[
https://issues.apache.org/jira/browse/FLINK-9196?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16457508#comment-16457508
]
ASF GitHub Bot commented on FLINK-9196:
---------------------------------------
GitHub user GJL opened a pull request:
https://github.com/apache/flink/pull/5938
[FLINK-9196][flip6, yarn] Cleanup application files when deregistering YARN
AM
## What is the purpose of the change
*Ensure that YARN application files are removed if cluster is shutdown.*
cc: @StephanEwen @tillrohrmann
## Brief change log
- *Enable graceful cluster shut down via HTTP.*
- *Remove Flink application files from remote file system when the
YarnResourceManager deregisters the YARN ApplicationMaster.
## Verifying this change
This change added tests and can be verified as follows:
- *Manually verified that files are removed from HDFS when running stream
(attached/detached) and batch jobs (attached).*
- *Manually verified that files are removed from HDFS when running
stopping a yarn session gracefully.*
## Does this pull request potentially affect one of the following parts:
- Dependencies (does it add or upgrade a dependency): (yes / **no**)
- The public API, i.e., is any changed class annotated with
`@Public(Evolving)`: (yes / **no**)
- The serializers: (yes / **no** / don't know)
- The runtime per-record code paths (performance sensitive): (yes /
**no** / don't know)
- Anything that affects deployment or recovery: JobManager (and its
components), Checkpointing, Yarn/Mesos, ZooKeeper: (**yes** / no / don't know)
- The S3 file system connector: (yes / **no** / don't know)
## Documentation
- Does this pull request introduce a new feature? (yes / **no**)
- If yes, how is the feature documented? (**not applicable** / docs /
JavaDocs / not documented)
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/GJL/flink FLINK-9196
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/flink/pull/5938.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 #5938
----
commit 6f0c0aed8a5b54814ed2e0fa761f06317592e4b3
Author: gyao <gary@...>
Date: 2018-04-19T08:29:43Z
[hotfix] Replace String concatenation with Slf4j placeholders.
commit 34b5b40fec62502579a3f3804839c1e9d1e95952
Author: gyao <gary@...>
Date: 2018-04-19T09:03:20Z
[hotfix] Indent method parameters.
commit bcb0f24ec587c15287c6144d1c088a5327d98c6d
Author: gyao <gary@...>
Date: 2018-04-19T09:04:27Z
[hotfix] Remove unnecessary int cast.
commit 264b3e664fe84583ab8e372824f6d4424627e6e1
Author: gyao <gary@...>
Date: 2018-04-19T09:05:05Z
[hotfix] Fix raw types warning.
commit 1b6eb96b3d287a20ea86606fd01b5e10564c3f5d
Author: gyao <gary@...>
Date: 2018-04-19T09:18:32Z
[hotfix][tests] Rename UtilsTest to YarnFlinkResourceManagerTest.
Test was misnamed.
commit e8d43ff72a2861713db934fe42163fac6d9ecb8d
Author: gyao <gary@...>
Date: 2018-04-26T15:38:20Z
[hotfix][mesos] Delete unused class FlinkMesosSessionCli.
commit a4f9a5c6a44f08aa5f4a8dbbfb28a0bdb562b8c5
Author: gyao <gary@...>
Date: 2018-04-26T15:44:56Z
[hotfix][yarn] Remove unused field appReport in YarnClusterClient.
commit 1260dfac974670f325b21d175e1e29064530bb53
Author: gyao <gary@...>
Date: 2018-04-19T10:07:54Z
[FLINK-9196][flip6, yarn] Cleanup application files when deregistering YARN
AM
Enable graceful cluster shut down via HTTP.
Remove Flink application files from remote file system when the
YarnResourceManager deregisters the YARN ApplicationMaster.
----
> YARN: Flink binaries are not deleted from HDFS after cluster shutdown
> ---------------------------------------------------------------------
>
> Key: FLINK-9196
> URL: https://issues.apache.org/jira/browse/FLINK-9196
> Project: Flink
> Issue Type: Bug
> Components: YARN
> Affects Versions: 1.5.0
> Reporter: Gary Yao
> Assignee: Gary Yao
> Priority: Blocker
> Labels: flip-6
> Fix For: 1.5.0
>
> Attachments: 0001-xxx.patch
>
>
> When deploying on YARN in flip6 mode, the Flink binaries are not deleted from
> HDFS after the cluster shuts down.
> *Steps to reproduce*
> # Submit job in YARN job mode, non-detached:
> {noformat} HADOOP_CLASSPATH=`hadoop classpath` bin/flink run -m yarn-cluster
> -yjm 2048 -ytm 2048 ./examples/streaming/WordCount.jar {noformat}
> # Check contents of {{/user/hadoop/.flink/<application_id>}} on HDFS after
> job is finished:
> {noformat}
> [hadoop@ip-172-31-43-78 flink-1.5.0]$ hdfs dfs -ls
> /user/hadoop/.flink/application_1523966184826_0016
> Found 6 items
> -rw-r--r-- 1 hadoop hadoop 583 2018-04-17 14:54
> /user/hadoop/.flink/application_1523966184826_0016/90cf5b3a-039e-4d52-8266-4e9563d74827-taskmanager-conf.yaml
> -rw-r--r-- 1 hadoop hadoop 332 2018-04-17 14:54
> /user/hadoop/.flink/application_1523966184826_0016/application_1523966184826_0016-flink-conf.yaml3818971235442577934.tmp
> -rw-r--r-- 1 hadoop hadoop 89779342 2018-04-02 17:08
> /user/hadoop/.flink/application_1523966184826_0016/flink-dist_2.11-1.5.0.jar
> drwxrwxrwx - hadoop hadoop 0 2018-04-17 14:54
> /user/hadoop/.flink/application_1523966184826_0016/lib
> -rw-r--r-- 1 hadoop hadoop 1939 2018-04-02 15:37
> /user/hadoop/.flink/application_1523966184826_0016/log4j.properties
> -rw-r--r-- 1 hadoop hadoop 2331 2018-04-02 15:37
> /user/hadoop/.flink/application_1523966184826_0016/logback.xml
> {noformat}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
