On Wed, Apr 4, 2018 at 12:10 PM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:


2018-03-21 11:01 GMT+01:00 Eyal Edri <eedri@redhat.com>:


On Wed, Mar 21, 2018 at 11:56 AM, Yedidyah Bar David <didi@redhat.com> wrote:
On Wed, Mar 21, 2018 at 11:48 AM, Code Review <gerrit@ovirt.org> wrote:
> From Jenkins CI:
>
> Jenkins CI has posted comments on this change.
>
> Change subject: Test hosted-engine cleanup
> ......................................................................
>
>
> Patch Set 19:
>
> Build Failed
>
> http://jenkins.ovirt.org/job/ovirt-system-tests_master_check-patch-el7-x86_64/4616/ : FAILURE

This is a retrigger of my previous mail to infra. Failed with:

09:48:27 CLEANUP: Stop all running containers and remove unwanted images
...

09:48:37 Error response from daemon: conflict: unable to delete
782fe38c8050 (cannot be forced) - image has dependent child images
09:48:37 Error response from daemon: conflict: unable to delete
bcec0ae8107e (cannot be forced) - image has dependent child images
09:48:37 ERROR: Failed to clean docker images

I have already seen this before.

Perhaps we should automatically reprovision all slaves periodically?
They should be cattle, not pets, right? :-)
I already suggested this in the past and there was some objection...

Unfourtunately oVIrt knows more how to handle pets than cattle :) 

Can't we use a stateless pool for this scope?

We tried it in downstream for some time, it didn't work well and we no longer use it.
Barak wrote the flow, he can share more info if you want.
 
 
This topic has indeed been discussed numerous times, since we use mock, we don't need to reprovision slaves, but there are some cleanups being done that might fail occasionally.
Our plan is to move away from mock to containers, so hopefully jobs will run on clean containers running on OpenShift, so I don't think its worth investing too much in the current solution over other priorities,

I don't recall this plan has been discussed, can you link to the thread about this?

I'm not sure if there was a thread on it, but there is a long-standing feature in JIRA [1], there are not immediate resources or plan for starting it though, we were hoping to get cooperation also from other CI teams 
on that front, but doesn't seem the case so far.


[1] https://ovirt-jira.atlassian.net/browse/OVIRT-873
 

 
Especially since it doesn't fail often, and each time it fails we improve the cleanup to handle the issue. 
 

>
> --
> To view, visit https://gerrit.ovirt.org/88331
> To unsubscribe, visit https://gerrit.ovirt.org/settings
>
> Gerrit-Project: ovirt-system-tests
> Gerrit-Branch: master
> Gerrit-MessageType: comment
> Gerrit-Change-Id: I5876a419dc2437de51149b1fe02ffd900e6ab29e
> Gerrit-Change-Number: 88331
> Gerrit-PatchSet: 19
> Gerrit-Owner: Yedidyah Bar David <didi@redhat.com>
> Gerrit-Reviewer: Jenkins CI
> Gerrit-Reviewer: Sandro Bonazzola <sbonazzo@redhat.com>
> Gerrit-Reviewer: Yedidyah Bar David <didi@redhat.com>
> Gerrit-Comment-Date: Wed, 21 Mar 2018 09:48:41 +0000
> Gerrit-HasComments: No



--
Didi
_______________________________________________
Infra mailing list
Infra@ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra



--

Eyal edri


MANAGER

RHV DevOps

EMEA VIRTUALIZATION R&D


Red Hat EMEA

TRIED. TESTED. TRUSTED.
phone: +972-9-7692018
irc: eedri (on #tlv #rhev-dev #rhev-integ)

_______________________________________________
Infra mailing list
Infra@ovirt.org
http://lists.ovirt.org/mailman/listinfo/infra




--

SANDRO BONAZZOLA

ASSOCIATE MANAGER, SOFTWARE ENGINEERING, EMEA ENG VIRTUALIZATION R&D

Red Hat EMEA

sbonazzo@redhat.com   




--

Eyal edri


MANAGER

RHV DevOps

EMEA VIRTUALIZATION R&D


Red Hat EMEA

TRIED. TESTED. TRUSTED.
phone: +972-9-7692018
irc: eedri (on #tlv #rhev-dev #rhev-integ)