[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15708503#comment-15708503 ] ASF GitHub Bot commented on FLINK-5197: --- Github user tillrohrmann closed the pull request at: https://github.com/apache/flink/pull/2895 > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15708500#comment-15708500 ] ASF GitHub Bot commented on FLINK-5197: --- Github user tillrohrmann closed the pull request at: https://github.com/apache/flink/pull/2896 > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15708502#comment-15708502 ] ASF GitHub Bot commented on FLINK-5197: --- Github user tillrohrmann commented on the issue: https://github.com/apache/flink/pull/2895 Forgot to include the closing tag in the commit. > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15708494#comment-15708494 ] ASF GitHub Bot commented on FLINK-5197: --- Github user tillrohrmann commented on the issue: https://github.com/apache/flink/pull/2895 Merging this PR. @uce reviewed the backport. > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15708485#comment-15708485 ] ASF GitHub Bot commented on FLINK-5197: --- Github user tillrohrmann commented on the issue: https://github.com/apache/flink/pull/2896 Thanks for the review @uce. Failing test cases are unrelated. Merging this PR. > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15705885#comment-15705885 ] ASF GitHub Bot commented on FLINK-5197: --- Github user uce commented on the issue: https://github.com/apache/flink/pull/2896 Good catch! The `RemoveJob` cannot succeed since is also checking the `currentJobs` that are checked for `JobStatusChanged` already. So in the end, the only case where this actually triggers removal is when it interfers with a recovered job as you say. +1 to merge for 1.1 and #2895 for 1.2. > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15705540#comment-15705540 ] ASF GitHub Bot commented on FLINK-5197: --- GitHub user tillrohrmann opened a pull request: https://github.com/apache/flink/pull/2896 [FLINK-5197] [jm] Ignore outdated JobStatusChanged messages Backport of #2895 for release 1.1 branch. Outdated JobStatusChanged messages no longer trigger a RemoveJob message but are logged and ignored. This has the advantage, that an outdated JobStatusChanged message cannot interfere with a recovered job which can have the same job id. Review @uce. You can merge this pull request into a Git repository by running: $ git pull https://github.com/tillrohrmann/flink backportFixJobStatusChangedMessage Alternatively you can review and apply these changes as the patch at: https://github.com/apache/flink/pull/2896.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #2896 commit 4a2f948224fe628c721adc4fae24199b0296c80f Author: Till RohrmannDate: 2016-11-29T15:02:29Z [FLINK-5197] [jm] Ignore outdated JobStatusChanged messages Outdated JobStatusChanged messages no longer trigger a RemoveJob message but are logged and ignored. This has the advantage, that an outdated JobStatusChanged message cannot interfere with a recovered job which can have the same job id. > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (FLINK-5197) Late JobStatusChanged messages can interfere with running jobs
[ https://issues.apache.org/jira/browse/FLINK-5197?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15705538#comment-15705538 ] ASF GitHub Bot commented on FLINK-5197: --- GitHub user tillrohrmann opened a pull request: https://github.com/apache/flink/pull/2895 [FLINK-5197] [jm] Ignore outdated JobStatusChanged messages Outdated JobStatusChanged messages no longer trigger a RemoveJob message but are logged and ignored. This has the advantage, that an outdated JobStatusChanged message cannot interfere with a recovered job which can have the same job id. You can merge this pull request into a Git repository by running: $ git pull https://github.com/tillrohrmann/flink fixJobStatusChangedMessage Alternatively you can review and apply these changes as the patch at: https://github.com/apache/flink/pull/2895.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #2895 commit 490cef46380178a2296c2f743b9eb91154967463 Author: Till RohrmannDate: 2016-11-29T15:02:29Z [FLINK-5197] [jm] Ignore outdated JobStatusChanged messages Outdated JobStatusChanged messages no longer trigger a RemoveJob message but are logged and ignored. This has the advantage, that an outdated JobStatusChanged message cannot interfere with a recovered job which can have the same job id. > Late JobStatusChanged messages can interfere with running jobs > -- > > Key: FLINK-5197 > URL: https://issues.apache.org/jira/browse/FLINK-5197 > Project: Flink > Issue Type: Bug > Components: JobManager >Affects Versions: 1.2.0, 1.1.3 >Reporter: Till Rohrmann >Assignee: Till Rohrmann >Priority: Minor > Fix For: 1.2.0, 1.1.4 > > > When the {{JobManager}} receives a {{JobStatusChanged}} message, it will look > up the {{ExecutionGraph}} for the given {{JobID}}. If there is no > {{ExecutionGraph}}, then a {{RemoveJob}} message is sent to itself. In the > general case, this is not problematic, because the {{RemoveJob}} message > won't do anything if there is no {{ExecutionGraph}}. However, since this is > an asynchronous call, it can happen that the corresponding job of the > {{JobID}} is recovered before receiving the {{RemoveJob}} message. In this > case, the newly recovered job would be removed. > I propose to change the behaviour such that a {{JobStatusChanged}} for a > non-existing {{ExecutionGraph}} will be simply ignored. -- This message was sent by Atlassian JIRA (v6.3.4#6332)