[jira] [Resolved] (FLINK-9567) Flink does not release resource in Yarn Cluster mode
[ https://issues.apache.org/jira/browse/FLINK-9567?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Till Rohrmann resolved FLINK-9567. -- Resolution: Fixed Fix Version/s: (was: 1.5.1) (was: 1.6.0) 1.5.5 1.6.2 1.7.0 Fixed via 1.7.0: 0cf776be4483e7f939de0a7f2b1fe3263a14c6fb 1.6.2: 806d3a424c0ad66f49282d0496599597d0f9f0c0 1.5.5: 69cae4f6daf6580ca83c452c7c3ad33489551a36 > Flink does not release resource in Yarn Cluster mode > > > Key: FLINK-9567 > URL: https://issues.apache.org/jira/browse/FLINK-9567 > Project: Flink > Issue Type: Bug > Components: Cluster Management, YARN >Affects Versions: 1.5.0 >Reporter: Shimin Yang >Assignee: Shimin Yang >Priority: Critical > Labels: pull-request-available > Fix For: 1.7.0, 1.6.2, 1.5.5 > > Attachments: FlinkYarnProblem, fulllog.txt > > > After restart the Job Manager in Yarn Cluster mode, sometimes Flink does not > release task manager containers in some specific case. In the worst case, I > had a job configured to 5 task managers, but possess more than 100 containers > in the end. Although the task didn't failed, but it affect other jobs in the > Yarn Cluster. > In the first log I posted, the container with id 24 is the reason why Yarn > did not release resources. As the container was killed before restart, but it > has not received the callback of *onContainerComplete* in > *YarnResourceManager* which should be called by *AMRMAsyncClient* of Yarn. > After restart, as we can see in line 347 of FlinkYarnProblem log, > 2018-06-14 22:50:47,846 WARN akka.remote.ReliableDeliverySupervisor - > Association with remote system [akka.tcp://flink@bd-r1hdp69:30609] has > failed, address is now gated for [50] ms. Reason: [Disassociated] > Flink lost the connection of container 24 which is on bd-r1hdp69 machine. > When it try to call *closeTaskManagerConnection* in *onContainerComplete*, it > did not has the connection to TaskManager on container 24, so it just ignore > the close of TaskManger. > 2018-06-14 22:50:51,812 DEBUG org.apache.flink.yarn.YarnResourceManager - No > open TaskExecutor connection container_1528707394163_29461_02_24. > Ignoring close TaskExecutor connection. > However, bafore calling *closeTaskManagerConnection,* it already called > *requestYarnContainer* which lead to *numPendingContainerRequests variable > in* *YarnResourceManager* increased by 1. > As the excessive container return is determined by the > *numPendingContainerRequests* variable in *YarnResourceManager*, it cannot > return this container although it is not required. Meanwhile, the restart > logic has already allocated enough containers for Task Managers, Flink will > possess the extra container for a long time for nothing. > In the full log, the job ended with 7 containers while only 3 are running > TaskManagers. > ps: Another strange thing I found is that when sometimes request for a yarn > container, it will return much more than requested. Is it a normal scenario > for AMRMAsyncClient? -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Resolved] (FLINK-9567) Flink does not release resource in Yarn Cluster mode
[ https://issues.apache.org/jira/browse/FLINK-9567?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Till Rohrmann resolved FLINK-9567. -- Resolution: Fixed Fix Version/s: 1.5.1 Fixed via 1.6.0: 49156e8d6c2cb223e350760ad69d1898d67dacab 1.5.1: 4b0aca5681e749cc8b0389d389e1fac736653eb5 > Flink does not release resource in Yarn Cluster mode > > > Key: FLINK-9567 > URL: https://issues.apache.org/jira/browse/FLINK-9567 > Project: Flink > Issue Type: Bug > Components: Cluster Management, YARN >Affects Versions: 1.5.0 >Reporter: Shimin Yang >Assignee: Shimin Yang >Priority: Critical > Labels: pull-request-available > Fix For: 1.6.0, 1.5.1 > > Attachments: FlinkYarnProblem, fulllog.txt > > > After restart the Job Manager in Yarn Cluster mode, sometimes Flink does not > release task manager containers in some specific case. In the worst case, I > had a job configured to 5 task managers, but possess more than 100 containers > in the end. Although the task didn't failed, but it affect other jobs in the > Yarn Cluster. > In the first log I posted, the container with id 24 is the reason why Yarn > did not release resources. As the container was killed before restart, but it > has not received the callback of *onContainerComplete* in > *YarnResourceManager* which should be called by *AMRMAsyncClient* of Yarn. > After restart, as we can see in line 347 of FlinkYarnProblem log, > 2018-06-14 22:50:47,846 WARN akka.remote.ReliableDeliverySupervisor - > Association with remote system [akka.tcp://flink@bd-r1hdp69:30609] has > failed, address is now gated for [50] ms. Reason: [Disassociated] > Flink lost the connection of container 24 which is on bd-r1hdp69 machine. > When it try to call *closeTaskManagerConnection* in *onContainerComplete*, it > did not has the connection to TaskManager on container 24, so it just ignore > the close of TaskManger. > 2018-06-14 22:50:51,812 DEBUG org.apache.flink.yarn.YarnResourceManager - No > open TaskExecutor connection container_1528707394163_29461_02_24. > Ignoring close TaskExecutor connection. > However, bafore calling *closeTaskManagerConnection,* it already called > *requestYarnContainer* which lead to *numPendingContainerRequests variable > in* *YarnResourceManager* increased by 1. > As the excessive container return is determined by the > *numPendingContainerRequests* variable in *YarnResourceManager*, it cannot > return this container although it is not required. Meanwhile, the restart > logic has already allocated enough containers for Task Managers, Flink will > possess the extra container for a long time for nothing. > In the full log, the job ended with 7 containers while only 3 are running > TaskManagers. > ps: Another strange thing I found is that when sometimes request for a yarn > container, it will return much more than requested. Is it a normal scenario > for AMRMAsyncClient? -- This message was sent by Atlassian JIRA (v7.6.3#76005)
