[jira] [Created] (FLINK-1063) Race condition in NettyConnectionManager

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[jira] [Created] (FLINK-1063) Race condition in NettyConnectionManager

Shang Yuanchun (Jira)
Ufuk Celebi created FLINK-1063:
----------------------------------

             Summary: Race condition in NettyConnectionManager
                 Key: FLINK-1063
                 URL: https://issues.apache.org/jira/browse/FLINK-1063
             Project: Flink
          Issue Type: Bug
          Components: Distributed Runtime
    Affects Versions: 0.6-incubating
            Reporter: Ufuk Celebi
            Assignee: Ufuk Celebi


The TCP channel queuing mechanism in {{NettyConnectionManager}} has a race condition, which may result in re-ordering of envelopes at the receiver (the dreaded {{"Expected data packet X but received Y"}} exception).

Thanks to [AHeise|https://github.com/AHeise] for reporting the problem.

The problem has been introduced with commits 52512636444902497e47ccbfb1cabaffb3e23343 ... 32d168f439bdb5dfab02a3ab2d12e87d0622a67e.

I will revert the respective commits and implement a fall back, which limits TCP channel multiplexing and immediately closes TCP channels.



--
This message was sent by Atlassian JIRA
(v6.2#6252)