[jira] [Created] (FLINK-12426) TM occasionally hang in deploying state

[Shang Yuanchun (Jira)]

Qi created FLINK-12426:
--------------------------

             Summary: TM occasionally hang in deploying state
                 Key: FLINK-12426
                 URL: https://issues.apache.org/jira/browse/FLINK-12426
             Project: Flink
          Issue Type: Bug
            Reporter: Qi


Hi all,
 
We use Flink batch and start thousands of jobs per day. Occasionally we observed some stuck jobs, due to some TM hang in “DEPLOYING” state. 
 
It seems that the TM is calling BlobClient to download jars from JM/BlobServer. Under hood it’s calling Socket.connect() and then Socket.read() to retrieve results. 
 
These jobs usually have many TM slots (1~2k). We checked the TM log and dumped the TM thread. It indeed hung on socket read to download jar from Blob server. 
 
We're using Flink 1.5 but this may also affect later versions since related code are not changed much. We've tried to add socket timeout in BlobClient, but still no luck.
 
————————
TM log
————————
...
INFO org.apache.flink.runtime.taskexecutor.TaskExecutor - Received task DataSource (at createInput(ExecutionEnvironment.java:548) (our.code)) (184/2000). INFO org.apache.flink.runtime.taskmanager.Task - DataSource (at createInput(ExecutionEnvironment.java:548) (our.code)) (184/2000) switched from CREATED to DEPLOYING. INFO org.apache.flink.runtime.taskmanager.Task - Creating FileSystem stream leak safety net for task DataSource (at createInput(ExecutionEnvironment.java:548) (our.code)) (184/2000) [DEPLOYING] INFO org.apache.flink.runtime.taskmanager.Task - Loading JAR files for task DataSource (at createInput(ExecutionEnvironment.java:548) (our.code)) (184/2000) [DEPLOYING]. INFO org.apache.flink.runtime.blob.BlobClient - Downloading 19e65c0caa41f264f9ffe4ca2a48a434/p-3ecd6341bf97d5512b14c93f6c9f51f682b6db26-37d5e69d156ee00a924c1ebff0c0d280 from some-host-ip-port
{color:#222222}no more logs...{color}
 
————————
TM thread dump:
————————
_"DataSource (at createInput(ExecutionEnvironment.java:548) (our.code)) (1999/2000)" #72 prio=5 os_prio=0 tid=0x00007fb9a1521000 nid=0xa0994 runnable [0x00007fb97cfbf000]_
   _java.lang.Thread.State: RUNNABLE_
        _at java.net.SocketInputStream.socketRead0(Native Method)_
        _at java.net.SocketInputStream.socketRead(SocketInputStream.java:116)_
        _at java.net.SocketInputStream.read(SocketInputStream.java:171)_
        _at java.net.SocketInputStream.read(SocketInputStream.java:141)_
        _at org.apache.flink.runtime.blob.BlobInputStream.read(BlobInputStream.java:152)_
        _at org.apache.flink.runtime.blob.BlobInputStream.read(BlobInputStream.java:140)_
        _at org.apache.flink.runtime.blob.BlobClient.downloadFromBlobServer(BlobClient.java:170)_
        _at org.apache.flink.runtime.blob.AbstractBlobCache.getFileInternal(AbstractBlobCache.java:181)_
        _at org.apache.flink.runtime.blob.PermanentBlobCache.getFile(PermanentBlobCache.java:206)_
        _at org.apache.flink.runtime.execution.librarycache.BlobLibraryCacheManager.registerTask(BlobLibraryCacheManager.java:120)_
        _- locked <0x000000078ab60ba8> (a java.lang.Object)_
        _at org.apache.flink.runtime.taskmanager.Task.createUserCodeClassloader(Task.java:893)_
        _at org.apache.flink.runtime.taskmanager.Task.run(Task.java:585)_
        _at java.lang.Thread.run(Thread.java:748)_
_————————_
 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)