[
https://issues.apache.org/jira/browse/HADOOP-18007?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Gautham Banasandra updated HADOOP-18007:
----------------------------------------
Description:
Consider the Github comment by hadoop-yetus for different platform runs -
||Platform||hadoop-yetus report||
|Centos 7|https://github.com/apache/hadoop/pull/3563#issuecomment-947151110|
|Centos 8|https://github.com/apache/hadoop/pull/3563#issuecomment-947239385|
|Debian 10|https://github.com/apache/hadoop/pull/3563#issuecomment-947340406|
|Ubuntu Focal|https://github.com/apache/hadoop/pull/3563#issuecomment-947464004|
Notice that the Docker subsystem points to the same file for all the platforms -
!image-2021-11-12-21-13-18-997.png!
To illustrate the issue here, upon clicking on the link to Dockerfile in
[Centos 7's hadoop-yetus
report|https://github.com/apache/hadoop/pull/3563#issuecomment-947151110] will
navigate to Ubuntu Focal's Dockerfile (since Ubuntu Focal wass the most
recently run platform for that pipeline run).
Please note that this issue applies to all the links that contain "out/" that
appear in the Yetus summary.
As a side note, this issue doesn't happen when the Yetus run fails. Consider
https://github.com/apache/hadoop/pull/3563#issuecomment-946215556. Here the
Yetus run has failed for Centos 7. The link to Dockerfile points to Centos 7's
Dockerfile since Centos 7 is the last platform that was run for that pipeline.
To fix this issue, we need to do the following two things -
*+Archive the artifacts even when the Yetus run is successful+*
Currently, we're archiving the artifacts only when the run fails -
https://github.com/apache/hadoop/blob/bccf2f3ef4c8f09f010656f9061a4e323daf132b/dev-support/Jenkinsfile#L142-L148
(Hence, we're able to navigate to the right Dockerfile when the Yetus run
fails for a platform). We need to archive the artifacts even when the run
succeeds so that we're able to navigate to it irrespective of success/failure
of the Yetus run.
*+Expose different endpoints for each platform+*
At the end of the Yetus's pre-commit run, we see the following table -
|| Subsystem || Report/Notes ||
| cc |
https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-3650/1/*artifact/out*/results-compile-cc-hadoop-yarn-project_hadoop-yarn_hadoop-yarn-server-jdkUbuntu-11.0.11+9-Ubuntu-0ubuntu2.20.04.txt
|
| unit |
https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-3650/1/*artifact/out*/patch-unit-hadoop-yarn-project_hadoop-yarn_hadoop-yarn-server_hadoop-yarn-server-resourcemanager.txt
|
We need a way to change the URL to /artifact/*centos-7*/out/,
/artifact/*ubuntu-focal*/out/ and so on.
was:
Consider the Github comment by hadoop-yetus for different platform runs -
||Platform||hadoop-yetus report||
|Centos 7|https://github.com/apache/hadoop/pull/3563#issuecomment-947151110|
|Centos 8|https://github.com/apache/hadoop/pull/3563#issuecomment-947239385|
|Debian 10|https://github.com/apache/hadoop/pull/3563#issuecomment-947340406|
|Ubuntu Focal|https://github.com/apache/hadoop/pull/3563#issuecomment-947464004|
Notice that the Docker subsystem points to the same file for all the platforms -
!image-2021-11-12-21-13-18-997.png!
To illustrate the issue here, upon clicking on the link to Dockerfile in
[Centos 7's hadoop-yetus
report|https://github.com/apache/hadoop/pull/3563#issuecomment-947151110] will
navigate to Ubuntu Focal's Dockerfile (since Ubuntu Focal wass the most
recently run platform for that pipeline run).
Please note that this issue applies to all the links that contain "out/" that
appear in the Yetus summary.
As a side note, this issue doesn't happen when the Yetus run fails. Consider
https://github.com/apache/hadoop/pull/3563#issuecomment-946215556. Here the
Yetus run has failed for Centos 7. The link to Dockerfile points to Centos 7's
Dockerfile since Centos 7 is the last platform that was run for that pipeline.
To fix this issue, we need to do 2 things -
*+Archive the artifacts even when the Yetus run is successful+*
Currently, we're archiving the artifacts only when the run fails -
https://github.com/apache/hadoop/blob/bccf2f3ef4c8f09f010656f9061a4e323daf132b/dev-support/Jenkinsfile#L142-L148
(Hence, we're able to navigate to the right Dockerfile when the Yetus run
fails for a platform). We need to archive the artifacts even when the run
succeeds so that we're able to navigate to it irrespective of success/failure
of the Yetus run.
*+Expose different endpoints for each platform+*
At the end of the Yetus's pre-commit run, we see the following table -
|| Subsystem || Report/Notes ||
| cc |
https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-3650/1/*artifact/out*/results-compile-cc-hadoop-yarn-project_hadoop-yarn_hadoop-yarn-server-jdkUbuntu-11.0.11+9-Ubuntu-0ubuntu2.20.04.txt
|
| unit |
https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-3650/1/*artifact/out*/patch-unit-hadoop-yarn-project_hadoop-yarn_hadoop-yarn-server_hadoop-yarn-server-resourcemanager.txt
|
We need a way to change the URL to /artifact/*centos-7*/out/,
/artifact/*ubuntu-focal*/out/ and so on.
> Use platform specific endpoints for CI report
> ---------------------------------------------
>
> Key: HADOOP-18007
> URL: https://issues.apache.org/jira/browse/HADOOP-18007
> Project: Hadoop Common
> Issue Type: Bug
> Components: build
> Affects Versions: 3.4.0
> Environment: Centos 7, Centos 8, Debian 10, Ubuntu Focal
> Reporter: Gautham Banasandra
> Assignee: Gautham Banasandra
> Priority: Major
> Attachments: image-2021-11-12-21-13-18-997.png
>
>
> Consider the Github comment by hadoop-yetus for different platform runs -
> ||Platform||hadoop-yetus report||
> |Centos 7|https://github.com/apache/hadoop/pull/3563#issuecomment-947151110|
> |Centos 8|https://github.com/apache/hadoop/pull/3563#issuecomment-947239385|
> |Debian 10|https://github.com/apache/hadoop/pull/3563#issuecomment-947340406|
> |Ubuntu
> Focal|https://github.com/apache/hadoop/pull/3563#issuecomment-947464004|
>
> Notice that the Docker subsystem points to the same file for all the
> platforms -
> !image-2021-11-12-21-13-18-997.png!
> To illustrate the issue here, upon clicking on the link to Dockerfile in
> [Centos 7's hadoop-yetus
> report|https://github.com/apache/hadoop/pull/3563#issuecomment-947151110]
> will navigate to Ubuntu Focal's Dockerfile (since Ubuntu Focal wass the most
> recently run platform for that pipeline run).
> Please note that this issue applies to all the links that contain "out/" that
> appear in the Yetus summary.
> As a side note, this issue doesn't happen when the Yetus run fails. Consider
> https://github.com/apache/hadoop/pull/3563#issuecomment-946215556. Here the
> Yetus run has failed for Centos 7. The link to Dockerfile points to Centos
> 7's Dockerfile since Centos 7 is the last platform that was run for that
> pipeline.
> To fix this issue, we need to do the following two things -
> *+Archive the artifacts even when the Yetus run is successful+*
> Currently, we're archiving the artifacts only when the run fails -
> https://github.com/apache/hadoop/blob/bccf2f3ef4c8f09f010656f9061a4e323daf132b/dev-support/Jenkinsfile#L142-L148
> (Hence, we're able to navigate to the right Dockerfile when the Yetus run
> fails for a platform). We need to archive the artifacts even when the run
> succeeds so that we're able to navigate to it irrespective of success/failure
> of the Yetus run.
> *+Expose different endpoints for each platform+*
> At the end of the Yetus's pre-commit run, we see the following table -
> || Subsystem || Report/Notes ||
> | cc |
> https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-3650/1/*artifact/out*/results-compile-cc-hadoop-yarn-project_hadoop-yarn_hadoop-yarn-server-jdkUbuntu-11.0.11+9-Ubuntu-0ubuntu2.20.04.txt
> |
> | unit |
> https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-3650/1/*artifact/out*/patch-unit-hadoop-yarn-project_hadoop-yarn_hadoop-yarn-server_hadoop-yarn-server-resourcemanager.txt
> |
> We need a way to change the URL to /artifact/*centos-7*/out/,
> /artifact/*ubuntu-focal*/out/ and so on.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
