Re-enabling automatic backfilling

Armen Zambrano G.

unread,

Oct 27, 2015, 4:34:07 PM10/27/15

to

Hello sheriffs,
As we spoke in our meeting I will soon be re-enabling automatic backfilling.
I will reply to this thread when deployed.
--
Zambrano Gasparnian, Armen
Automation & Tools Engineer
http://armenzg.blogspot.ca

Armen Zambrano G.

unread,

Oct 28, 2015, 2:38:02 PM10/28/15

to mozill...@lists.mozilla.org, Sheriffs, David Burns

tl;dr please let me know by the end of week what to do when there is a
backlog on a specific repo.

Hello sheriffs,
As you know, I enabled automated backfilling this morning and disabled
it a bit later.

What happened is that philor commented on the high number of Windows
pending jobs and complained that automatic backfilling would make it worse.

At first, I thought it was unnecessary. If you look at [1] you can see
that only 25 more jobs were added by auto-backfilling over a 2 hours
period. That is really a drop in the bucket (at the time there were over
3,000 pending jobs).
After speaking with Callek, it seems that it was more about 500 jobs for
m-i. I was thinking in global pending versus per repo pending.
It seems that the inbound was closed for a bit and then re-opened once
that backlogged got better.

I want to discuss what should we do when there is a backlog of jobs.

Should we automatically close the tree when the pending count is above
certain threshold? (This removes the need for the sheriffs to keep an
eye on per platform per repo pending count).

Should automatic backfilling listen for a tree being closed?

Looking forward to discuss this with you,
Armen

On 15-10-27 04:34 PM, Armen Zambrano G. wrote:
> Hello sheriffs,
> As we spoke in our meeting I will soon be re-enabling automatic backfilling.
> I will reply to this thread when deployed.
>

[1]
Oct 28 07:31:37 fa8f6c1a_Windows 7 32-bit mozilla-inbound debug test
reftest will backfill [u'fa8f6c1aa893', u'1d67c4e4858f', u'9f16e62baa3a'].
Oct 28 07:33:38 fa8f6c1a_Windows 7 32-bit mozilla-inbound opt test
xpcshell will backfill [u'fa8f6c1aa893', u'1d67c4e4858f'].
Oct 28 07:52:11 58d4fc52_Windows 8 64-bit mozilla-inbound debug test
mochitest-other will backfill [u'58d4fc528b3b', u'2730cc97c6ec',
u'9a67e1d55e0d'].
Oct 28 08:15:12 1949b1c7_Windows 8 64-bit mozilla-inbound debug test
mochitest-devtools-chrome-5 will backfill [u'1949b1c7fccd',
u'df5f057d1277', u'113e563fd0e7'].
Oct 28 08:52:55 1949b1c7_Windows 8 64-bit mozilla-inbound debug test
mochitest-other will backfill [u'1949b1c7fccd', u'df5f057d1277',
u'113e563fd0e7', u'58d4fc528b3b', u'2730cc97c6ec', u'9a67e1d55e0d'].
Oct 28 09:17:05 692dbb16_Windows 8 64-bit mozilla-inbound opt test
mochitest-e10s-browser-chrome-3 will backfill [u'692dbb16a949',
u'c537a7eb9bbd', u'53952bbfad9a', u'1e9c356aafe6', u'b5acf46ad95b',
u'8d655f2ae624'].

Carsten Book

unread,

Oct 30, 2015, 11:44:08 AM10/30/15

to Armen Zambrano G., Sheriffs, dev-tree-...@lists.mozilla.org

Hi Armen,

yeah some kind of combination with automatic stating would be cool, would
avoid uneeded load i guess.

Cheers,

- Tomcat

On Wed, Oct 28, 2015 at 7:37 PM, Armen Zambrano G. <arm...@mozilla.com>
wrote:

> _______________________________________________
> dev-tree-management mailing list
> dev-tree-...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-tree-management
>