[JIRA] (OVIRT-620) Optimize how Jenkins fetchs ovirt-engine from gerrit

eyal edri [Administrator] (oVirt JIRA) jira at ovirt-jira.atlassian.net
Tue Jul 5 16:57:00 UTC 2016


    [ https://ovirt-jira.atlassian.net/browse/OVIRT-620?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18004#comment-18004 ] 

eyal edri [Administrator] commented on OVIRT-620:
-------------------------------------------------

I would say lets give it a try on the ovirt engine jobs
On Jul 5, 2016 3:15 PM, "Nadav Goldin (oVirt JIRA)" <



> Optimize how Jenkins fetchs ovirt-engine from gerrit
> ----------------------------------------------------
>
>                 Key: OVIRT-620
>                 URL: https://ovirt-jira.atlassian.net/browse/OVIRT-620
>             Project: oVirt - virtualization made easy
>          Issue Type: Sub-task
>            Reporter: Nadav Goldin
>            Assignee: infra
>
> ovirt-engine git repository is at the time of writing is of size 244MB, where ovirt-engine/.git alone takes 160MB.
> There are two things which might be done:
> 1. Find a way to optimize ovirt-engine.git/.git size, this needs to be explored deeply.
> 2. Check if it is possible to configure git plugin to not download the entire history, although Jenkins attempts to optimize that, it still looks like the git plugin is downloading the entire history(this is what a random check on the slaves using du ../jenkins/workspace/...../.git -h -s showed)
> for instance the following command takes only 91Mb:
> {quote}git clone --depth 1 --branch master {quote}
> Although this will not solve real network/storage issues, it could help reduce those problems and in the overall decrease jobs runtime. As Gerrit is still on AWS, this is a considerable amount of traffic for each job we might be able to reduce.



--
This message was sent by Atlassian JIRA
(v1000.126.1#100004)



More information about the Infra mailing list