You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "ASF GitHub Bot (JIRA)" <ji...@apache.org> on 2017/04/25 17:59:04 UTC
[jira] [Commented] (FLINK-5867) The implementation of
RestartPipelinedRegionStrategy
[ https://issues.apache.org/jira/browse/FLINK-5867?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15983350#comment-15983350 ]
ASF GitHub Bot commented on FLINK-5867:
---------------------------------------
GitHub user StephanEwen opened a pull request:
https://github.com/apache/flink/pull/3773
[FLINK-5867] [FLINK-5866] [flip-1] Implement FailoverStrategy for pipelined regions
This is based on #3772 , the relevant commits are the latter four.
The majority of the work has been done by @tiemsn , with some rebasing and additions from me.
# Pipelined Region Failover
As described in [FLIP-1](https://cwiki.apache.org/confluence/display/FLINK/FLIP-1+%3A+Fine+Grained+Recovery+from+Task+Failures), this pull request implements the detection of pipelined regions in the `ExecutionGraph` and failover within these pipelined regions.
![st0-nzqia5abpwrgaogpllw](https://cloud.githubusercontent.com/assets/1727146/25399938/54fda5a4-29f1-11e7-9efe-5d845644089f.png)
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/StephanEwen/incubator-flink flip-1-pipelined-regions
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/flink/pull/3773.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 #3773
----
commit ef7fd9964c1c74feb4641e57a138c54558b2449c
Author: Stephan Ewen <se...@apache.org>
Date: 2017-03-21T18:13:34Z
[FLINK-5869] [flip-1] Add basic abstraction for Failover Strategies to ExecutionGraph
- Rename 'ExecutionGraph.fail()' to 'ExecutionGraph.failGlobally()' to differentiate from fine grained failures/recovery
- Add base class for FailoverStrategy
- Add default implementation (restart all tasks)
- Add logic to load the failover strategy from the configuration
commit c04a8a312098fddce14e392b8d9dbf396b1df3f3
Author: Stephan Ewen <se...@apache.org>
Date: 2017-03-29T20:49:54Z
[FLINK-6340] [flip-1] Add a termination future to the Execution
commit 92d3f7e1025dc3c3499730bda8e8a9acfd3b5c13
Author: shuai.xus <sh...@alibaba-inc.com>
Date: 2017-04-18T06:15:29Z
[FLINK-5867] [flip-1] Support restarting only pipelined sub-regions of the ExecutionGraph on task failure
commit 456600d5e37724bbcc7d570f6828e3fef6298483
Author: shuai.xus <sh...@alibaba-inc.com>
Date: 2017-04-20T21:56:53Z
[FLINK-5867] [flip-1] Add tests for pipelined failover region construction
commit 622f07e0efc82bf13f12ae1960a35ecc48c865c1
Author: Stephan Ewen <se...@apache.org>
Date: 2017-04-20T22:02:19Z
[FLINK-5867] [flip-1] Improve performance of Pipelined Failover Region construction
This method exploits the fact that verties are already in topological order.
commit 39402583df8b4c51016c72f968772cbbdd6c92e3
Author: shuai.xus <sh...@alibaba-inc.com>
Date: 2017-04-25T07:42:48Z
[FLINK-5867] [flip-1] Correct some JavaDocs for RestartIndividualStrategy
----
> The implementation of RestartPipelinedRegionStrategy
> ----------------------------------------------------
>
> Key: FLINK-5867
> URL: https://issues.apache.org/jira/browse/FLINK-5867
> Project: Flink
> Issue Type: Sub-task
> Components: JobManager
> Reporter: shuai.xu
> Assignee: shuai.xu
>
> The RestartPipelinedRegionStrategy's responsibility is the following:
> 1. Calculate all FailoverRegions and their relations when initializing.
> 2. Listen for the failure of the job and executions, and find corresponding FailoverRegions to do the failover.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)