infra-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Allen Wittenauer (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (INFRA-15920) docker purge is too aggressive on the Hadoop nodes
Date Thu, 01 Feb 2018 18:02:00 GMT

    [ https://issues.apache.org/jira/browse/INFRA-15920?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16349014#comment-16349014
] 

Allen Wittenauer commented on INFRA-15920:
------------------------------------------

Another build broken by INFRA's code:

https://builds.apache.org/view/H-L/view/HBase/job/PreCommit-HBASE-Build/11330/console

> docker purge is too aggressive on the Hadoop nodes
> --------------------------------------------------
>
>                 Key: INFRA-15920
>                 URL: https://issues.apache.org/jira/browse/INFRA-15920
>             Project: Infrastructure
>          Issue Type: Improvement
>          Components: Jenkins
>            Reporter: Allen Wittenauer
>            Priority: Major
>
> On the Hadoop machines, the cron-based 'docker purge' is too aggressive.   We are starting
to see 'docker build' commands due to a race condition caused by the cron job.  
> e.g.:
> https://builds.apache.org/job/PreCommit-HADOOP-Build/14039/console
> The message:
> unknown parent image ID sha256:a67d5d1330f42e610a78f1cc35f2f50dfae116b2fa458c0d323c5ad5db8a1247
> correlates to an earlier message:
>  ---> a67d5d1330f4
> Ideally, the purge on these machines would just be turned off given that Yetus has been
keeping them clean for years now.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message