>>>>> table well.
>>>>>>>>> （2）The temporal behavior is triggered by temporal join syntax
>>>>> rather
>>>>>>>> than in DDL, all Flink DDL table are dynamic table logically
>>>>> including
>>>>>>>> temporal table. If we decide to use "TEMPORAL" keyword and treats
>>>>>>> changelog
>>>>>>>> as temporal table, other tables backed queue like Kafka should
>>> also
>>>>> use
>>>>>>>> "TEMPORAL" keyword.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
>>>>> with
>>>>>>> 2.1
>>>>>>>> may confuse users much. If we take a second to think about, for
>>>>>>> source/sink
>>>>>>>> table which may backed queue (like kafka) or DB (like MySQL), we
>>>> did
>>>>>> not
>>>>>>>> add any keyword in DDL to specify they are source or sinks, it
>>>> works
>>>>>>> well.
>>>>>>>>> I think temporal table is the third one,  kafka data source and
>>>> DB
>>>>>> data
>>>>>>>> source can play as a source/sink/temporal table depends on the
>>>>>>>> position/syntax that user put them in the query. The above rates
>>>>> table
>>>>>>>>>     - can be a source table if user put it at `SELECT * FROM
>>>>> rates;`
>>>>>>>>>     - can be a temporal table if user put it at `SELECT * FROM
>>>>>> orders
>>>>>>>> JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
>>>>>>>>>              ON orders.currency = rates.currency;`
>>>>>>>>>     - can be sink table if user put is at `INSERT INTO rates
>>>>> SELECT
>>>>>> *
>>>>>>>> FROM …; `
>>>>>>>>> From these cases, we found all tables defined in Flink should
>>> be
>>>>>>>> dynamic table logically, the source/sink/temporal role depends on
>>>> the
>>>>>>>> position/syntax in user’s query.
>>>>>>>>>       In fact we have used similar syntax for current lookup
>>>>> table,
>>>>>> we
>>>>>>>> didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
>>>>> trigger
>>>>>>> the
>>>>>>>> temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
>>>> in
>>>>>>> query.
>>>>>>>>>
>>>>>>>>> So, I prefer to resolve the small divergence with “CREATE
>>> TABLE”
>>>>>> which
>>>>>>>>> (1) is more unified with our source/sink/temporal dynamic table
>>>>>>>> conceptually,
>>>>>>>>> (2) is aligned with current lookup table,
>>>>>>>>> (3) also make users learn less keyword.
>>>>>>>>>
>>>>>>>>> WDYT？
>>>>>>>>>
>>>>>>>>> Best,
>>>>>>>>> Leonard Xu
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>>
>>>>>> Konstantin Knauf
>>>>>>
>>>>>> https://twitter.com/snntrable
>>>>>>
>>>>>> https://github.com/knaufk
>>>>>>
>>>>>
>>>>
>>>
>

[Jark Wu-2]

Hi all,

Thanks Leonard for summarizing our discussion. I want to share more of my
thoughts:

* rowtime is a column in the its schema, so the rowtime of DELETE event is
the value of the previous image.
* operation time is the time when the DML statements happen in databases,
so the operation time of DELETE events is the time when it happens.
* rowtime can't be used as operation time for history tracking
* operation time can't be used as rowtime (can't apply window on the
operation time)
* rowtime and operation time are orthogonal concepts and used in different
scenarios.
* operation time implicitly means it is monotonically increasing, we don't
need watermark syntax to specify the out of boundness for it.

======================================================================
So, conclusion from my side so far:

* watermark/rowtime + primary key + changelog source != versioned temporal
table
* operation time + primary key + changelog source == versioned temporal
table
* We may need something like 'PERIOD FOR SYSTEM_TIME(op_ts)' to define the
operation time

======================================================================
However, there is still a pending question I don't have answer:

Assuming you are doing a MIN aggregate on the operation time, that doesn't
work because the DELETE/UPDATE_BEFORE doesn't hold
the previous value of operation time and thus can't retract.

The operation time in fact should be metadata information (just like
RowKind) which shouldn't be in the shema, and can't be accessed in queries.
But the PERIOD FOR SYSTEM_TIME syntax is in the schema part and should
refer to a field in the schema...

======================================================================

Anyway, let's focus on the operation_time vs rowtime problem first. Let me
know what's your thought!

Best,
Jark

On Wed, 24 Jun 2020 at 23:49, Leonard Xu <[hidden email]> wrote:

> Hi, kurt, Fabian
>
> After an offline discussion with Jark, We think that the  'PERIOD FOR
> SYSTEM_TIME(operation_time)' statement might be needed now. Changelog table
> is superset of insert-only table, use PRIMARY KEY and rowtime may work well
> in insert-only or upsert source but has some problem in changelog table.
>
> 'PERIOD FOR SYSTEM_TIME(operation_time)' in a temporal table
> defines/maintains  the valid time of each row, the rowtime can not play the
> history tracking function well.
>
> *# 1.*operation time (version time) *vs* rowtime (watermark)
>
> I will take an example to explain. The following changelog records came
> from database table using debezium tool:
> { "before":  null
>   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> "12:00:01"},
>   "op":       "c",  //INSERT
>   "ts_ms": 1592971201000 // 2020-06-24 12:00:02
> }
> { "before": {"currency": "Euro", "rate": 114, "gmt_modified": "12:00:05"},
>   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> "12:00:05"},
>   "op":       "u",  //UPDATE
>   "ts_ms": 1592971206000 // 2020-06-24 12:00:06
> }
>
> { "before": {"currency": "Euro", "rate": 118, "gmt_modified": "12:00:05"},
>   "after":     null,
>   "op":        "d",  //DELETE
>   "ts_ms":  1593000011000  // 2020-06-24 20:00:11
> }
>
> The rowtime should be the "gmt_modified" field that belongs to the
> original record，the "ts_ms" is the the operation time when the DML
> statement happen in the DB. For DELETE changelog record, its "gmt_modified"
> field (12:00:05) can not reflect the real operation time (20:00:11).
>
> In temporal join case, we should maintain the valid time of each row. For
> a DELETE event, we should use the operation time of DELETE as the “end
> time” of the row. That says, the record {"currency": "Euro", "rate": 118}
> is not exist anymore after “20:00:11”, not “12:00:05”.
>
> we would not access the record {"currency": "Euro", "rate": 118,
> "gmt_modified": "12:00:05"} when rowtime is bigger than (12:00:05) if we
> use rowtime to track the history version, because the DELETE changelog
> record also has rowtime (12:00:05) and will clear the record in state. In
> fact, the expected result is that the record expires until (20:00:11) when
> the record is deleted rather than the last update time(20:00:11) in
> materialized state.
>
> From this case, I found rowtime and operation time should be orthogonal in
> temporal table scenario. The operation time should be strictly
> monotonically increasing  (no out of order) and only be used for tracking a
> history version of a changelog table, every history version of changelog
> table equals a database table snapshot due to the stream-table duality.
>
> *# 2.*The semantic of rowtime and watermark on changelog
>
> The rowtime and watermark can also be defined on a changelog table just
> like other source backed queue, Flink supports cascaded window aggregation
> (with ) in SQL like:
> SELECT
>      TUMBLE_ROWTIME(rowtime, INTERVAL '60' SECOND),
>      MAX(rate) AS rate
> FROM (
>        SELECT
>           MAX(rate) AS rate,
>           TUMBLE_ROWTIME(rowtime, INTERVAL '5' SECOND) AS `rowtime`
>        FROM currency
>             GROUP BY TUMBLE(rowtime, INTERVAL '5' SECOND)
>     )
>     GROUP BY TUMBLE(rowtime, INTERVAL '60' SECOND
>
> We can think of the output of the first window aggregation as a changelog
> source of the second window aggregation. There are INSERT/UPDATE/DELETE
> messages and also watermarks in the changelog stream. And the rowtime in
> the changelog stream is the `TUMBLE_ROWTIME` value (just like the
> `gmt_modified` column in DB).
>
> *#  summary*
>
>    1. we should use ‘PERIOD FOR SYSTEM_TIME(operation_time) syntax to
>    track history version by operation time rather than rowtime in temporal
>    table scenario.
>    2. we also support define a rowtime(watermark) on changelog table, but
>    the rowtime will not be used to track the history of changelog stream.
>
>
>
> WDYT? please correct me if I am wrong.
>
>
> Best,
>
> Leonard
>
>
>
>
> 在 2020年6月24日，11:31，Leonard Xu <[hidden email]> 写道：
>
> Hi, everyone
>
> Thanks Fabian,Kurt for making the multiple version(event time) clear, I
> also like the 'PERIOD FOR SYSTEM' syntax which supported in SQL standard. I
> think we can add some explanation of the multiple version support in the
> future section of FLIP.
>
> For the PRIMARY KEY semantic, I agree with Jark's point that the semantic
> should unify both on changelog source and insert-only source.
>
> Currently, Flink supports PRIMARY KEY after FLIP-87, Flink uses PRIMARY
> KEY NOT ENFORCED because Flink does not own the data like other DBMS therefore
> Flink won't validate/enforce the key integrity and only trusts the external
> systems. It is expected user and external system/application should make
> sure no deduplicated records happened when using NOT ENFORCED.
>
> (a) For PRIMARY KEY NOT ENFORCED semantic on changelog source:
> It means the materialized changelogs (INSERT/UPDATE/DELETE) should be
> unique on the primary key constraints.Flink assumes messages are in order
> on the primary key. Flink will use the PRIMARY KEY for some optimization,
> e.g. use the PRIMARY KEY to update the materialized state by key in
> temporal join operator.
>
>
> (b) For PRIMARY KEY NOT ENFORCED semantic on insert-only source:
> It means records should be unique on the primary key constraints. If there
> are INSERT records with duplicate primary key columns, the result of SQL
> query might be nondeterministic because it broken the PRIMARY KEY
> constraints.
>
> Cheers,
> Leonard
>
>
> 在 2020年6月23日，23:35，Fabian Hueske <[hidden email]> 写道：
>
> Thanks Kurt,
>
> Yes, you are right.
> The `PERIOD FOR SYSTEM_TIME` that you linked before corresponds to the
> VERSION clause that I used and would explicitly define the versioning of a
> table.
> I didn't know that the `PERIOD FOR SYSTEM_TIME` cause is already defined by
> the SQL standard.
> I think we would need a slightly different syntax though because (so far)
> the validity of a row is determined by its own timestamp and the timestamp
> of the next row.
>
> Adding a clause later solves the ambiguity issue for tables with multiple
> event-time attributes.
> However, I'd feel more comfortable having such a cause and an explicit
> definition of the temporal property from the beginning.
> I guess this is a matter of personal preference so I'll go with the
> majority if we decide that every table that has a primary key and an
> event-time attribute should be usable in an event-time temporal table join.
>
> Thanks, Fabian
>
>
> Am Di., 23. Juni 2020 um 16:58 Uhr schrieb Kurt Young <[hidden email]>:
>
> Hi Fabian,
>
> I agree with you that implicitly letting event time to be the version of
> the table will
> work in most cases, but not for all. That's the reason I mentioned `PERIOD
> FOR` [1]
> syntax in my first email, which is already in sql standard to represent the
> validity of
> each row in the table.
>
> If the event time can't be used, or multiple event time are defined, we
> could still add
> this syntax in the future.
>
> What do you think?
>
> [1]
>
>
> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
> Best,
> Kurt
>
>
> On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]> wrote:
>
> Hi everyone,
>
> Every table with a primary key and an event-time attribute provides what
>
> is
>
> needed for an event-time temporal table join.
> I agree that, from a technical point of view, the TEMPORAL keyword is not
> required.
>
> I'm more sceptical about implicitly deriving the versioning information
>
> of
>
> a (temporal) table as the table's only event-time attribute.
> In the query
>
> SELECT *
> FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
> WHERE o.currency = r.currency
>
> the syntax of the temporal table join does not explicitly reference the
> version of the temporal rates table.
> Hence, the system needs a way to derive the version of temporal table.
>
> Implicitly using the (only) event-time attribute of a temporal table
>
> (rates
>
> in the example above) to identify the right version works in most cases,
> but probably not in all.
> * What if a table has more than one event-time attribute? (TableSchema is
> designed to support multiple watermarks; queries with interval joins
> produce tables with multiple event-time attributes, ...)
> * What if the table does not have an event-time attribute in its schema
>
> but
>
> the version should only be provided as meta data?
>
> We could add a clause to define the version of a table, such as:
>
> CREATE TABLE rates (
>   currency CHAR(3) NOT NULL PRIMARY KEY,
>   rate DOUBLE,
>   rowtime TIMESTAMP,
>   WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
> VERSION (rowtime)
> WITH (...);
>
> The presence of a the VERSION clause (or whatever syntax) would
>
> explicitly
>
> define the version of a (temporal) table.
> It would also render the need for the TEMPORAL keyword superfluous
>
> because
>
> there would be another indicator that a table can be used in a temporal
> table join.
>
> I'm OK with not adding the TEMPORAL keyword, but I recommend that we
>
> think
>
> again about the proposed implicit definition of a table's version and how
> it might limit use in the future.
>
> Cheers,
> Fabian
>
> Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:
>
> I'm also +1 for not adding the TEMPORAL keyword.
>
> +1 to make the PRIMARY KEY semantic clear for sources.
> From my point of view:
>
> 1) PRIMARY KEY on changelog souruce:
> It means that when the changelogs (INSERT/UPDATE/DELETE) are
>
> materialized,
>
> the materialized table should be unique on the primary key columns.
> Flink assumes messages are in order on the primary key. Flink doesn't
> validate/enforces the key integrity, but simply trust it (thus NOT
> ENFORCED).
> Flink will use the PRIMARY KEY for some optimization, e.g. use the
>
> PRIMARY
>
> KEY to update the materilized state by key in temporal join operator.
>
> 2) PRIMARY KEY on insert-only source:
> I prefer to have the same semantic to the batch source and changelog
> source, that it implies that records are not duplicate on the primary
>
> key.
>
> Flink just simply trust the primary key constraint, and doesn't valid
>
> it.
>
> If there is duplicate primary keys with INSERT changeflag, then result
>
> of
>
> Flink query might be wrong.
>
> If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
>
> duplicate
>
> primary keys with INSERT changeflag, when we migrate this case to
>
> temporal
>
> table DDL,
> I think this source should emit INSERT/UPDATE (UPSERT) messages instead
>
> of
>
> INSERT-only messages,  e.g. a Kafka compacted topic source?
>
> Best,
> Jark
>
>
> On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
>
> wrote:
>
>
> Hi everyone,
>
> I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
>
> Best,
>
> Konstantin
>
> On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]>
>
> wrote:
>
>
> I agree with Timo, semantic about primary key needs more thought
>
> and
>
> discussion, especially after FLIP-95 and FLIP-105.
>
> Best,
> Kurt
>
>
> On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
>
> wrote:
>
>
> Hi Leonard,
>
> thanks for the summary.
>
> After reading all of the previous arguments and working on
>
> FLIP-95. I
>
> would also lean towards the conclusion of not adding the TEMPORAL
>
> keyword.
>
>
> After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can
>
> be
>
> represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
>
> FOR
>
> SYSTEM_TIME AS OF t would trigger the internal materialization
>
> and
>
> "temporal" logic.
>
> However, we should discuss the meaning of PRIMARY KEY again in
>
> this
>
> case. In a TEMPORAL TABLE scenario, the source would emit
>
> duplicate
>
> primary keys with INSERT changeflag but at different point in
>
> time.
>
> Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> changelog semantics of FLIP-95 and FLIP-105 don't work well with
>
> a
>
> primary key declaration.
>
> Regards,
> Timo
>
>
> On 20.06.20 17:08, Leonard Xu wrote:
>
> Hi everyone,
>
> Thanks for the nice discussion. I’d like to move forward the
>
> work,
>
> please let me simply summarize the main opinion and current
>
> divergences.
>
>
> 1. The agreements have been achieved:
>
> 1.1 The motivation we're discussing temporal table DDL is just
>
> for
>
> creating temporal table in pure SQL to replace pre-process
>
> temporal
>
> table
>
> in YAML/Table API for usability.
>
> 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD
>
> FOR
>
> SYSTEM_TIME” is to make user understand easily.
>
> 1.3 For append-only table, it can convert to changelog table
>
> which
>
> has
>
> been discussed in FLIP-105, we assume the following temporal
>
> table
>
> is
>
> comes
>
> from changelog (Jark, fabian, Timo).
>
> 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
>
> instead
>
> of
>
> the current `LATERAL TABLE(rates(x))`  has come to an
>
> agreement(Fabian,
>
> Timo, Seth, Konstantin, Kurt).
>
>
> 2. The small divergence :
>
> About the definition syntax of the temporal table,
>
> CREATE [TEMPORAL] TABLE rates (
>    currency CHAR(3) NOT NULL PRIMARY KEY,
>    rate DOUBLE,
>    rowtime TIMESTAMP,
>    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> WITH (...);
>
> there is small divergence whether add "TEMPORAL" keyword or
>
> not.
>
>
> 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo,
>
> Fabian,
>
> Seth),
>
> the main advantages are:
>
> （1）"TEMPORAL" keyword is intuitive to indicate the history
>
> tracking
>
> semantics.
>
> （2）"TEMPORAL" keyword illustrates that queries can visit the
>
> previous
>
> versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
>
> keyword.
>
>
> 2.2 the other is using "CREATE TABLE"(Kurt), the main
>
> advantages
>
> are:
>
> （1）Just primary key and time attribute can track previous
>
> versions
>
> of a
>
> table well.
>
> （2）The temporal behavior is triggered by temporal join syntax
>
> rather
>
> than in DDL, all Flink DDL table are dynamic table logically
>
> including
>
> temporal table. If we decide to use "TEMPORAL" keyword and treats
>
> changelog
>
> as temporal table, other tables backed queue like Kafka should
>
> also
>
> use
>
> "TEMPORAL" keyword.
>
>
>
> IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
>
> with
>
> 2.1
>
> may confuse users much. If we take a second to think about, for
>
> source/sink
>
> table which may backed queue (like kafka) or DB (like MySQL), we
>
> did
>
> not
>
> add any keyword in DDL to specify they are source or sinks, it
>
> works
>
> well.
>
> I think temporal table is the third one,  kafka data source and
>
> DB
>
> data
>
> source can play as a source/sink/temporal table depends on the
> position/syntax that user put them in the query. The above rates
>
> table
>
>     - can be a source table if user put it at `SELECT * FROM
>
> rates;`
>
>     - can be a temporal table if user put it at `SELECT * FROM
>
> orders
>
> JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
>
>              ON orders.currency = rates.currency;`
>     - can be sink table if user put is at `INSERT INTO rates
>
> SELECT
>
> *
>
> FROM …; `
>
> From these cases, we found all tables defined in Flink should
>
> be
>
> dynamic table logically, the source/sink/temporal role depends on
>
> the
>
> position/syntax in user’s query.
>
>       In fact we have used similar syntax for current lookup
>
> table,
>
> we
>
> didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
>
> trigger
>
> the
>
> temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
>
> in
>
> query.
>
>
> So, I prefer to resolve the small divergence with “CREATE
>
> TABLE”
>
> which
>
> (1) is more unified with our source/sink/temporal dynamic table
>
> conceptually,
>
> (2) is aligned with current lookup table,
> (3) also make users learn less keyword.
>
> WDYT？
>
> Best,
> Leonard Xu
>
>
>
>
>
>
>
> --
>
> Konstantin Knauf
>
> https://twitter.com/snntrable
>
> https://github.com/knaufk
>
>
>
>
>
>
>

[Jingsong Li]

Thanks for your discussion.

Looks like the problem is supporting the versioned temporal table for the
changelog source.

I want to share more of my thoughts:

When I think about changelog sources, I treat it as a view like: "CREATE
VIEW changelog_table AS SELECT ... FROM origin_table GROUP BY ..." (Some
queries produce changelog records).

Does this view support window aggregation? No...
Does this view support versioned temporal tables? No...

Because both window aggregation and versioned temporal tables require a
time attribute.

So can we give this view a new time attribute?
1. No, keep it not supported.
2. Using processing time.
3. there is an operation time in this view, something like processing time
when modifying the origin table. Treat this operation time as rowtime.
4. Introduce a new time attribute concept: operation time. Assuming it
monotonically increases, no watermark.

NOTE: For the versioned temporal table, there is a time-relation between
these two tables. This time attribute must be something user perceived.

I am slightly +1 for #1 and #2.
For #1: If users really want to support the versioned temporal table for
the changelog source. They can change the definition. And make the
changelog source as a regular table, then they have an operation time field
in the table schema, they can use this field as a rowtime field.
For #2: This versioned temporal table is joined using the processing-time
way, it means we assume records come in a monotonically way, I think it is
good to match changelog concept.

-1 for #3 and #4.
It can work, but l think it is hard to understand what is the rowtime
attribute after "changing" the table.
And I don't think it is not worth creating another concept for users.

Best,
Jingsong Lee

On Thu, Jun 25, 2020 at 10:30 PM Jark Wu <[hidden email]> wrote:

> Hi all,
>
> Thanks Leonard for summarizing our discussion. I want to share more of my
> thoughts:
>
> * rowtime is a column in the its schema, so the rowtime of DELETE event is
> the value of the previous image.
> * operation time is the time when the DML statements happen in databases,
> so the operation time of DELETE events is the time when it happens.
> * rowtime can't be used as operation time for history tracking
> * operation time can't be used as rowtime (can't apply window on the
> operation time)
> * rowtime and operation time are orthogonal concepts and used in different
> scenarios.
> * operation time implicitly means it is monotonically increasing, we don't
> need watermark syntax to specify the out of boundness for it.
>
> ======================================================================
> So, conclusion from my side so far:
>
> * watermark/rowtime + primary key + changelog source != versioned temporal
> table
> * operation time + primary key + changelog source == versioned temporal
> table
> * We may need something like 'PERIOD FOR SYSTEM_TIME(op_ts)' to define the
> operation time
>
> ======================================================================
> However, there is still a pending question I don't have answer:
>
> Assuming you are doing a MIN aggregate on the operation time, that doesn't
> work because the DELETE/UPDATE_BEFORE doesn't hold
> the previous value of operation time and thus can't retract.
>
> The operation time in fact should be metadata information (just like
> RowKind) which shouldn't be in the shema, and can't be accessed in queries.
> But the PERIOD FOR SYSTEM_TIME syntax is in the schema part and should
> refer to a field in the schema...
>
> ======================================================================
>
> Anyway, let's focus on the operation_time vs rowtime problem first. Let me
> know what's your thought!
>
> Best,
> Jark
>
> On Wed, 24 Jun 2020 at 23:49, Leonard Xu <[hidden email]> wrote:
>
> > Hi, kurt, Fabian
> >
> > After an offline discussion with Jark, We think that the  'PERIOD FOR
> > SYSTEM_TIME(operation_time)' statement might be needed now. Changelog
> table
> > is superset of insert-only table, use PRIMARY KEY and rowtime may work
> well
> > in insert-only or upsert source but has some problem in changelog table.
> >
> > 'PERIOD FOR SYSTEM_TIME(operation_time)' in a temporal table
> > defines/maintains  the valid time of each row, the rowtime can not play
> the
> > history tracking function well.
> >
> > *# 1.*operation time (version time) *vs* rowtime (watermark)
> >
> > I will take an example to explain. The following changelog records came
> > from database table using debezium tool:
> > { "before":  null
> >   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> > "12:00:01"},
> >   "op":       "c",  //INSERT
> >   "ts_ms": 1592971201000 // 2020-06-24 12:00:02
> > }
> > { "before": {"currency": "Euro", "rate": 114, "gmt_modified":
> "12:00:05"},
> >   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> > "12:00:05"},
> >   "op":       "u",  //UPDATE
> >   "ts_ms": 1592971206000 // 2020-06-24 12:00:06
> > }
> >
> > { "before": {"currency": "Euro", "rate": 118, "gmt_modified":
> "12:00:05"},
> >   "after":     null,
> >   "op":        "d",  //DELETE
> >   "ts_ms":  1593000011000  // 2020-06-24 20:00:11
> > }
> >
> > The rowtime should be the "gmt_modified" field that belongs to the
> > original record，the "ts_ms" is the the operation time when the DML
> > statement happen in the DB. For DELETE changelog record, its
> "gmt_modified"
> > field (12:00:05) can not reflect the real operation time (20:00:11).
> >
> > In temporal join case, we should maintain the valid time of each row. For
> > a DELETE event, we should use the operation time of DELETE as the “end
> > time” of the row. That says, the record {"currency": "Euro", "rate": 118}
> > is not exist anymore after “20:00:11”, not “12:00:05”.
> >
> > we would not access the record {"currency": "Euro", "rate": 118,
> > "gmt_modified": "12:00:05"} when rowtime is bigger than (12:00:05) if we
> > use rowtime to track the history version, because the DELETE changelog
> > record also has rowtime (12:00:05) and will clear the record in state. In
> > fact, the expected result is that the record expires until (20:00:11)
> when
> > the record is deleted rather than the last update time(20:00:11) in
> > materialized state.
> >
> > From this case, I found rowtime and operation time should be orthogonal
> in
> > temporal table scenario. The operation time should be strictly
> > monotonically increasing  (no out of order) and only be used for
> tracking a
> > history version of a changelog table, every history version of changelog
> > table equals a database table snapshot due to the stream-table duality.
> >
> > *# 2.*The semantic of rowtime and watermark on changelog
> >
> > The rowtime and watermark can also be defined on a changelog table just
> > like other source backed queue, Flink supports cascaded window
> aggregation
> > (with ) in SQL like:
> > SELECT
> >      TUMBLE_ROWTIME(rowtime, INTERVAL '60' SECOND),
> >      MAX(rate) AS rate
> > FROM (
> >        SELECT
> >           MAX(rate) AS rate,
> >           TUMBLE_ROWTIME(rowtime, INTERVAL '5' SECOND) AS `rowtime`
> >        FROM currency
> >             GROUP BY TUMBLE(rowtime, INTERVAL '5' SECOND)
> >     )
> >     GROUP BY TUMBLE(rowtime, INTERVAL '60' SECOND
> >
> > We can think of the output of the first window aggregation as a changelog
> > source of the second window aggregation. There are INSERT/UPDATE/DELETE
> > messages and also watermarks in the changelog stream. And the rowtime in
> > the changelog stream is the `TUMBLE_ROWTIME` value (just like the
> > `gmt_modified` column in DB).
> >
> > *#  summary*
> >
> >    1. we should use ‘PERIOD FOR SYSTEM_TIME(operation_time) syntax to
> >    track history version by operation time rather than rowtime in
> temporal
> >    table scenario.
> >    2. we also support define a rowtime(watermark) on changelog table, but
> >    the rowtime will not be used to track the history of changelog stream.
> >
> >
> >
> > WDYT? please correct me if I am wrong.
> >
> >
> > Best,
> >
> > Leonard
> >
> >
> >
> >
> > 在 2020年6月24日，11:31，Leonard Xu <[hidden email]> 写道：
> >
> > Hi, everyone
> >
> > Thanks Fabian,Kurt for making the multiple version(event time) clear, I
> > also like the 'PERIOD FOR SYSTEM' syntax which supported in SQL
> standard. I
> > think we can add some explanation of the multiple version support in the
> > future section of FLIP.
> >
> > For the PRIMARY KEY semantic, I agree with Jark's point that the semantic
> > should unify both on changelog source and insert-only source.
> >
> > Currently, Flink supports PRIMARY KEY after FLIP-87, Flink uses PRIMARY
> > KEY NOT ENFORCED because Flink does not own the data like other DBMS
> therefore
> > Flink won't validate/enforce the key integrity and only trusts the
> external
> > systems. It is expected user and external system/application should make
> > sure no deduplicated records happened when using NOT ENFORCED.
> >
> > (a) For PRIMARY KEY NOT ENFORCED semantic on changelog source:
> > It means the materialized changelogs (INSERT/UPDATE/DELETE) should be
> > unique on the primary key constraints.Flink assumes messages are in order
> > on the primary key. Flink will use the PRIMARY KEY for some optimization,
> > e.g. use the PRIMARY KEY to update the materialized state by key in
> > temporal join operator.
> >
> >
> > (b) For PRIMARY KEY NOT ENFORCED semantic on insert-only source:
> > It means records should be unique on the primary key constraints. If
> there
> > are INSERT records with duplicate primary key columns, the result of SQL
> > query might be nondeterministic because it broken the PRIMARY KEY
> > constraints.
> >
> > Cheers,
> > Leonard
> >
> >
> > 在 2020年6月23日，23:35，Fabian Hueske <[hidden email]> 写道：
> >
> > Thanks Kurt,
> >
> > Yes, you are right.
> > The `PERIOD FOR SYSTEM_TIME` that you linked before corresponds to the
> > VERSION clause that I used and would explicitly define the versioning of
> a
> > table.
> > I didn't know that the `PERIOD FOR SYSTEM_TIME` cause is already defined
> by
> > the SQL standard.
> > I think we would need a slightly different syntax though because (so far)
> > the validity of a row is determined by its own timestamp and the
> timestamp
> > of the next row.
> >
> > Adding a clause later solves the ambiguity issue for tables with multiple
> > event-time attributes.
> > However, I'd feel more comfortable having such a cause and an explicit
> > definition of the temporal property from the beginning.
> > I guess this is a matter of personal preference so I'll go with the
> > majority if we decide that every table that has a primary key and an
> > event-time attribute should be usable in an event-time temporal table
> join.
> >
> > Thanks, Fabian
> >
> >
> > Am Di., 23. Juni 2020 um 16:58 Uhr schrieb Kurt Young <[hidden email]
> >:
> >
> > Hi Fabian,
> >
> > I agree with you that implicitly letting event time to be the version of
> > the table will
> > work in most cases, but not for all. That's the reason I mentioned
> `PERIOD
> > FOR` [1]
> > syntax in my first email, which is already in sql standard to represent
> the
> > validity of
> > each row in the table.
> >
> > If the event time can't be used, or multiple event time are defined, we
> > could still add
> > this syntax in the future.
> >
> > What do you think?
> >
> > [1]
> >
> >
> >
> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
> > Best,
> > Kurt
> >
> >
> > On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]> wrote:
> >
> > Hi everyone,
> >
> > Every table with a primary key and an event-time attribute provides what
> >
> > is
> >
> > needed for an event-time temporal table join.
> > I agree that, from a technical point of view, the TEMPORAL keyword is not
> > required.
> >
> > I'm more sceptical about implicitly deriving the versioning information
> >
> > of
> >
> > a (temporal) table as the table's only event-time attribute.
> > In the query
> >
> > SELECT *
> > FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
> > WHERE o.currency = r.currency
> >
> > the syntax of the temporal table join does not explicitly reference the
> > version of the temporal rates table.
> > Hence, the system needs a way to derive the version of temporal table.
> >
> > Implicitly using the (only) event-time attribute of a temporal table
> >
> > (rates
> >
> > in the example above) to identify the right version works in most cases,
> > but probably not in all.
> > * What if a table has more than one event-time attribute? (TableSchema is
> > designed to support multiple watermarks; queries with interval joins
> > produce tables with multiple event-time attributes, ...)
> > * What if the table does not have an event-time attribute in its schema
> >
> > but
> >
> > the version should only be provided as meta data?
> >
> > We could add a clause to define the version of a table, such as:
> >
> > CREATE TABLE rates (
> >   currency CHAR(3) NOT NULL PRIMARY KEY,
> >   rate DOUBLE,
> >   rowtime TIMESTAMP,
> >   WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
> > VERSION (rowtime)
> > WITH (...);
> >
> > The presence of a the VERSION clause (or whatever syntax) would
> >
> > explicitly
> >
> > define the version of a (temporal) table.
> > It would also render the need for the TEMPORAL keyword superfluous
> >
> > because
> >
> > there would be another indicator that a table can be used in a temporal
> > table join.
> >
> > I'm OK with not adding the TEMPORAL keyword, but I recommend that we
> >
> > think
> >
> > again about the proposed implicit definition of a table's version and how
> > it might limit use in the future.
> >
> > Cheers,
> > Fabian
> >
> > Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:
> >
> > I'm also +1 for not adding the TEMPORAL keyword.
> >
> > +1 to make the PRIMARY KEY semantic clear for sources.
> > From my point of view:
> >
> > 1) PRIMARY KEY on changelog souruce:
> > It means that when the changelogs (INSERT/UPDATE/DELETE) are
> >
> > materialized,
> >
> > the materialized table should be unique on the primary key columns.
> > Flink assumes messages are in order on the primary key. Flink doesn't
> > validate/enforces the key integrity, but simply trust it (thus NOT
> > ENFORCED).
> > Flink will use the PRIMARY KEY for some optimization, e.g. use the
> >
> > PRIMARY
> >
> > KEY to update the materilized state by key in temporal join operator.
> >
> > 2) PRIMARY KEY on insert-only source:
> > I prefer to have the same semantic to the batch source and changelog
> > source, that it implies that records are not duplicate on the primary
> >
> > key.
> >
> > Flink just simply trust the primary key constraint, and doesn't valid
> >
> > it.
> >
> > If there is duplicate primary keys with INSERT changeflag, then result
> >
> > of
> >
> > Flink query might be wrong.
> >
> > If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
> >
> > duplicate
> >
> > primary keys with INSERT changeflag, when we migrate this case to
> >
> > temporal
> >
> > table DDL,
> > I think this source should emit INSERT/UPDATE (UPSERT) messages instead
> >
> > of
> >
> > INSERT-only messages,  e.g. a Kafka compacted topic source?
> >
> > Best,
> > Jark
> >
> >
> > On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
> >
> > wrote:
> >
> >
> > Hi everyone,
> >
> > I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
> >
> > Best,
> >
> > Konstantin
> >
> > On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]>
> >
> > wrote:
> >
> >
> > I agree with Timo, semantic about primary key needs more thought
> >
> > and
> >
> > discussion, especially after FLIP-95 and FLIP-105.
> >
> > Best,
> > Kurt
> >
> >
> > On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
> >
> > wrote:
> >
> >
> > Hi Leonard,
> >
> > thanks for the summary.
> >
> > After reading all of the previous arguments and working on
> >
> > FLIP-95. I
> >
> > would also lean towards the conclusion of not adding the TEMPORAL
> >
> > keyword.
> >
> >
> > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can
> >
> > be
> >
> > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
> >
> > FOR
> >
> > SYSTEM_TIME AS OF t would trigger the internal materialization
> >
> > and
> >
> > "temporal" logic.
> >
> > However, we should discuss the meaning of PRIMARY KEY again in
> >
> > this
> >
> > case. In a TEMPORAL TABLE scenario, the source would emit
> >
> > duplicate
> >
> > primary keys with INSERT changeflag but at different point in
> >
> > time.
> >
> > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > changelog semantics of FLIP-95 and FLIP-105 don't work well with
> >
> > a
> >
> > primary key declaration.
> >
> > Regards,
> > Timo
> >
> >
> > On 20.06.20 17:08, Leonard Xu wrote:
> >
> > Hi everyone,
> >
> > Thanks for the nice discussion. I’d like to move forward the
> >
> > work,
> >
> > please let me simply summarize the main opinion and current
> >
> > divergences.
> >
> >
> > 1. The agreements have been achieved:
> >
> > 1.1 The motivation we're discussing temporal table DDL is just
> >
> > for
> >
> > creating temporal table in pure SQL to replace pre-process
> >
> > temporal
> >
> > table
> >
> > in YAML/Table API for usability.
> >
> > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD
> >
> > FOR
> >
> > SYSTEM_TIME” is to make user understand easily.
> >
> > 1.3 For append-only table, it can convert to changelog table
> >
> > which
> >
> > has
> >
> > been discussed in FLIP-105, we assume the following temporal
> >
> > table
> >
> > is
> >
> > comes
> >
> > from changelog (Jark, fabian, Timo).
> >
> > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
> >
> > instead
> >
> > of
> >
> > the current `LATERAL TABLE(rates(x))`  has come to an
> >
> > agreement(Fabian,
> >
> > Timo, Seth, Konstantin, Kurt).
> >
> >
> > 2. The small divergence :
> >
> > About the definition syntax of the temporal table,
> >
> > CREATE [TEMPORAL] TABLE rates (
> >    currency CHAR(3) NOT NULL PRIMARY KEY,
> >    rate DOUBLE,
> >    rowtime TIMESTAMP,
> >    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > WITH (...);
> >
> > there is small divergence whether add "TEMPORAL" keyword or
> >
> > not.
> >
> >
> > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo,
> >
> > Fabian,
> >
> > Seth),
> >
> > the main advantages are:
> >
> > （1）"TEMPORAL" keyword is intuitive to indicate the history
> >
> > tracking
> >
> > semantics.
> >
> > （2）"TEMPORAL" keyword illustrates that queries can visit the
> >
> > previous
> >
> > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
> >
> > keyword.
> >
> >
> > 2.2 the other is using "CREATE TABLE"(Kurt), the main
> >
> > advantages
> >
> > are:
> >
> > （1）Just primary key and time attribute can track previous
> >
> > versions
> >
> > of a
> >
> > table well.
> >
> > （2）The temporal behavior is triggered by temporal join syntax
> >
> > rather
> >
> > than in DDL, all Flink DDL table are dynamic table logically
> >
> > including
> >
> > temporal table. If we decide to use "TEMPORAL" keyword and treats
> >
> > changelog
> >
> > as temporal table, other tables backed queue like Kafka should
> >
> > also
> >
> > use
> >
> > "TEMPORAL" keyword.
> >
> >
> >
> > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
> >
> > with
> >
> > 2.1
> >
> > may confuse users much. If we take a second to think about, for
> >
> > source/sink
> >
> > table which may backed queue (like kafka) or DB (like MySQL), we
> >
> > did
> >
> > not
> >
> > add any keyword in DDL to specify they are source or sinks, it
> >
> > works
> >
> > well.
> >
> > I think temporal table is the third one,  kafka data source and
> >
> > DB
> >
> > data
> >
> > source can play as a source/sink/temporal table depends on the
> > position/syntax that user put them in the query. The above rates
> >
> > table
> >
> >     - can be a source table if user put it at `SELECT * FROM
> >
> > rates;`
> >
> >     - can be a temporal table if user put it at `SELECT * FROM
> >
> > orders
> >
> > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> >
> >              ON orders.currency = rates.currency;`
> >     - can be sink table if user put is at `INSERT INTO rates
> >
> > SELECT
> >
> > *
> >
> > FROM …; `
> >
> > From these cases, we found all tables defined in Flink should
> >
> > be
> >
> > dynamic table logically, the source/sink/temporal role depends on
> >
> > the
> >
> > position/syntax in user’s query.
> >
> >       In fact we have used similar syntax for current lookup
> >
> > table,
> >
> > we
> >
> > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
> >
> > trigger
> >
> > the
> >
> > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
> >
> > in
> >
> > query.
> >
> >
> > So, I prefer to resolve the small divergence with “CREATE
> >
> > TABLE”
> >
> > which
> >
> > (1) is more unified with our source/sink/temporal dynamic table
> >
> > conceptually,
> >
> > (2) is aligned with current lookup table,
> > (3) also make users learn less keyword.
> >
> > WDYT？
> >
> > Best,
> > Leonard Xu
> >
> >
> >
> >
> >
> >
> >
> > --
> >
> > Konstantin Knauf
> >
> > https://twitter.com/snntrable
> >
> > https://github.com/knaufk
> >
> >
> >
> >
> >
> >
> >
>

--
Best, Jingsong Lee

[Konstantin Knauf-4]

Hi everyone,

well, this got complicated :) Let me add my thoughts:

* Temporal Table Joins are already quite hard to understand for many users.
If need be, we should trade off for simplicity.

* The important case is the *event time *temporal join. In my understanding
processing time temporal joins are comparably easy, no history tracking is
needed, etc.

* It seems that for regular upsert streams with an event time attribute,
everyone agrees that it works. There are olny questions about multiple
event time attributes, which we could in my opinion postpone for future
work.

* For changelog streams, which specify an event time column explicitly, it
should be possible to use it for event time temporal tables. I understand
that deletion can not be handled properly, but we could - for example -
handle this exactly like an upsert stream, i.e. ignore deletions. This is a
limitation, but it is at least easy to understand and acceptable for many
use cases, I believe. Alternatively, one could also use the "ts_ms" for
deletion, which would always be larger than the event time.

CREATE TABLE currency_rates (
  id BIGINT,
  name STRING,
  rate DECIMAL(10, 5),  time TIMESTAMP(3),   WATERMARK FOR time AS ...) WITH (
 'connector' = 'kafka',
 ...
 'format' = 'debezium-json')


* For changelog streams without an event time attribute (the more common
case?), it would be great if we can support temporal table joins based on
"ts_ms" (in the debezium case). One option could be to "simply" extract
"ts_ms" and make it possible to use it as an event time column. Then we
would again be in the above case. Thinking about it, this could even be
addressed in [1], which is also planned for Flink 1.12 as far as I know. *
This could look something like:

CREATE TABLE topic_products (
  id BIGINT,
  name STRING,
  description STRING,
  weight DECIMAL(10, 2),   time TIMESTAMP(3)) WITH (
 'connector' = 'kafka',
 ...
 'format' = 'debezium-json'
 'timestamp' = 'time' )


I hope I roughly understood your concerns and made sense in my comments.
Looking forward to what you think.

Cheers,

Konstantin


[1]
https://cwiki.apache.org/confluence/display/FLINK/FLIP-107%3A+Reading+table+columns+from+different+parts+of+source+records
*  In this case, the DELETE statement could theoretically actually be
handled properly, because the "ts_ms" is used throughout.

On Sun, Jun 28, 2020 at 8:05 AM Jingsong Li <[hidden email]> wrote:

> Thanks for your discussion.
>
> Looks like the problem is supporting the versioned temporal table for the
> changelog source.
>
> I want to share more of my thoughts:
>
> When I think about changelog sources, I treat it as a view like: "CREATE
> VIEW changelog_table AS SELECT ... FROM origin_table GROUP BY ..." (Some
> queries produce changelog records).
>
> Does this view support window aggregation? No...
> Does this view support versioned temporal tables? No...
>
> Because both window aggregation and versioned temporal tables require a
> time attribute.
>
> So can we give this view a new time attribute?
> 1. No, keep it not supported.
> 2. Using processing time.
> 3. there is an operation time in this view, something like processing time
> when modifying the origin table. Treat this operation time as rowtime.
> 4. Introduce a new time attribute concept: operation time. Assuming it
> monotonically increases, no watermark.
>
> NOTE: For the versioned temporal table, there is a time-relation between
> these two tables. This time attribute must be something user perceived.
>
> I am slightly +1 for #1 and #2.
> For #1: If users really want to support the versioned temporal table for
> the changelog source. They can change the definition. And make the
> changelog source as a regular table, then they have an operation time field
> in the table schema, they can use this field as a rowtime field.
> For #2: This versioned temporal table is joined using the processing-time
> way, it means we assume records come in a monotonically way, I think it is
> good to match changelog concept.
>
> -1 for #3 and #4.
> It can work, but l think it is hard to understand what is the rowtime
> attribute after "changing" the table.
> And I don't think it is not worth creating another concept for users.
>
> Best,
> Jingsong Lee
>
> On Thu, Jun 25, 2020 at 10:30 PM Jark Wu <[hidden email]> wrote:
>
> > Hi all,
> >
> > Thanks Leonard for summarizing our discussion. I want to share more of my
> > thoughts:
> >
> > * rowtime is a column in the its schema, so the rowtime of DELETE event
> is
> > the value of the previous image.
> > * operation time is the time when the DML statements happen in databases,
> > so the operation time of DELETE events is the time when it happens.
> > * rowtime can't be used as operation time for history tracking
> > * operation time can't be used as rowtime (can't apply window on the
> > operation time)
> > * rowtime and operation time are orthogonal concepts and used in
> different
> > scenarios.
> > * operation time implicitly means it is monotonically increasing, we
> don't
> > need watermark syntax to specify the out of boundness for it.
> >
> > ======================================================================
> > So, conclusion from my side so far:
> >
> > * watermark/rowtime + primary key + changelog source != versioned
> temporal
> > table
> > * operation time + primary key + changelog source == versioned temporal
> > table
> > * We may need something like 'PERIOD FOR SYSTEM_TIME(op_ts)' to define
> the
> > operation time
> >
> > ======================================================================
> > However, there is still a pending question I don't have answer:
> >
> > Assuming you are doing a MIN aggregate on the operation time, that
> doesn't
> > work because the DELETE/UPDATE_BEFORE doesn't hold
> > the previous value of operation time and thus can't retract.
> >
> > The operation time in fact should be metadata information (just like
> > RowKind) which shouldn't be in the shema, and can't be accessed in
> queries.
> > But the PERIOD FOR SYSTEM_TIME syntax is in the schema part and should
> > refer to a field in the schema...
> >
> > ======================================================================
> >
> > Anyway, let's focus on the operation_time vs rowtime problem first. Let
> me
> > know what's your thought!
> >
> > Best,
> > Jark
> >
> > On Wed, 24 Jun 2020 at 23:49, Leonard Xu <[hidden email]> wrote:
> >
> > > Hi, kurt, Fabian
> > >
> > > After an offline discussion with Jark, We think that the  'PERIOD FOR
> > > SYSTEM_TIME(operation_time)' statement might be needed now. Changelog
> > table
> > > is superset of insert-only table, use PRIMARY KEY and rowtime may work
> > well
> > > in insert-only or upsert source but has some problem in changelog
> table.
> > >
> > > 'PERIOD FOR SYSTEM_TIME(operation_time)' in a temporal table
> > > defines/maintains  the valid time of each row, the rowtime can not play
> > the
> > > history tracking function well.
> > >
> > > *# 1.*operation time (version time) *vs* rowtime (watermark)
> > >
> > > I will take an example to explain. The following changelog records came
> > > from database table using debezium tool:
> > > { "before":  null
> > >   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> > > "12:00:01"},
> > >   "op":       "c",  //INSERT
> > >   "ts_ms": 1592971201000 // 2020-06-24 12:00:02
> > > }
> > > { "before": {"currency": "Euro", "rate": 114, "gmt_modified":
> > "12:00:05"},
> > >   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> > > "12:00:05"},
> > >   "op":       "u",  //UPDATE
> > >   "ts_ms": 1592971206000 // 2020-06-24 12:00:06
> > > }
> > >
> > > { "before": {"currency": "Euro", "rate": 118, "gmt_modified":
> > "12:00:05"},
> > >   "after":     null,
> > >   "op":        "d",  //DELETE
> > >   "ts_ms":  1593000011000  // 2020-06-24 20:00:11
> > > }
> > >
> > > The rowtime should be the "gmt_modified" field that belongs to the
> > > original record，the "ts_ms" is the the operation time when the DML
> > > statement happen in the DB. For DELETE changelog record, its
> > "gmt_modified"
> > > field (12:00:05) can not reflect the real operation time (20:00:11).
> > >
> > > In temporal join case, we should maintain the valid time of each row.
> For
> > > a DELETE event, we should use the operation time of DELETE as the “end
> > > time” of the row. That says, the record {"currency": "Euro", "rate":
> 118}
> > > is not exist anymore after “20:00:11”, not “12:00:05”.
> > >
> > > we would not access the record {"currency": "Euro", "rate": 118,
> > > "gmt_modified": "12:00:05"} when rowtime is bigger than (12:00:05) if
> we
> > > use rowtime to track the history version, because the DELETE changelog
> > > record also has rowtime (12:00:05) and will clear the record in state.
> In
> > > fact, the expected result is that the record expires until (20:00:11)
> > when
> > > the record is deleted rather than the last update time(20:00:11) in
> > > materialized state.
> > >
> > > From this case, I found rowtime and operation time should be orthogonal
> > in
> > > temporal table scenario. The operation time should be strictly
> > > monotonically increasing  (no out of order) and only be used for
> > tracking a
> > > history version of a changelog table, every history version of
> changelog
> > > table equals a database table snapshot due to the stream-table duality.
> > >
> > > *# 2.*The semantic of rowtime and watermark on changelog
> > >
> > > The rowtime and watermark can also be defined on a changelog table just
> > > like other source backed queue, Flink supports cascaded window
> > aggregation
> > > (with ) in SQL like:
> > > SELECT
> > >      TUMBLE_ROWTIME(rowtime, INTERVAL '60' SECOND),
> > >      MAX(rate) AS rate
> > > FROM (
> > >        SELECT
> > >           MAX(rate) AS rate,
> > >           TUMBLE_ROWTIME(rowtime, INTERVAL '5' SECOND) AS `rowtime`
> > >        FROM currency
> > >             GROUP BY TUMBLE(rowtime, INTERVAL '5' SECOND)
> > >     )
> > >     GROUP BY TUMBLE(rowtime, INTERVAL '60' SECOND
> > >
> > > We can think of the output of the first window aggregation as a
> changelog
> > > source of the second window aggregation. There are INSERT/UPDATE/DELETE
> > > messages and also watermarks in the changelog stream. And the rowtime
> in
> > > the changelog stream is the `TUMBLE_ROWTIME` value (just like the
> > > `gmt_modified` column in DB).
> > >
> > > *#  summary*
> > >
> > >    1. we should use ‘PERIOD FOR SYSTEM_TIME(operation_time) syntax to
> > >    track history version by operation time rather than rowtime in
> > temporal
> > >    table scenario.
> > >    2. we also support define a rowtime(watermark) on changelog table,
> but
> > >    the rowtime will not be used to track the history of changelog
> stream.
> > >
> > >
> > >
> > > WDYT? please correct me if I am wrong.
> > >
> > >
> > > Best,
> > >
> > > Leonard
> > >
> > >
> > >
> > >
> > > 在 2020年6月24日，11:31，Leonard Xu <[hidden email]> 写道：
> > >
> > > Hi, everyone
> > >
> > > Thanks Fabian,Kurt for making the multiple version(event time) clear, I
> > > also like the 'PERIOD FOR SYSTEM' syntax which supported in SQL
> > standard. I
> > > think we can add some explanation of the multiple version support in
> the
> > > future section of FLIP.
> > >
> > > For the PRIMARY KEY semantic, I agree with Jark's point that the
> semantic
> > > should unify both on changelog source and insert-only source.
> > >
> > > Currently, Flink supports PRIMARY KEY after FLIP-87, Flink uses PRIMARY
> > > KEY NOT ENFORCED because Flink does not own the data like other DBMS
> > therefore
> > > Flink won't validate/enforce the key integrity and only trusts the
> > external
> > > systems. It is expected user and external system/application should
> make
> > > sure no deduplicated records happened when using NOT ENFORCED.
> > >
> > > (a) For PRIMARY KEY NOT ENFORCED semantic on changelog source:
> > > It means the materialized changelogs (INSERT/UPDATE/DELETE) should be
> > > unique on the primary key constraints.Flink assumes messages are in
> order
> > > on the primary key. Flink will use the PRIMARY KEY for some
> optimization,
> > > e.g. use the PRIMARY KEY to update the materialized state by key in
> > > temporal join operator.
> > >
> > >
> > > (b) For PRIMARY KEY NOT ENFORCED semantic on insert-only source:
> > > It means records should be unique on the primary key constraints. If
> > there
> > > are INSERT records with duplicate primary key columns, the result of
> SQL
> > > query might be nondeterministic because it broken the PRIMARY KEY
> > > constraints.
> > >
> > > Cheers,
> > > Leonard
> > >
> > >
> > > 在 2020年6月23日，23:35，Fabian Hueske <[hidden email]> 写道：
> > >
> > > Thanks Kurt,
> > >
> > > Yes, you are right.
> > > The `PERIOD FOR SYSTEM_TIME` that you linked before corresponds to the
> > > VERSION clause that I used and would explicitly define the versioning
> of
> > a
> > > table.
> > > I didn't know that the `PERIOD FOR SYSTEM_TIME` cause is already
> defined
> > by
> > > the SQL standard.
> > > I think we would need a slightly different syntax though because (so
> far)
> > > the validity of a row is determined by its own timestamp and the
> > timestamp
> > > of the next row.
> > >
> > > Adding a clause later solves the ambiguity issue for tables with
> multiple
> > > event-time attributes.
> > > However, I'd feel more comfortable having such a cause and an explicit
> > > definition of the temporal property from the beginning.
> > > I guess this is a matter of personal preference so I'll go with the
> > > majority if we decide that every table that has a primary key and an
> > > event-time attribute should be usable in an event-time temporal table
> > join.
> > >
> > > Thanks, Fabian
> > >
> > >
> > > Am Di., 23. Juni 2020 um 16:58 Uhr schrieb Kurt Young <
> [hidden email]
> > >:
> > >
> > > Hi Fabian,
> > >
> > > I agree with you that implicitly letting event time to be the version
> of
> > > the table will
> > > work in most cases, but not for all. That's the reason I mentioned
> > `PERIOD
> > > FOR` [1]
> > > syntax in my first email, which is already in sql standard to represent
> > the
> > > validity of
> > > each row in the table.
> > >
> > > If the event time can't be used, or multiple event time are defined, we
> > > could still add
> > > this syntax in the future.
> > >
> > > What do you think?
> > >
> > > [1]
> > >
> > >
> > >
> >
> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
> > > Best,
> > > Kurt
> > >
> > >
> > > On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]>
> wrote:
> > >
> > > Hi everyone,
> > >
> > > Every table with a primary key and an event-time attribute provides
> what
> > >
> > > is
> > >
> > > needed for an event-time temporal table join.
> > > I agree that, from a technical point of view, the TEMPORAL keyword is
> not
> > > required.
> > >
> > > I'm more sceptical about implicitly deriving the versioning information
> > >
> > > of
> > >
> > > a (temporal) table as the table's only event-time attribute.
> > > In the query
> > >
> > > SELECT *
> > > FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
> > > WHERE o.currency = r.currency
> > >
> > > the syntax of the temporal table join does not explicitly reference the
> > > version of the temporal rates table.
> > > Hence, the system needs a way to derive the version of temporal table.
> > >
> > > Implicitly using the (only) event-time attribute of a temporal table
> > >
> > > (rates
> > >
> > > in the example above) to identify the right version works in most
> cases,
> > > but probably not in all.
> > > * What if a table has more than one event-time attribute? (TableSchema
> is
> > > designed to support multiple watermarks; queries with interval joins
> > > produce tables with multiple event-time attributes, ...)
> > > * What if the table does not have an event-time attribute in its schema
> > >
> > > but
> > >
> > > the version should only be provided as meta data?
> > >
> > > We could add a clause to define the version of a table, such as:
> > >
> > > CREATE TABLE rates (
> > >   currency CHAR(3) NOT NULL PRIMARY KEY,
> > >   rate DOUBLE,
> > >   rowtime TIMESTAMP,
> > >   WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
> > > VERSION (rowtime)
> > > WITH (...);
> > >
> > > The presence of a the VERSION clause (or whatever syntax) would
> > >
> > > explicitly
> > >
> > > define the version of a (temporal) table.
> > > It would also render the need for the TEMPORAL keyword superfluous
> > >
> > > because
> > >
> > > there would be another indicator that a table can be used in a temporal
> > > table join.
> > >
> > > I'm OK with not adding the TEMPORAL keyword, but I recommend that we
> > >
> > > think
> > >
> > > again about the proposed implicit definition of a table's version and
> how
> > > it might limit use in the future.
> > >
> > > Cheers,
> > > Fabian
> > >
> > > Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]>:
> > >
> > > I'm also +1 for not adding the TEMPORAL keyword.
> > >
> > > +1 to make the PRIMARY KEY semantic clear for sources.
> > > From my point of view:
> > >
> > > 1) PRIMARY KEY on changelog souruce:
> > > It means that when the changelogs (INSERT/UPDATE/DELETE) are
> > >
> > > materialized,
> > >
> > > the materialized table should be unique on the primary key columns.
> > > Flink assumes messages are in order on the primary key. Flink doesn't
> > > validate/enforces the key integrity, but simply trust it (thus NOT
> > > ENFORCED).
> > > Flink will use the PRIMARY KEY for some optimization, e.g. use the
> > >
> > > PRIMARY
> > >
> > > KEY to update the materilized state by key in temporal join operator.
> > >
> > > 2) PRIMARY KEY on insert-only source:
> > > I prefer to have the same semantic to the batch source and changelog
> > > source, that it implies that records are not duplicate on the primary
> > >
> > > key.
> > >
> > > Flink just simply trust the primary key constraint, and doesn't valid
> > >
> > > it.
> > >
> > > If there is duplicate primary keys with INSERT changeflag, then result
> > >
> > > of
> > >
> > > Flink query might be wrong.
> > >
> > > If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
> > >
> > > duplicate
> > >
> > > primary keys with INSERT changeflag, when we migrate this case to
> > >
> > > temporal
> > >
> > > table DDL,
> > > I think this source should emit INSERT/UPDATE (UPSERT) messages instead
> > >
> > > of
> > >
> > > INSERT-only messages,  e.g. a Kafka compacted topic source?
> > >
> > > Best,
> > > Jark
> > >
> > >
> > > On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
> > >
> > > wrote:
> > >
> > >
> > > Hi everyone,
> > >
> > > I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
> > >
> > > Best,
> > >
> > > Konstantin
> > >
> > > On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]>
> > >
> > > wrote:
> > >
> > >
> > > I agree with Timo, semantic about primary key needs more thought
> > >
> > > and
> > >
> > > discussion, especially after FLIP-95 and FLIP-105.
> > >
> > > Best,
> > > Kurt
> > >
> > >
> > > On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
> > >
> > > wrote:
> > >
> > >
> > > Hi Leonard,
> > >
> > > thanks for the summary.
> > >
> > > After reading all of the previous arguments and working on
> > >
> > > FLIP-95. I
> > >
> > > would also lean towards the conclusion of not adding the TEMPORAL
> > >
> > > keyword.
> > >
> > >
> > > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can
> > >
> > > be
> > >
> > > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
> > >
> > > FOR
> > >
> > > SYSTEM_TIME AS OF t would trigger the internal materialization
> > >
> > > and
> > >
> > > "temporal" logic.
> > >
> > > However, we should discuss the meaning of PRIMARY KEY again in
> > >
> > > this
> > >
> > > case. In a TEMPORAL TABLE scenario, the source would emit
> > >
> > > duplicate
> > >
> > > primary keys with INSERT changeflag but at different point in
> > >
> > > time.
> > >
> > > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > > changelog semantics of FLIP-95 and FLIP-105 don't work well with
> > >
> > > a
> > >
> > > primary key declaration.
> > >
> > > Regards,
> > > Timo
> > >
> > >
> > > On 20.06.20 17:08, Leonard Xu wrote:
> > >
> > > Hi everyone,
> > >
> > > Thanks for the nice discussion. I’d like to move forward the
> > >
> > > work,
> > >
> > > please let me simply summarize the main opinion and current
> > >
> > > divergences.
> > >
> > >
> > > 1. The agreements have been achieved:
> > >
> > > 1.1 The motivation we're discussing temporal table DDL is just
> > >
> > > for
> > >
> > > creating temporal table in pure SQL to replace pre-process
> > >
> > > temporal
> > >
> > > table
> > >
> > > in YAML/Table API for usability.
> > >
> > > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD
> > >
> > > FOR
> > >
> > > SYSTEM_TIME” is to make user understand easily.
> > >
> > > 1.3 For append-only table, it can convert to changelog table
> > >
> > > which
> > >
> > > has
> > >
> > > been discussed in FLIP-105, we assume the following temporal
> > >
> > > table
> > >
> > > is
> > >
> > > comes
> > >
> > > from changelog (Jark, fabian, Timo).
> > >
> > > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
> > >
> > > instead
> > >
> > > of
> > >
> > > the current `LATERAL TABLE(rates(x))`  has come to an
> > >
> > > agreement(Fabian,
> > >
> > > Timo, Seth, Konstantin, Kurt).
> > >
> > >
> > > 2. The small divergence :
> > >
> > > About the definition syntax of the temporal table,
> > >
> > > CREATE [TEMPORAL] TABLE rates (
> > >    currency CHAR(3) NOT NULL PRIMARY KEY,
> > >    rate DOUBLE,
> > >    rowtime TIMESTAMP,
> > >    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > > WITH (...);
> > >
> > > there is small divergence whether add "TEMPORAL" keyword or
> > >
> > > not.
> > >
> > >
> > > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo,
> > >
> > > Fabian,
> > >
> > > Seth),
> > >
> > > the main advantages are:
> > >
> > > （1）"TEMPORAL" keyword is intuitive to indicate the history
> > >
> > > tracking
> > >
> > > semantics.
> > >
> > > （2）"TEMPORAL" keyword illustrates that queries can visit the
> > >
> > > previous
> > >
> > > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
> > >
> > > keyword.
> > >
> > >
> > > 2.2 the other is using "CREATE TABLE"(Kurt), the main
> > >
> > > advantages
> > >
> > > are:
> > >
> > > （1）Just primary key and time attribute can track previous
> > >
> > > versions
> > >
> > > of a
> > >
> > > table well.
> > >
> > > （2）The temporal behavior is triggered by temporal join syntax
> > >
> > > rather
> > >
> > > than in DDL, all Flink DDL table are dynamic table logically
> > >
> > > including
> > >
> > > temporal table. If we decide to use "TEMPORAL" keyword and treats
> > >
> > > changelog
> > >
> > > as temporal table, other tables backed queue like Kafka should
> > >
> > > also
> > >
> > > use
> > >
> > > "TEMPORAL" keyword.
> > >
> > >
> > >
> > > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
> > >
> > > with
> > >
> > > 2.1
> > >
> > > may confuse users much. If we take a second to think about, for
> > >
> > > source/sink
> > >
> > > table which may backed queue (like kafka) or DB (like MySQL), we
> > >
> > > did
> > >
> > > not
> > >
> > > add any keyword in DDL to specify they are source or sinks, it
> > >
> > > works
> > >
> > > well.
> > >
> > > I think temporal table is the third one,  kafka data source and
> > >
> > > DB
> > >
> > > data
> > >
> > > source can play as a source/sink/temporal table depends on the
> > > position/syntax that user put them in the query. The above rates
> > >
> > > table
> > >
> > >     - can be a source table if user put it at `SELECT * FROM
> > >
> > > rates;`
> > >
> > >     - can be a temporal table if user put it at `SELECT * FROM
> > >
> > > orders
> > >
> > > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> > >
> > >              ON orders.currency = rates.currency;`
> > >     - can be sink table if user put is at `INSERT INTO rates
> > >
> > > SELECT
> > >
> > > *
> > >
> > > FROM …; `
> > >
> > > From these cases, we found all tables defined in Flink should
> > >
> > > be
> > >
> > > dynamic table logically, the source/sink/temporal role depends on
> > >
> > > the
> > >
> > > position/syntax in user’s query.
> > >
> > >       In fact we have used similar syntax for current lookup
> > >
> > > table,
> > >
> > > we
> > >
> > > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
> > >
> > > trigger
> > >
> > > the
> > >
> > > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
> > >
> > > in
> > >
> > > query.
> > >
> > >
> > > So, I prefer to resolve the small divergence with “CREATE
> > >
> > > TABLE”
> > >
> > > which
> > >
> > > (1) is more unified with our source/sink/temporal dynamic table
> > >
> > > conceptually,
> > >
> > > (2) is aligned with current lookup table,
> > > (3) also make users learn less keyword.
> > >
> > > WDYT？
> > >
> > > Best,
> > > Leonard Xu
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > --
> > >
> > > Konstantin Knauf
> > >
> > > https://twitter.com/snntrable
> > >
> > > https://github.com/knaufk
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> >
>
>
> --
> Best, Jingsong Lee
>

--

Konstantin Knauf

https://twitter.com/snntrable

https://github.com/knaufk

[Seth Wiesman-4]

It is clear there are a lot of edge cases with temporal tables that need to
be carefully thought out. If we go at this problem from the perspective of
what a majority of users need to accomplish in production, I believe there
is a simpler version of this problem we can solve that can be expanded in
the future.

The most important practical use case is denormalizing star schemas. A user
has the main data stream, their fact table, that needs to be processed.
Before applying specific business logic the dimension stream needs to be
joined with one or more dimension streams. The canonical example of this
being joining transactions with currency rates. I'm not saying all this to
be pedantic but to make the point that if we can solve this practical use
case in a way that may be extended in the future I believe that will
already be immensely useful for most users.

In this case, the data is most likely coming from CDC or something that
approximates it and contains a clearly defined event time column.

For this common use case the syntax would only need to support:

   - Single event time column
   - Joining with externally defined Upsert streams. In the first version,
   a stream could only be used as a temporal table if the join was the first
   operation after reading from an external source. We could disallow using
   streams post flink aggregation as temporal tables in the beginning until
   there is a larger consensus of what timestamp to use.

Seth

On Thu, Jul 2, 2020 at 11:49 AM Konstantin Knauf <[hidden email]> wrote:

> Hi everyone,
>
> well, this got complicated :) Let me add my thoughts:
>
> * Temporal Table Joins are already quite hard to understand for many users.
> If need be, we should trade off for simplicity.
>
> * The important case is the *event time *temporal join. In my understanding
> processing time temporal joins are comparably easy, no history tracking is
> needed, etc.
>
> * It seems that for regular upsert streams with an event time attribute,
> everyone agrees that it works. There are olny questions about multiple
> event time attributes, which we could in my opinion postpone for future
> work.
>
> * For changelog streams, which specify an event time column explicitly, it
> should be possible to use it for event time temporal tables. I understand
> that deletion can not be handled properly, but we could - for example -
> handle this exactly like an upsert stream, i.e. ignore deletions. This is a
> limitation, but it is at least easy to understand and acceptable for many
> use cases, I believe. Alternatively, one could also use the "ts_ms" for
> deletion, which would always be larger than the event time.
>
> CREATE TABLE currency_rates (
>   id BIGINT,
>   name STRING,
>   rate DECIMAL(10, 5),  time TIMESTAMP(3),   WATERMARK FOR time AS ...)
> WITH (
>  'connector' = 'kafka',
>  ...
>  'format' = 'debezium-json')
>
>
> * For changelog streams without an event time attribute (the more common
> case?), it would be great if we can support temporal table joins based on
> "ts_ms" (in the debezium case). One option could be to "simply" extract
> "ts_ms" and make it possible to use it as an event time column. Then we
> would again be in the above case. Thinking about it, this could even be
> addressed in [1], which is also planned for Flink 1.12 as far as I know. *
> This could look something like:
>
> CREATE TABLE topic_products (
>   id BIGINT,
>   name STRING,
>   description STRING,
>   weight DECIMAL(10, 2),   time TIMESTAMP(3)) WITH (
>  'connector' = 'kafka',
>  ...
>  'format' = 'debezium-json'
>  'timestamp' = 'time' )
>
>
> I hope I roughly understood your concerns and made sense in my comments.
> Looking forward to what you think.
>
> Cheers,
>
> Konstantin
>
>
> [1]
>
> https://cwiki.apache.org/confluence/display/FLINK/FLIP-107%3A+Reading+table+columns+from+different+parts+of+source+records
> *  In this case, the DELETE statement could theoretically actually be
> handled properly, because the "ts_ms" is used throughout.
>
> On Sun, Jun 28, 2020 at 8:05 AM Jingsong Li <[hidden email]>
> wrote:
>
> > Thanks for your discussion.
> >
> > Looks like the problem is supporting the versioned temporal table for the
> > changelog source.
> >
> > I want to share more of my thoughts:
> >
> > When I think about changelog sources, I treat it as a view like: "CREATE
> > VIEW changelog_table AS SELECT ... FROM origin_table GROUP BY ..." (Some
> > queries produce changelog records).
> >
> > Does this view support window aggregation? No...
> > Does this view support versioned temporal tables? No...
> >
> > Because both window aggregation and versioned temporal tables require a
> > time attribute.
> >
> > So can we give this view a new time attribute?
> > 1. No, keep it not supported.
> > 2. Using processing time.
> > 3. there is an operation time in this view, something like processing
> time
> > when modifying the origin table. Treat this operation time as rowtime.
> > 4. Introduce a new time attribute concept: operation time. Assuming it
> > monotonically increases, no watermark.
> >
> > NOTE: For the versioned temporal table, there is a time-relation between
> > these two tables. This time attribute must be something user perceived.
> >
> > I am slightly +1 for #1 and #2.
> > For #1: If users really want to support the versioned temporal table for
> > the changelog source. They can change the definition. And make the
> > changelog source as a regular table, then they have an operation time
> field
> > in the table schema, they can use this field as a rowtime field.
> > For #2: This versioned temporal table is joined using the processing-time
> > way, it means we assume records come in a monotonically way, I think it
> is
> > good to match changelog concept.
> >
> > -1 for #3 and #4.
> > It can work, but l think it is hard to understand what is the rowtime
> > attribute after "changing" the table.
> > And I don't think it is not worth creating another concept for users.
> >
> > Best,
> > Jingsong Lee
> >
> > On Thu, Jun 25, 2020 at 10:30 PM Jark Wu <[hidden email]> wrote:
> >
> > > Hi all,
> > >
> > > Thanks Leonard for summarizing our discussion. I want to share more of
> my
> > > thoughts:
> > >
> > > * rowtime is a column in the its schema, so the rowtime of DELETE event
> > is
> > > the value of the previous image.
> > > * operation time is the time when the DML statements happen in
> databases,
> > > so the operation time of DELETE events is the time when it happens.
> > > * rowtime can't be used as operation time for history tracking
> > > * operation time can't be used as rowtime (can't apply window on the
> > > operation time)
> > > * rowtime and operation time are orthogonal concepts and used in
> > different
> > > scenarios.
> > > * operation time implicitly means it is monotonically increasing, we
> > don't
> > > need watermark syntax to specify the out of boundness for it.
> > >
> > > ======================================================================
> > > So, conclusion from my side so far:
> > >
> > > * watermark/rowtime + primary key + changelog source != versioned
> > temporal
> > > table
> > > * operation time + primary key + changelog source == versioned temporal
> > > table
> > > * We may need something like 'PERIOD FOR SYSTEM_TIME(op_ts)' to define
> > the
> > > operation time
> > >
> > > ======================================================================
> > > However, there is still a pending question I don't have answer:
> > >
> > > Assuming you are doing a MIN aggregate on the operation time, that
> > doesn't
> > > work because the DELETE/UPDATE_BEFORE doesn't hold
> > > the previous value of operation time and thus can't retract.
> > >
> > > The operation time in fact should be metadata information (just like
> > > RowKind) which shouldn't be in the shema, and can't be accessed in
> > queries.
> > > But the PERIOD FOR SYSTEM_TIME syntax is in the schema part and should
> > > refer to a field in the schema...
> > >
> > > ======================================================================
> > >
> > > Anyway, let's focus on the operation_time vs rowtime problem first. Let
> > me
> > > know what's your thought!
> > >
> > > Best,
> > > Jark
> > >
> > > On Wed, 24 Jun 2020 at 23:49, Leonard Xu <[hidden email]> wrote:
> > >
> > > > Hi, kurt, Fabian
> > > >
> > > > After an offline discussion with Jark, We think that the  'PERIOD FOR
> > > > SYSTEM_TIME(operation_time)' statement might be needed now. Changelog
> > > table
> > > > is superset of insert-only table, use PRIMARY KEY and rowtime may
> work
> > > well
> > > > in insert-only or upsert source but has some problem in changelog
> > table.
> > > >
> > > > 'PERIOD FOR SYSTEM_TIME(operation_time)' in a temporal table
> > > > defines/maintains  the valid time of each row, the rowtime can not
> play
> > > the
> > > > history tracking function well.
> > > >
> > > > *# 1.*operation time (version time) *vs* rowtime (watermark)
> > > >
> > > > I will take an example to explain. The following changelog records
> came
> > > > from database table using debezium tool:
> > > > { "before":  null
> > > >   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> > > > "12:00:01"},
> > > >   "op":       "c",  //INSERT
> > > >   "ts_ms": 1592971201000 // 2020-06-24 12:00:02
> > > > }
> > > > { "before": {"currency": "Euro", "rate": 114, "gmt_modified":
> > > "12:00:05"},
> > > >   "after":    {"currency": "Euro", "rate": 118, "gmt_modified":
> > > > "12:00:05"},
> > > >   "op":       "u",  //UPDATE
> > > >   "ts_ms": 1592971206000 // 2020-06-24 12:00:06
> > > > }
> > > >
> > > > { "before": {"currency": "Euro", "rate": 118, "gmt_modified":
> > > "12:00:05"},
> > > >   "after":     null,
> > > >   "op":        "d",  //DELETE
> > > >   "ts_ms":  1593000011000  // 2020-06-24 20:00:11
> > > > }
> > > >
> > > > The rowtime should be the "gmt_modified" field that belongs to the
> > > > original record，the "ts_ms" is the the operation time when the DML
> > > > statement happen in the DB. For DELETE changelog record, its
> > > "gmt_modified"
> > > > field (12:00:05) can not reflect the real operation time (20:00:11).
> > > >
> > > > In temporal join case, we should maintain the valid time of each row.
> > For
> > > > a DELETE event, we should use the operation time of DELETE as the
> “end
> > > > time” of the row. That says, the record {"currency": "Euro", "rate":
> > 118}
> > > > is not exist anymore after “20:00:11”, not “12:00:05”.
> > > >
> > > > we would not access the record {"currency": "Euro", "rate": 118,
> > > > "gmt_modified": "12:00:05"} when rowtime is bigger than (12:00:05) if
> > we
> > > > use rowtime to track the history version, because the DELETE
> changelog
> > > > record also has rowtime (12:00:05) and will clear the record in
> state.
> > In
> > > > fact, the expected result is that the record expires until (20:00:11)
> > > when
> > > > the record is deleted rather than the last update time(20:00:11) in
> > > > materialized state.
> > > >
> > > > From this case, I found rowtime and operation time should be
> orthogonal
> > > in
> > > > temporal table scenario. The operation time should be strictly
> > > > monotonically increasing  (no out of order) and only be used for
> > > tracking a
> > > > history version of a changelog table, every history version of
> > changelog
> > > > table equals a database table snapshot due to the stream-table
> duality.
> > > >
> > > > *# 2.*The semantic of rowtime and watermark on changelog
> > > >
> > > > The rowtime and watermark can also be defined on a changelog table
> just
> > > > like other source backed queue, Flink supports cascaded window
> > > aggregation
> > > > (with ) in SQL like:
> > > > SELECT
> > > >      TUMBLE_ROWTIME(rowtime, INTERVAL '60' SECOND),
> > > >      MAX(rate) AS rate
> > > > FROM (
> > > >        SELECT
> > > >           MAX(rate) AS rate,
> > > >           TUMBLE_ROWTIME(rowtime, INTERVAL '5' SECOND) AS `rowtime`
> > > >        FROM currency
> > > >             GROUP BY TUMBLE(rowtime, INTERVAL '5' SECOND)
> > > >     )
> > > >     GROUP BY TUMBLE(rowtime, INTERVAL '60' SECOND
> > > >
> > > > We can think of the output of the first window aggregation as a
> > changelog
> > > > source of the second window aggregation. There are
> INSERT/UPDATE/DELETE
> > > > messages and also watermarks in the changelog stream. And the rowtime
> > in
> > > > the changelog stream is the `TUMBLE_ROWTIME` value (just like the
> > > > `gmt_modified` column in DB).
> > > >
> > > > *#  summary*
> > > >
> > > >    1. we should use ‘PERIOD FOR SYSTEM_TIME(operation_time) syntax to
> > > >    track history version by operation time rather than rowtime in
> > > temporal
> > > >    table scenario.
> > > >    2. we also support define a rowtime(watermark) on changelog table,
> > but
> > > >    the rowtime will not be used to track the history of changelog
> > stream.
> > > >
> > > >
> > > >
> > > > WDYT? please correct me if I am wrong.
> > > >
> > > >
> > > > Best,
> > > >
> > > > Leonard
> > > >
> > > >
> > > >
> > > >
> > > > 在 2020年6月24日，11:31，Leonard Xu <[hidden email]> 写道：
> > > >
> > > > Hi, everyone
> > > >
> > > > Thanks Fabian,Kurt for making the multiple version(event time)
> clear, I
> > > > also like the 'PERIOD FOR SYSTEM' syntax which supported in SQL
> > > standard. I
> > > > think we can add some explanation of the multiple version support in
> > the
> > > > future section of FLIP.
> > > >
> > > > For the PRIMARY KEY semantic, I agree with Jark's point that the
> > semantic
> > > > should unify both on changelog source and insert-only source.
> > > >
> > > > Currently, Flink supports PRIMARY KEY after FLIP-87, Flink uses
> PRIMARY
> > > > KEY NOT ENFORCED because Flink does not own the data like other DBMS
> > > therefore
> > > > Flink won't validate/enforce the key integrity and only trusts the
> > > external
> > > > systems. It is expected user and external system/application should
> > make
> > > > sure no deduplicated records happened when using NOT ENFORCED.
> > > >
> > > > (a) For PRIMARY KEY NOT ENFORCED semantic on changelog source:
> > > > It means the materialized changelogs (INSERT/UPDATE/DELETE) should be
> > > > unique on the primary key constraints.Flink assumes messages are in
> > order
> > > > on the primary key. Flink will use the PRIMARY KEY for some
> > optimization,
> > > > e.g. use the PRIMARY KEY to update the materialized state by key in
> > > > temporal join operator.
> > > >
> > > >
> > > > (b) For PRIMARY KEY NOT ENFORCED semantic on insert-only source:
> > > > It means records should be unique on the primary key constraints. If
> > > there
> > > > are INSERT records with duplicate primary key columns, the result of
> > SQL
> > > > query might be nondeterministic because it broken the PRIMARY KEY
> > > > constraints.
> > > >
> > > > Cheers,
> > > > Leonard
> > > >
> > > >
> > > > 在 2020年6月23日，23:35，Fabian Hueske <[hidden email]> 写道：
> > > >
> > > > Thanks Kurt,
> > > >
> > > > Yes, you are right.
> > > > The `PERIOD FOR SYSTEM_TIME` that you linked before corresponds to
> the
> > > > VERSION clause that I used and would explicitly define the versioning
> > of
> > > a
> > > > table.
> > > > I didn't know that the `PERIOD FOR SYSTEM_TIME` cause is already
> > defined
> > > by
> > > > the SQL standard.
> > > > I think we would need a slightly different syntax though because (so
> > far)
> > > > the validity of a row is determined by its own timestamp and the
> > > timestamp
> > > > of the next row.
> > > >
> > > > Adding a clause later solves the ambiguity issue for tables with
> > multiple
> > > > event-time attributes.
> > > > However, I'd feel more comfortable having such a cause and an
> explicit
> > > > definition of the temporal property from the beginning.
> > > > I guess this is a matter of personal preference so I'll go with the
> > > > majority if we decide that every table that has a primary key and an
> > > > event-time attribute should be usable in an event-time temporal table
> > > join.
> > > >
> > > > Thanks, Fabian
> > > >
> > > >
> > > > Am Di., 23. Juni 2020 um 16:58 Uhr schrieb Kurt Young <
> > [hidden email]
> > > >:
> > > >
> > > > Hi Fabian,
> > > >
> > > > I agree with you that implicitly letting event time to be the version
> > of
> > > > the table will
> > > > work in most cases, but not for all. That's the reason I mentioned
> > > `PERIOD
> > > > FOR` [1]
> > > > syntax in my first email, which is already in sql standard to
> represent
> > > the
> > > > validity of
> > > > each row in the table.
> > > >
> > > > If the event time can't be used, or multiple event time are defined,
> we
> > > > could still add
> > > > this syntax in the future.
> > > >
> > > > What do you think?
> > > >
> > > > [1]
> > > >
> > > >
> > > >
> > >
> >
> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
> > > > Best,
> > > > Kurt
> > > >
> > > >
> > > > On Tue, Jun 23, 2020 at 9:12 PM Fabian Hueske <[hidden email]>
> > wrote:
> > > >
> > > > Hi everyone,
> > > >
> > > > Every table with a primary key and an event-time attribute provides
> > what
> > > >
> > > > is
> > > >
> > > > needed for an event-time temporal table join.
> > > > I agree that, from a technical point of view, the TEMPORAL keyword is
> > not
> > > > required.
> > > >
> > > > I'm more sceptical about implicitly deriving the versioning
> information
> > > >
> > > > of
> > > >
> > > > a (temporal) table as the table's only event-time attribute.
> > > > In the query
> > > >
> > > > SELECT *
> > > > FROM orders o, rates r FOR SYSTEM_TIME AS OF o.ordertime
> > > > WHERE o.currency = r.currency
> > > >
> > > > the syntax of the temporal table join does not explicitly reference
> the
> > > > version of the temporal rates table.
> > > > Hence, the system needs a way to derive the version of temporal
> table.
> > > >
> > > > Implicitly using the (only) event-time attribute of a temporal table
> > > >
> > > > (rates
> > > >
> > > > in the example above) to identify the right version works in most
> > cases,
> > > > but probably not in all.
> > > > * What if a table has more than one event-time attribute?
> (TableSchema
> > is
> > > > designed to support multiple watermarks; queries with interval joins
> > > > produce tables with multiple event-time attributes, ...)
> > > > * What if the table does not have an event-time attribute in its
> schema
> > > >
> > > > but
> > > >
> > > > the version should only be provided as meta data?
> > > >
> > > > We could add a clause to define the version of a table, such as:
> > > >
> > > > CREATE TABLE rates (
> > > >   currency CHAR(3) NOT NULL PRIMARY KEY,
> > > >   rate DOUBLE,
> > > >   rowtime TIMESTAMP,
> > > >   WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE),
> > > > VERSION (rowtime)
> > > > WITH (...);
> > > >
> > > > The presence of a the VERSION clause (or whatever syntax) would
> > > >
> > > > explicitly
> > > >
> > > > define the version of a (temporal) table.
> > > > It would also render the need for the TEMPORAL keyword superfluous
> > > >
> > > > because
> > > >
> > > > there would be another indicator that a table can be used in a
> temporal
> > > > table join.
> > > >
> > > > I'm OK with not adding the TEMPORAL keyword, but I recommend that we
> > > >
> > > > think
> > > >
> > > > again about the proposed implicit definition of a table's version and
> > how
> > > > it might limit use in the future.
> > > >
> > > > Cheers,
> > > > Fabian
> > > >
> > > > Am Mo., 22. Juni 2020 um 16:14 Uhr schrieb Jark Wu <[hidden email]
> >:
> > > >
> > > > I'm also +1 for not adding the TEMPORAL keyword.
> > > >
> > > > +1 to make the PRIMARY KEY semantic clear for sources.
> > > > From my point of view:
> > > >
> > > > 1) PRIMARY KEY on changelog souruce:
> > > > It means that when the changelogs (INSERT/UPDATE/DELETE) are
> > > >
> > > > materialized,
> > > >
> > > > the materialized table should be unique on the primary key columns.
> > > > Flink assumes messages are in order on the primary key. Flink doesn't
> > > > validate/enforces the key integrity, but simply trust it (thus NOT
> > > > ENFORCED).
> > > > Flink will use the PRIMARY KEY for some optimization, e.g. use the
> > > >
> > > > PRIMARY
> > > >
> > > > KEY to update the materilized state by key in temporal join operator.
> > > >
> > > > 2) PRIMARY KEY on insert-only source:
> > > > I prefer to have the same semantic to the batch source and changelog
> > > > source, that it implies that records are not duplicate on the primary
> > > >
> > > > key.
> > > >
> > > > Flink just simply trust the primary key constraint, and doesn't valid
> > > >
> > > > it.
> > > >
> > > > If there is duplicate primary keys with INSERT changeflag, then
> result
> > > >
> > > > of
> > > >
> > > > Flink query might be wrong.
> > > >
> > > > If this is a TEMPORAL TABLE FUNCTION scenario, that source emits
> > > >
> > > > duplicate
> > > >
> > > > primary keys with INSERT changeflag, when we migrate this case to
> > > >
> > > > temporal
> > > >
> > > > table DDL,
> > > > I think this source should emit INSERT/UPDATE (UPSERT) messages
> instead
> > > >
> > > > of
> > > >
> > > > INSERT-only messages,  e.g. a Kafka compacted topic source?
> > > >
> > > > Best,
> > > > Jark
> > > >
> > > >
> > > > On Mon, 22 Jun 2020 at 17:04, Konstantin Knauf <[hidden email]>
> > > >
> > > > wrote:
> > > >
> > > >
> > > > Hi everyone,
> > > >
> > > > I also agree with Leonard/Kurt's proposal for CREATE TEMPORAL TABLE.
> > > >
> > > > Best,
> > > >
> > > > Konstantin
> > > >
> > > > On Mon, Jun 22, 2020 at 10:53 AM Kurt Young <[hidden email]>
> > > >
> > > > wrote:
> > > >
> > > >
> > > > I agree with Timo, semantic about primary key needs more thought
> > > >
> > > > and
> > > >
> > > > discussion, especially after FLIP-95 and FLIP-105.
> > > >
> > > > Best,
> > > > Kurt
> > > >
> > > >
> > > > On Mon, Jun 22, 2020 at 4:45 PM Timo Walther <[hidden email]>
> > > >
> > > > wrote:
> > > >
> > > >
> > > > Hi Leonard,
> > > >
> > > > thanks for the summary.
> > > >
> > > > After reading all of the previous arguments and working on
> > > >
> > > > FLIP-95. I
> > > >
> > > > would also lean towards the conclusion of not adding the TEMPORAL
> > > >
> > > > keyword.
> > > >
> > > >
> > > > After FLIP-95, what we considered as a CREATE TEMPORAL TABLE can
> > > >
> > > > be
> > > >
> > > > represented as a CREATE TABLE with PRIMARY KEY and WATERMARK. The
> > > >
> > > > FOR
> > > >
> > > > SYSTEM_TIME AS OF t would trigger the internal materialization
> > > >
> > > > and
> > > >
> > > > "temporal" logic.
> > > >
> > > > However, we should discuss the meaning of PRIMARY KEY again in
> > > >
> > > > this
> > > >
> > > > case. In a TEMPORAL TABLE scenario, the source would emit
> > > >
> > > > duplicate
> > > >
> > > > primary keys with INSERT changeflag but at different point in
> > > >
> > > > time.
> > > >
> > > > Currently, we require a PRIMARY KEY NOT ENFORCED declaration. The
> > > > changelog semantics of FLIP-95 and FLIP-105 don't work well with
> > > >
> > > > a
> > > >
> > > > primary key declaration.
> > > >
> > > > Regards,
> > > > Timo
> > > >
> > > >
> > > > On 20.06.20 17:08, Leonard Xu wrote:
> > > >
> > > > Hi everyone,
> > > >
> > > > Thanks for the nice discussion. I’d like to move forward the
> > > >
> > > > work,
> > > >
> > > > please let me simply summarize the main opinion and current
> > > >
> > > > divergences.
> > > >
> > > >
> > > > 1. The agreements have been achieved:
> > > >
> > > > 1.1 The motivation we're discussing temporal table DDL is just
> > > >
> > > > for
> > > >
> > > > creating temporal table in pure SQL to replace pre-process
> > > >
> > > > temporal
> > > >
> > > > table
> > > >
> > > > in YAML/Table API for usability.
> > > >
> > > > 1.2 The reason we use "TEMPORAL" keyword rather than “PERIOD
> > > >
> > > > FOR
> > > >
> > > > SYSTEM_TIME” is to make user understand easily.
> > > >
> > > > 1.3 For append-only table, it can convert to changelog table
> > > >
> > > > which
> > > >
> > > > has
> > > >
> > > > been discussed in FLIP-105, we assume the following temporal
> > > >
> > > > table
> > > >
> > > > is
> > > >
> > > > comes
> > > >
> > > > from changelog (Jark, fabian, Timo).
> > > >
> > > > 1.4 For temporal join syntax, using "FOR SYSTEM_TIME AS OF x"
> > > >
> > > > instead
> > > >
> > > > of
> > > >
> > > > the current `LATERAL TABLE(rates(x))`  has come to an
> > > >
> > > > agreement(Fabian,
> > > >
> > > > Timo, Seth, Konstantin, Kurt).
> > > >
> > > >
> > > > 2. The small divergence :
> > > >
> > > > About the definition syntax of the temporal table,
> > > >
> > > > CREATE [TEMPORAL] TABLE rates (
> > > >    currency CHAR(3) NOT NULL PRIMARY KEY,
> > > >    rate DOUBLE,
> > > >    rowtime TIMESTAMP,
> > > >    WATERMARK FOR rowtime AS rowtime - INTERVAL '5' MINUTE)
> > > > WITH (...);
> > > >
> > > > there is small divergence whether add "TEMPORAL" keyword or
> > > >
> > > > not.
> > > >
> > > >
> > > > 2.1  one opinion is using "CREATE TEMPORAL TABLE" (Timo,
> > > >
> > > > Fabian,
> > > >
> > > > Seth),
> > > >
> > > > the main advantages are:
> > > >
> > > > （1）"TEMPORAL" keyword is intuitive to indicate the history
> > > >
> > > > tracking
> > > >
> > > > semantics.
> > > >
> > > > （2）"TEMPORAL" keyword illustrates that queries can visit the
> > > >
> > > > previous
> > > >
> > > > versions of a table like other DBMS use "PERIOD FOR SYSTEM_TIME"
> > > >
> > > > keyword.
> > > >
> > > >
> > > > 2.2 the other is using "CREATE TABLE"(Kurt), the main
> > > >
> > > > advantages
> > > >
> > > > are:
> > > >
> > > > （1）Just primary key and time attribute can track previous
> > > >
> > > > versions
> > > >
> > > > of a
> > > >
> > > > table well.
> > > >
> > > > （2）The temporal behavior is triggered by temporal join syntax
> > > >
> > > > rather
> > > >
> > > > than in DDL, all Flink DDL table are dynamic table logically
> > > >
> > > > including
> > > >
> > > > temporal table. If we decide to use "TEMPORAL" keyword and treats
> > > >
> > > > changelog
> > > >
> > > > as temporal table, other tables backed queue like Kafka should
> > > >
> > > > also
> > > >
> > > > use
> > > >
> > > > "TEMPORAL" keyword.
> > > >
> > > >
> > > >
> > > > IMO, the statement “CREATE TEMPORARY TEMPORAL TABLE...” follows
> > > >
> > > > with
> > > >
> > > > 2.1
> > > >
> > > > may confuse users much. If we take a second to think about, for
> > > >
> > > > source/sink
> > > >
> > > > table which may backed queue (like kafka) or DB (like MySQL), we
> > > >
> > > > did
> > > >
> > > > not
> > > >
> > > > add any keyword in DDL to specify they are source or sinks, it
> > > >
> > > > works
> > > >
> > > > well.
> > > >
> > > > I think temporal table is the third one,  kafka data source and
> > > >
> > > > DB
> > > >
> > > > data
> > > >
> > > > source can play as a source/sink/temporal table depends on the
> > > > position/syntax that user put them in the query. The above rates
> > > >
> > > > table
> > > >
> > > >     - can be a source table if user put it at `SELECT * FROM
> > > >
> > > > rates;`
> > > >
> > > >     - can be a temporal table if user put it at `SELECT * FROM
> > > >
> > > > orders
> > > >
> > > > JOIN rates FOR SYSTEM_TIME AS OF orders.proctime
> > > >
> > > >              ON orders.currency = rates.currency;`
> > > >     - can be sink table if user put is at `INSERT INTO rates
> > > >
> > > > SELECT
> > > >
> > > > *
> > > >
> > > > FROM …; `
> > > >
> > > > From these cases, we found all tables defined in Flink should
> > > >
> > > > be
> > > >
> > > > dynamic table logically, the source/sink/temporal role depends on
> > > >
> > > > the
> > > >
> > > > position/syntax in user’s query.
> > > >
> > > >       In fact we have used similar syntax for current lookup
> > > >
> > > > table,
> > > >
> > > > we
> > > >
> > > > didn’t add “LOOKUP" or “TEMPORAL" keyword for lookup table and
> > > >
> > > > trigger
> > > >
> > > > the
> > > >
> > > > temporal join from the position/syntax(“FOR SYSTEM_TIME AS OF x")
> > > >
> > > > in
> > > >
> > > > query.
> > > >
> > > >
> > > > So, I prefer to resolve the small divergence with “CREATE
> > > >
> > > > TABLE”
> > > >
> > > > which
> > > >
> > > > (1) is more unified with our source/sink/temporal dynamic table
> > > >
> > > > conceptually,
> > > >
> > > > (2) is aligned with current lookup table,
> > > > (3) also make users learn less keyword.
> > > >
> > > > WDYT？
> > > >
> > > > Best,
> > > > Leonard Xu
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > --
> > > >
> > > > Konstantin Knauf
> > > >
> > > > https://twitter.com/snntrable
> > > >
> > > > https://github.com/knaufk
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > >
> >
> >
> > --
> > Best, Jingsong Lee
> >
>
>
> --
>
> Konstantin Knauf
>
> https://twitter.com/snntrable
>
> https://github.com/knaufk
>

[Leonard Xu]

Thanks Jingsong, Jark, Knauf, Seth for sharing your thoughts.

Although we discussed many details about the concept, I think it’s worth to clarify the semantic from long term goals. Temporal table concept was first imported in SQL:2011, I made some investigation of Temporal Table work mechanism in traditional DBMS which implements it like SQL Server[1], PostgreSQL[2]
 
In DBMS, Temporal Table is implemented as a pair of tables, a current table and a history table.The current table contains the current value of each row, the history table contains each previous value for each row. Each row contains a time range constructed by RowStartTime and RowEndTime to define the period validity of the row. The RowStartTime and RowEndTime is changed by DBMS when a DML operation happened, Given a simple temporal table in SQL Server to show how it works:
        CREATE TABLE dbo.currency (
    [currency]  VARCHAR(10) NOT NULL PRIMARY KEY,
    [rate] INT,
    [RowStart] DATETIME2 GENERATED ALWAYS AS ROW START,
    [RowEnd] DATETIME2 GENERATED ALWAYS AS ROW END,
    PERIOD FOR SYSTEM_TIME (RowStart, RowEnd)
    ) WITH
        (SYSTEM_VERSIONING = ON (HISTORY_TABLE = dbo.currency_History));

1> select * from currency;                                                                       // The initial  test data, the RowEndTime is the max value of timestamp type
    currency   rate        RowStart                               RowEnd
    ---------- ----------- -------------------------------------- --------------------------------------
    Euro               114            2020-06-29 15:06:24.7459246            9999-12-31 23:59:59.9999999
    US Dollar       102            2020-06-29 15:06:24.7503288            9999-12-31 23:59:59.9999999
1> UPDATE dbo.currency SET [rate] = 118 WHERE currency = 'Euro’;   // UPDATE Euro currency
2> select * from currency_History;                                                                  // The history table increased a record that represents the validity period of record (Euro,114)
    currency   rate        RowStart                               RowEnd
    ---------- ----------- -------------------------------------- --------------------------------------
    Euro               114            2020-06-29 15:06:24.7459246            2020-06-29 15:07:01.1245406
1> DELETE FROM dbo.currency WHERE currency = 'Euro’;           // DELETE Euro currency
1> select * from currency_History;
    currency   rate        RowStart                               RowEnd                 // The history table also increased another record that represents the validity period of record (Euro, 118)
    ---------- ----------- -------------------------------------- --------------------------------------
    Euro               114            2020-06-29 15:06:24.7459246            2020-06-29 15:07:01.1245406
    Euro               118            2020-06-29 15:07:01.1245406            2020-06-29 15:07:11.2981995
1> select * from currency;
    currency   rate        RowStart                               RowEnd                 // Current table only keep the latest value
    ---------- ----------- -------------------------------------- --------------------------------------
    US Dollar       102            2020-06-29 15:06:24.7503288            9999-12-31 23:59:59.9999999

The history table is very important for history version tracking, pleas note the DELETE operation also increase a record in history table and the record’s RowEndTime is the system time that the DELETE operation happened. In one word, temporal table use time range [RowStart, RowEnd) to mark period validity, store all versions’ records in history table for history tracking, use DBMS operation time to change the RowStart or RowEnd.

Back to our Flink World, temporal table with event time attribute works well in data source that contains INSERT, UPDATE messages except DELETE currently. Let us see what happened in DELETE message  scenario(i.e. changelog source), both DBMS Temporal Table and other general table can capture data change by CDC tools and have same format, I used debezuim to capture a SQL server table changes:

1> select * from currency;
    currency   rate        RowStart                               RowEnd
    ---------- ----------- -------------------------------------- --------------------------------------
    Euro               118            2020-06-29 15:07:01.1245406            9999-12-31 23:59:59.9999999
    US Dollar          102            2020-06-29 15:06:24.7503288         9999-12-31 23:59:59.9999999
1> DELETE FROM dbo.currency WHERE currency = 'Euro’;               // DELETE Euro currency
1> select * from currency_History;
    currency   rate        RowStart                               RowEnd
    ---------- ----------- -------------------------------------- --------------------------------------
    Euro               118            2020-06-29 15:07:01.1245406            2020-06-29 15:07:11.2981995

{ // The DELETE record produced by  CDC tools(both debezuim and canal are same)
    "before": {
      "currency": "Euro",
      "rate": 118,
      "RowStart": 1593443221124540600,  //2020-06-29 15:07:01.1245406
      "RowEnd": -4852116231933722724   //9999-12-31 23:59:59.9999999
    },
    "after": null,
    "op": "d”,  // DELETE operation
    "ts_ms": 15934432361354,                //2020-06-29 15:07:16.354, the ’ts_ms’ value is bigger than the record delete operation time(2020-06-29 15:07:11.2981995)
    "transaction": null
}

The main problem is that the DELETE record only contains current table message which  does not contain the expected RowEnd (2020-06-29 15:07:11.2981995) in history table. Without the exact RowEndTime, it’s impossible to obtain exact previous version of temporal table in Flink, `ts_ms` filed in CDC record is an approximate time of RowEndTime but  it depends on CDC tool status and can not equal the RowEndTime from semantics angle.

Current Temporal Table Function supports:
        (1) Define temporal table backed upsert data source with process time
        (2) Define temporal table backed upsert data source with event time
I think  the proposed Temporal table currently could support:
        (1) Define temporal table backed upsert(include delete) data source with process time
        (2) Define temporal table backed upsert  data source with event time
        (3) Do not support define temporal table backed  data source that contains DELETE message with event time. Because most CDC tools can not obtain the exact DELETE operation  time currently, the “ts_ms” field from meta is just an approximate time which will break event time semantics. And we can support it when CDC tools have the ability to
obtain/extract the DML operation time.
 
And this has get consensus from me, Jingsong, Jark and Kurt after offline discuss, the opinions from Knauf and Seth looks like same with us.

I’ll prepare a design doc for temporal table, thanks everyone involving and please let me know if you have any concern.

Best,
Leonard Xu
[1] https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15 <https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15>
[2] https://pgxn.org/dist/temporal_tables/ <https://pgxn.org/dist/temporal_tables/>

 

[Konstantin Knauf-4]

Hi Leonard,

Thank you for the summary. I don't fully understand the implications of
(3). Would we support a temporal join with a changelog stream with
event time semantics by ignoring DELETE messages or would it be completed
unsupported. I mean something like the following sequence of statements:

CREATE TABLE currency_rates (
  currencyId BIGINT PRIMARY KEY,
  rate DECIMAL(10, 2)) WITH (
 'connector' = 'kafka',
 'format' = 'debezium-json'
)
*CREATE* TABLE transactions (
  currencyId BIGINT,
  transactionTime TIMESTAMP(3)) WITH (

)


SELECT
  ...FROM
  transactions AS t
  JOIN currency_rates FOR SYSTEM_TIME AS OF t.transactionTime AS r
  ON r.currency = t.currency

Cheers,

Konstantin

On Fri, Jul 3, 2020 at 4:52 PM Leonard Xu <[hidden email]> wrote:

> Thanks Jingsong, Jark, Knauf, Seth for sharing your thoughts.
>
> Although we discussed many details about the concept, I think it’s worth
> to clarify the semantic from long term goals. Temporal table concept was
> first imported in SQL:2011, I made some investigation of Temporal Table
> work mechanism in traditional DBMS which implements it like SQL Server[1],
> PostgreSQL[2]
>
> In DBMS, Temporal Table is implemented as a pair of tables, *a current
> table and a history table*.The current table contains the current value
> of each row, the history table contains each previous value for each row.
> Each row contains a time range constructed by RowStartTime and RowEndTime
> to define the period validity of the row. The RowStartTime and RowEndTime
> is changed by DBMS when a DML operation happened, Given a simple temporal
> table in SQL Server to show how it works:
> CREATE TABLE dbo.currency (
> [currency] VARCHAR(10) NOT NULL PRIMARY KEY, [rate] INT, [RowStart]
> DATETIME2 GENERATED ALWAYS AS ROW START, [RowEnd] DATETIME2 GENERATED
> ALWAYS AS ROW END, PERIOD FOR SYSTEM_TIME (RowStart, RowEnd) ) WITH (SYSTEM_VERSIONING
> = ON (HISTORY_TABLE = dbo.currency_History));
>
> 1> select * from currency; // *The initial test data, the RowEndTime is
> the max value of timestamp type* currency rate RowStart RowEnd ----------
> ----------- --------------------------------------
> -------------------------------------- Euro 114 2020-06-29 15:06:24.7459246
> 9999-12-31 23:59:59.9999999 US Dollar 102 2020-06-29 15:06:24.7503288
> 9999-12-31 23:59:59.9999999 1>* UPDATE dbo.currency SET [rate] = 118
> WHERE currency = 'Euro’*; //* UPDATE **Euro currency* 2> select * from
> *currency_History*; // *The history table increased a record that
> represents the validity period of record (Euro,114)* currency rate
> RowStart RowEnd ---------- -----------
> --------------------------------------
> -------------------------------------- Euro 114 2020-06-29 15:06:24.7459246
> 2020-06-29 15:07:01.1245406 1> *DELETE FROM dbo.currency WHERE currency =
> 'Euro’;* //* DELETE **Euro currency* 1> select * from *currency_History*;
> currency rate RowStart RowEnd // *The history table also increased
> another record that represents the validity period of record (Euro, 118)*
> ---------- ----------- --------------------------------------
> -------------------------------------- Euro 114 2020-06-29 15:06:24.7459246
> 2020-06-29 15:07:01.1245406 Euro 118 2020-06-29 15:07:01.1245406 2020-06-29
> 15:07:11.2981995 1> select * from currency; currency rate RowStart RowEnd
> // *Current table only keep the latest value * ---------- -----------
> --------------------------------------
> -------------------------------------- US Dollar 102 2020-06-29
> 15:06:24.7503288 9999-12-31 23:59:59.9999999
>
> The history table is very important for history version tracking, pleas
> note the *DELETE* operation also increase a record in history table and
> the record’s RowEndTime is the system time that the DELETE operation
> happened. In one word, temporal table use time range [RowStart, RowEnd) to
> mark period validity, store all versions’ records in history table for
> history tracking, use DBMS operation time to change the RowStart or
> RowEnd.
>
> Back to our Flink World, temporal table with event time attribute works
> well in data source that contains INSERT, UPDATE messages except DELETE currently.
> Let us see what happened in DELETE message scenario(i.e. changelog
> source), both DBMS Temporal Table and other general table can capture data
> change by CDC tools and have same format, I used debezuim to capture a SQL
> server table changes:
>
> 1> select * from currency; currency rate RowStart RowEnd ----------
> ----------- --------------------------------------
> -------------------------------------- Euro 118 2020-06-29 15:07:01.1245406
> 9999-12-31 23:59:59.9999999 US Dollar 102 2020-06-29 15:06:24.7503288
> 9999-12-31 23:59:59.9999999 1>* DELETE FROM dbo.currency WHERE currency =
> 'Euro’; * //* DELETE **Euro currency* 1> select * from currency_History;
> currency rate RowStart RowEnd ---------- -----------
> --------------------------------------
> -------------------------------------- Euro 118 2020-06-29 15:07:01.1245406 *2020-06-29
> 15:07:11.2981995* { // *The DELETE record produced by CDC tools(both
> debezuim and canal are same)* "before": { "currency": "Euro", "rate":
> 118, "RowStart": 1593443221124540600, //2020-06-29 15:07:01.1245406
> "RowEnd": -4852116231933722724 //9999-12-31 23:59:59.9999999 }, "after":
> null, "op": "d”, // DELETE operation "ts_ms": 15934432361354, //*2020-06-29
> 15:07:16.354, the ’ts_ms’ value is bigger than the record delete operation
> time(**2020-06-29 15:07:11.2981995**)*
> * "transaction": null* }
>
> The main problem is that the *DELETE* record only contains current table
> message which  does not contain the expected RowEnd (*2020-06-29
> 15:07:11.2981995*) in history table. Without the exact RowEndTime, it’s
> impossible to obtain exact previous version of temporal table in Flink,
> `ts_ms` filed in CDC record is an approximate time of RowEndTime but  it
> depends on CDC tool status and can not equal the RowEndTime from semantics
> angle.
>
> Current Temporal Table Function supports:
> (1) Define temporal table backed upsert data source with process time
> (2) Define temporal table backed upsert data source with event time
> I think  the proposed Temporal table currently could support:
> (1) Define temporal table backed upsert(include delete) data source with
> process time
> (2) Define temporal table backed upsert  data source with event time
> (3) Do not support define temporal table backed  data source that
> contains DELETE message with event time. Because most CDC tools can not
> obtain the exact DELETE operation  time currently, the “ts_ms” field from
> meta is just an approximate time which will break event time semantics.
> And we can support it when CDC tools have the ability to
> obtain/extract the DML operation time.
>
> And this has get consensus from me, Jingsong, Jark and Kurt after offline
> discuss, the opinions from Knauf and Seth looks like same with us.
>
> I’ll prepare a design doc for temporal table, thanks everyone involving
> and please let me know if you have any concern.
>
> Best,
> Leonard Xu
> [1]
> https://docs.microsoft.com/en-us/sql/relational-databases/tables/temporal-tables?view=sql-server-ver15
> [2] https://pgxn.org/dist/temporal_tables/
>
>
>

--

Konstantin Knauf

https://twitter.com/snntrable

https://github.com/knaufk

[Leonard Xu]

Hi, Konstantin

> . Would we support a temporal join with a changelog stream with
> event time semantics by ignoring DELETE messages or would it be completed
> unsupported.

I don’t know the percentage of this feature in temporal scenarios.

Comparing to support the approximate event time join by ignoring DELETE message or by extracting an approximate event time for DELET message,  I’m not sure is this acceptable for user even if we have explained the limitation of approximate event time join, I tend to do not support this feature, because we can not ensure the semantic of event time and it may lead an incorrect result for user in some scenarios.

If the percentage is highly enough and most user cases can accept the approximate  event time, I'm ok to support it  for usability although it doesn’t implements the event time semantic strictly.

Cheers,
Leonard Xu  
 

[Seth Wiesman-4]

Hi Leonard,

Regarding DELETE operations I tend to have the opposite reaction. I spend a
lot of time working with production Flink users across a large number of
organizations and to say we don't support temporal tables that include
DELETEs will be a blocker for adoption. Even organizations that claim to
never delete rows still occasionally due so per  GDPR requests or other
regulations.

I actually do think users will understand the limitations. Flink today has
a very clear value proposition around correctness, your results are as
correct as the input data provided. This does not change under support for
DELETE records. Flink is providing the most correct results possible based
on the resolution of the fields as generated by 3rd party systems. As
Debezium and other CDC libraries become more accurate, so will Flink.

Seth

On Fri, Jul 3, 2020 at 11:00 PM Leonard Xu <[hidden email]> wrote:

> Hi, Konstantin
>
> . Would we support a temporal join with a changelog stream with
> event time semantics by ignoring DELETE messages or would it be completed
> unsupported.
>
>
> I don’t know the percentage of this feature in temporal scenarios.
>
> Comparing to support the approximate event time join by ignoring DELETE
> message or by extracting an approximate event time for DELET message,  I’m
> not sure is this acceptable for user even if we have explained the
> limitation of approximate event time join, I tend to do not support this
> feature, because we can not ensure the semantic of event time and it may
> lead an incorrect result for user in some scenarios.
>
> If the percentage is highly enough and most user cases can accept the
> approximate  event time, I'm ok to support it  for usability although it
> doesn’t implements the event time semantic strictly.
>
> Cheers,
> Leonard Xu
>
>
>

[Seth Wiesman-4]

As an aside, I conceptually view temporal table joins to be semantically
equivalent to look up table joins. They are just two different ways of
consuming the same data.

Seth

On Mon, Jul 6, 2020 at 8:56 AM Seth Wiesman <[hidden email]> wrote:

> Hi Leonard,
>
> Regarding DELETE operations I tend to have the opposite reaction. I spend
> a lot of time working with production Flink users across a large number of
> organizations and to say we don't support temporal tables that include
> DELETEs will be a blocker for adoption. Even organizations that claim to
> never delete rows still occasionally due so per  GDPR requests or other
> regulations.
>
> I actually do think users will understand the limitations. Flink today has
> a very clear value proposition around correctness, your results are as
> correct as the input data provided. This does not change under support for
> DELETE records. Flink is providing the most correct results possible based
> on the resolution of the fields as generated by 3rd party systems. As
> Debezium and other CDC libraries become more accurate, so will Flink.
>
> Seth
>
> On Fri, Jul 3, 2020 at 11:00 PM Leonard Xu <[hidden email]> wrote:
>
>> Hi, Konstantin
>>
>> . Would we support a temporal join with a changelog stream with
>> event time semantics by ignoring DELETE messages or would it be completed
>> unsupported.
>>
>>
>> I don’t know the percentage of this feature in temporal scenarios.
>>
>> Comparing to support the approximate event time join by ignoring DELETE
>> message or by extracting an approximate event time for DELET message,  I’m
>> not sure is this acceptable for user even if we have explained the
>> limitation of approximate event time join, I tend to do not support this
>> feature, because we can not ensure the semantic of event time and it may
>> lead an incorrect result for user in some scenarios.
>>
>> If the percentage is highly enough and most user cases can accept the
>> approximate  event time, I'm ok to support it  for usability although it
>> doesn’t implements the event time semantic strictly.
>>
>> Cheers,
>> Leonard Xu
>>
>>
>>

[Leonard Xu]

Hi, Seth

Thanks for your explanation of user cases, and you’re wright the look up join/table is one kind of temporal table join/table which tracks latest snapshot of external  DB-like tables, it's why we proposed use same temporal join syntax.

In fact, I have invested and checked Debezuim format and Canal format more these days, and we can obtain the extract DML operation time from their meta information which comes from DB bin-log.  Although extracting meta information from record is a part of FLIP-107 scope[1], at least we have a way to extract the correct operation time. Event we can obtain the expected operation time, there’re some problems.

(1) For support changelog source backed CDC tools, a problem is that can we use the temporal table as a general source table which may followed by some aggregation operations,  more accurate is wether the aggregation operator can use the DELETE record that we just updated the “correct” operation time to retract a record, maybe not. This will pull us back to the discussion of operation time VS event time, it’s a real cool but complicated topic see above discussion from mine and @Jark’s.

(2) For upsert source that defines PRIMARY KEY and may contains multiple records under a PK, the main problem is the  PK semantic，the multiple records under a PK broke the unique semantic on a table. We need to walk around this by (a) Adding another primary key keyword and explain the upsert semantic (b) Creating temporal table base on a view that is the deduplicated result of source table[2].

I’m working on (2), and if we want to support（1）i.e. support DELETE entirely, that’s really a big challenge but I also think wright thing for long term.

If we decide to do (1), we need import operation time concept firstly, we need change the codebase for deal the operation time header in many places secondly, and finally explain and tell users how to understand and use temporal table.

I’m a little worried about it’s valuable enough, I proposed only support (2) because it is a good replacement of current Temporal Table Function and will not introduce more concept and works.

Jark, Jingsong, Konstantin, WDYT?


Best,
Leonard Xu  
[1] https://cwiki.apache.org/confluence/display/FLINK/FLIP-107%3A+Reading+table+columns+from+different+parts+of+source+records#FLIP107:Readingtablecolumnsfromdifferentpartsofsourcerecords-Accessread-onlymetadatae.g.partition <https://cwiki.apache.org/confluence/display/FLINK/FLIP-107:+Reading+table+columns+from+different+parts+of+source+records#FLIP107:Readingtablecolumnsfromdifferentpartsofsourcerecords-Accessread-onlymetadatae.g.partition>
[2] https://ci.apache.org/projects/flink/flink-docs-master/dev/table/sql/queries.html#deduplication <https://ci.apache.org/projects/flink/flink-docs-master/dev/table/sql/queries.html#deduplication>

> 在 2020年7月6日，22:02，Seth Wiesman <[hidden email]> 写道：
>
> As an aside, I conceptually view temporal table joins to be semantically equivalent to look up table joins. They are just two different ways of consuming the same data.
>
> Seth
>
> On Mon, Jul 6, 2020 at 8:56 AM Seth Wiesman <[hidden email] <mailto:[hidden email]>> wrote:
> Hi Leonard,
>
> Regarding DELETE operations I tend to have the opposite reaction. I spend a lot of time working with production Flink users across a large number of organizations and to say we don't support temporal tables that include DELETEs will be a blocker for adoption. Even organizations that claim to never delete rows still occasionally due so per  GDPR requests or other regulations.
>
> I actually do think users will understand the limitations. Flink today has a very clear value proposition around correctness, your results are as correct as the input data provided. This does not change under support for DELETE records. Flink is providing the most correct results possible based on the resolution of the fields as generated by 3rd party systems. As Debezium and other CDC libraries become more accurate, so will Flink.
>
> Seth
>
> On Fri, Jul 3, 2020 at 11:00 PM Leonard Xu <[hidden email] <mailto:[hidden email]>> wrote:
> Hi, Konstantin
>
>> . Would we support a temporal join with a changelog stream with
>> event time semantics by ignoring DELETE messages or would it be completed
>> unsupported.
>
> I don’t know the percentage of this feature in temporal scenarios.
>
> Comparing to support the approximate event time join by ignoring DELETE message or by extracting an approximate event time for DELET message,  I’m not sure is this acceptable for user even if we have explained the limitation of approximate event time join, I tend to do not support this feature, because we can not ensure the semantic of event time and it may lead an incorrect result for user in some scenarios.
>
> If the percentage is highly enough and most user cases can accept the approximate  event time, I'm ok to support it  for usability although it doesn’t implements the event time semantic strictly.
>
> Cheers,
> Leonard Xu  
>  
>

[Benchao Li-2]

Hi everyone,

Thanks a lot for the great discussions so far.

After reading through the long discussion, I still have one question.
Currently the temporal table function supports both event time and proc
time joining.
If we use "FOR SYSTEM_TIME AS OF" syntax without "TEMPORAL" keyword in DDL,
does it mean we can only use temporal table function join with event time?
If we can, how do we distinguish it with current temporal table (also known
as dimension table)?

Maybe I'm missing something here. Correct me if I'm wrong.

Leonard Xu <[hidden email]> 于2020年7月6日周一 下午11:34写道：

> Hi, Seth
>
> Thanks for your explanation of user cases, and you’re wright the look up
> join/table is one kind of temporal table join/table which tracks latest
> snapshot of external  DB-like tables, it's why we proposed use same
> temporal join syntax.
>
> In fact, I have invested and checked Debezuim format and Canal format more
> these days, and we can obtain the extract DML operation time from their
> meta information which comes from DB bin-log.  Although extracting meta
> information from record is a part of FLIP-107 scope[1], at least we have a
> way to extract the correct operation time. Event we can obtain the expected
> operation time, there’re some problems.
>
> (1) For support changelog source backed CDC tools, a problem is that can
> we use the temporal table as a general source table which may followed by
> some aggregation operations,  more accurate is wether the aggregation
> operator can use the DELETE record that we just updated the “correct”
> operation time to retract a record, maybe not. This will pull us back to
> the discussion of operation time VS event time, it’s a real cool but
> complicated topic see above discussion from mine and @Jark’s.
>
> (2) For upsert source that defines PRIMARY KEY and may contains multiple
> records under a PK, the main problem is the  PK semantic，the multiple
> records under a PK broke the unique semantic on a table. We need to walk
> around this by (a) Adding another primary key keyword and explain the
> upsert semantic (b) Creating temporal table base on a view that is the
> deduplicated result of source table[2].
>
> I’m working on (2), and if we want to support（1）i.e. support DELETE
> entirely, that’s really a big challenge but I also think wright thing for
> long term.
>
> If we decide to do (1), we need import operation time concept firstly, we
> need change the codebase for deal the operation time header in many places
> secondly, and finally explain and tell users how to understand and use
> temporal table.
>
> I’m a little worried about it’s valuable enough, I proposed only support
> (2) because it is a good replacement of current Temporal Table Function and
> will not introduce more concept and works.
>
> Jark, Jingsong, Konstantin, WDYT?
>
>
> Best,
> Leonard Xu
> [1]
> https://cwiki.apache.org/confluence/display/FLINK/FLIP-107%3A+Reading+table+columns+from+different+parts+of+source+records#FLIP107:Readingtablecolumnsfromdifferentpartsofsourcerecords-Accessread-onlymetadatae.g.partition
> <
> https://cwiki.apache.org/confluence/display/FLINK/FLIP-107:+Reading+table+columns+from+different+parts+of+source+records#FLIP107:Readingtablecolumnsfromdifferentpartsofsourcerecords-Accessread-onlymetadatae.g.partition
> >
> [2]
> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/sql/queries.html#deduplication
> <
> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/sql/queries.html#deduplication
> >
>
>
> > 在 2020年7月6日，22:02，Seth Wiesman <[hidden email]> 写道：
> >
> > As an aside, I conceptually view temporal table joins to be semantically
> equivalent to look up table joins. They are just two different ways of
> consuming the same data.
> >
> > Seth
> >
> > On Mon, Jul 6, 2020 at 8:56 AM Seth Wiesman <[hidden email]
> <mailto:[hidden email]>> wrote:
> > Hi Leonard,
> >
> > Regarding DELETE operations I tend to have the opposite reaction. I
> spend a lot of time working with production Flink users across a large
> number of organizations and to say we don't support temporal tables that
> include DELETEs will be a blocker for adoption. Even organizations that
> claim to never delete rows still occasionally due so per  GDPR requests or
> other regulations.
> >
> > I actually do think users will understand the limitations. Flink today
> has a very clear value proposition around correctness, your results are as
> correct as the input data provided. This does not change under support for
> DELETE records. Flink is providing the most correct results possible based
> on the resolution of the fields as generated by 3rd party systems. As
> Debezium and other CDC libraries become more accurate, so will Flink.
> >
> > Seth
> >
> > On Fri, Jul 3, 2020 at 11:00 PM Leonard Xu <[hidden email] <mailto:
> [hidden email]>> wrote:
> > Hi, Konstantin
> >
> >> . Would we support a temporal join with a changelog stream with
> >> event time semantics by ignoring DELETE messages or would it be
> completed
> >> unsupported.
> >
> > I don’t know the percentage of this feature in temporal scenarios.
> >
> > Comparing to support the approximate event time join by ignoring DELETE
> message or by extracting an approximate event time for DELET message,  I’m
> not sure is this acceptable for user even if we have explained the
> limitation of approximate event time join, I tend to do not support this
> feature, because we can not ensure the semantic of event time and it may
> lead an incorrect result for user in some scenarios.
> >
> > If the percentage is highly enough and most user cases can accept the
> approximate  event time, I'm ok to support it  for usability although it
> doesn’t implements the event time semantic strictly.
> >
> > Cheers,
> > Leonard Xu
> >
> >
>
>

--

Best,
Benchao Li

[Seth Wiesman-4]

Hey Leonard,

Agreed, this is a fun discussion!

(1) For support changelog source backed CDC tools, a problem is that can we
> use the temporal table as a general source table which may followed by some
> aggregation operations,  more accurate is wether the aggregation operator
> can use the DELETE record that we just updated the “correct” operation time
> to retract a record, maybe not. This will pull us back to the discussion of
> operation time VS event time, it’s a real cool but complicated topic see
> above discussion from mine and @Jark’s.
>

I fully agree this is a complicated topic, however, I don't think its
actually a problem that needs to be solved for the first version of this
feature. My proposal is to disallow using upsert streams as temporal tables
if an aggregation operation has been applied. Going back to my notion that
temporal tables are a tool for performing streaming star schema
denormalization, the dimension table in a star schema is rarely aggregated
pre-join. In the case of a CDC stream of currency rates joined to
transactions, the CDC stream only needs to support filter pushdowns and
map-like transformations before being joined. I believe this is a
reasonable limitation we can impose that will unblock a large percentage of
use cases, and once we better understand the semantics of the correct
operation in a retraction the limitation can be removed in future versions
while remaining backward compatible.



CREATE TABLE currency_rates (
  currencyId BIGINT PRIMARY KEY,
  rate DECIMAL(10, 2)) WITH (
 'connector' = 'kafka',
 'format' = 'debezium-json'
)
*CREATE* TABLE transactions (
  currencyId BIGINT,
  transactionTime TIMESTAMP(3)) WITH (

)

-- Uner my proposal this query would be supported because the currency_rates

-- table is used in a temporal join without any aggregations having been applied

CREATE VIEW AS working_query
SELECT
  ...FROM
  transactions AS t
  JOIN currency_rates FOR SYSTEM_TIME AS OF t.transactionTime AS r
  ON r.currency = t.currencyId

-- However, this query would be rejected by the planner until we
determine the proper time semantics of a retacation

CREATE VIEW AS post_agg_stream SELECT currencyId, AVG(rate)* as *rate*
FROM *currency_rates

CREATE VIEW AS rejected_query
SELECT
  ...FROM
  transactions AS t
  JOIN currency_rates FOR SYSTEM_TIME AS OF t.transactionTime AS r
  ON r.currency = t.currency



(2) For upsert source that defines PRIMARY KEY and may contains multiple
> records under a PK, the main problem is the  PK semantic，the multiple
> records under a PK broke the unique semantic on a table. We need to walk
> around this by (a) Adding another primary key keyword and explain the
> upsert semantic (b) Creating temporal table base on a view that is the
> deduplicated result of source table[2].
>

This feels like more of a bikeshedding question than a blocker and I look
forward to seeing what you come up with!

Seth

On Mon, Jul 6, 2020 at 10:59 AM Benchao Li <[hidden email]> wrote:

> Hi everyone,
>
> Thanks a lot for the great discussions so far.
>
> After reading through the long discussion, I still have one question.
> Currently the temporal table function supports both event time and proc
> time joining.
> If we use "FOR SYSTEM_TIME AS OF" syntax without "TEMPORAL" keyword in DDL,
> does it mean we can only use temporal table function join with event time?
> If we can, how do we distinguish it with current temporal table (also
> known as dimension table)?
>
> Maybe I'm missing something here. Correct me if I'm wrong.
>
> Leonard Xu <[hidden email]> 于2020年7月6日周一 下午11:34写道：
>
>> Hi, Seth
>>
>> Thanks for your explanation of user cases, and you’re wright the look up
>> join/table is one kind of temporal table join/table which tracks latest
>> snapshot of external  DB-like tables, it's why we proposed use same
>> temporal join syntax.
>>
>> In fact, I have invested and checked Debezuim format and Canal format
>> more these days, and we can obtain the extract DML operation time from
>> their meta information which comes from DB bin-log.  Although extracting
>> meta information from record is a part of FLIP-107 scope[1], at least we
>> have a way to extract the correct operation time. Event we can obtain the
>> expected operation time, there’re some problems.
>>
>> (1) For support changelog source backed CDC tools, a problem is that can
>> we use the temporal table as a general source table which may followed by
>> some aggregation operations,  more accurate is wether the aggregation
>> operator can use the DELETE record that we just updated the “correct”
>> operation time to retract a record, maybe not. This will pull us back to
>> the discussion of operation time VS event time, it’s a real cool but
>> complicated topic see above discussion from mine and @Jark’s.
>>
>> (2) For upsert source that defines PRIMARY KEY and may contains multiple
>> records under a PK, the main problem is the  PK semantic，the multiple
>> records under a PK broke the unique semantic on a table. We need to walk
>> around this by (a) Adding another primary key keyword and explain the
>> upsert semantic (b) Creating temporal table base on a view that is the
>> deduplicated result of source table[2].
>>
>> I’m working on (2), and if we want to support（1）i.e. support DELETE
>> entirely, that’s really a big challenge but I also think wright thing for
>> long term.
>>
>> If we decide to do (1), we need import operation time concept firstly, we
>> need change the codebase for deal the operation time header in many places
>> secondly, and finally explain and tell users how to understand and use
>> temporal table.
>>
>> I’m a little worried about it’s valuable enough, I proposed only support
>> (2) because it is a good replacement of current Temporal Table Function and
>> will not introduce more concept and works.
>>
>> Jark, Jingsong, Konstantin, WDYT?
>>
>>
>> Best,
>> Leonard Xu
>> [1]
>> https://cwiki.apache.org/confluence/display/FLINK/FLIP-107%3A+Reading+table+columns+from+different+parts+of+source+records#FLIP107:Readingtablecolumnsfromdifferentpartsofsourcerecords-Accessread-onlymetadatae.g.partition
>> <
>> https://cwiki.apache.org/confluence/display/FLINK/FLIP-107:+Reading+table+columns+from+different+parts+of+source+records#FLIP107:Readingtablecolumnsfromdifferentpartsofsourcerecords-Accessread-onlymetadatae.g.partition
>> >
>> [2]
>> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/sql/queries.html#deduplication
>> <
>> https://ci.apache.org/projects/flink/flink-docs-master/dev/table/sql/queries.html#deduplication
>> >
>>
>>
>> > 在 2020年7月6日，22:02，Seth Wiesman <[hidden email]> 写道：
>> >
>> > As an aside, I conceptually view temporal table joins to be
>> semantically equivalent to look up table joins. They are just two different
>> ways of consuming the same data.
>> >
>> > Seth
>> >
>> > On Mon, Jul 6, 2020 at 8:56 AM Seth Wiesman <[hidden email]
>> <mailto:[hidden email]>> wrote:
>> > Hi Leonard,
>> >
>> > Regarding DELETE operations I tend to have the opposite reaction. I
>> spend a lot of time working with production Flink users across a large
>> number of organizations and to say we don't support temporal tables that
>> include DELETEs will be a blocker for adoption. Even organizations that
>> claim to never delete rows still occasionally due so per  GDPR requests or
>> other regulations.
>> >
>> > I actually do think users will understand the limitations. Flink today
>> has a very clear value proposition around correctness, your results are as
>> correct as the input data provided. This does not change under support for
>> DELETE records. Flink is providing the most correct results possible based
>> on the resolution of the fields as generated by 3rd party systems. As
>> Debezium and other CDC libraries become more accurate, so will Flink.
>> >
>> > Seth
>> >
>> > On Fri, Jul 3, 2020 at 11:00 PM Leonard Xu <[hidden email] <mailto:
>> [hidden email]>> wrote:
>> > Hi, Konstantin
>> >
>> >> . Would we support a temporal join with a changelog stream with
>> >> event time semantics by ignoring DELETE messages or would it be
>> completed
>> >> unsupported.
>> >
>> > I don’t know the percentage of this feature in temporal scenarios.
>> >
>> > Comparing to support the approximate event time join by ignoring DELETE
>> message or by extracting an approximate event time for DELET message,  I’m
>> not sure is this acceptable for user even if we have explained the
>> limitation of approximate event time join, I tend to do not support this
>> feature, because we can not ensure the semantic of event time and it may
>> lead an incorrect result for user in some scenarios.
>> >
>> > If the percentage is highly enough and most user cases can accept the
>> approximate  event time, I'm ok to support it  for usability although it
>> doesn’t implements the event time semantic strictly.
>> >
>> > Cheers,
>> > Leonard Xu
>> >
>> >
>>
>>
>
> --
>
> Best,
> Benchao Li
>