1186617 - releng changes to decommission pandas

Status: RESOLVED FIXED

(Infrastructure & Operations Graveyard :: CIDuty, task)

Description

[Kim Moir [:kmoir] ET]

remove platform buildbot-configs, buildbotcustom, devices.json in tools, slave health etc.

Comment 2

[Kim Moir [:kmoir] ET]

actually I will use this bug to remove dead panda code

Comment 3

[Kim Moir [:kmoir] ET]

Attachment: bug1186617tools.patch

patches to remove references to pandas, foopies, bmm and mozpool from tools repo

Comment 4

[Kim Moir [:kmoir] ET]

Attachment: bug1186617puppet.patch

Comment 5

[Kim Moir [:kmoir] ET]

Attachment: bug1186617bbcustom.patch

Comment 6

[Kim Moir [:kmoir] ET]

Attachment: bug1186617bbcustom.patch

Comment 7

[Kim Moir [:kmoir] ET]

Attachment: bug1186617bbcustom.patch

Comment 8

[Kim Moir [:kmoir] ET]

Attachment: bug1186617sh.patch

Comment 9

[Kim Moir [:kmoir] ET]

Attachment: bug1186617mh.patch

remove mh scripts on inbound for panda and panda talos

Comment 10

[Joel Maher ( :jmaher ) (UTC -8)]

Comment on attachment 8729155 [details] [diff] [review]
bug1186617mh.patch

Review of attachment 8729155 [details] [diff] [review]:
-----------------------------------------------------------------

so beautiful!

Comment 11

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/d83f22ef3d5c

Comment 12

[Kim Moir [:kmoir] ET]

I removed the pandas and bm89 and 100-102 from the slavealloc databases too

Comment 13

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/d83f22ef3d5c

Comment 14

[Chris Cooper [:coop] (he/him)]

Comment on attachment 8729131 [details] [diff] [review]
bug1186617sh.patch

Review of attachment 8729131 [details] [diff] [review]:
-----------------------------------------------------------------

We'll want to watch when this lands that it doesn't break on historical panda data in the db. Problems should appear as unknown entries on the index page.

Comment 15

[Release Engineering SlaveAPI Service]

In production: https://hg.mozilla.org/build/buildbotcustom/rev/695299094ba0

Comment 16

[Chris Cooper [:coop] (he/him)]

Attachment: [tools] Remove foopies from end-to-end reconfig script

Streamlining our reconfig process a bit now that foopies are no more.

Comment 17

[Kim Moir [:kmoir] ET]

Comment on attachment 8729623 [details] [diff] [review]
[tools] Remove foopies from end-to-end reconfig script

Thanks!

I actually included this in https://bug1186617.bmoattachments.org/attachment.cgi?id=8729107 yesterday but didn't ask for review yet

Comment 18

[Chris Cooper [:coop] (he/him)]

Comment on attachment 8729623 [details] [diff] [review]
[tools] Remove foopies from end-to-end reconfig script

(In reply to Kim Moir [:kmoir] from comment #17) 
> I actually included this in
> https://bug1186617.bmoattachments.org/attachment.cgi?id=8729107 yesterday
> but didn't ask for review yet

OK, let's go with your more complete patch. Let me know when you want review.

Comment 19

[Chris Cooper [:coop] (he/him)]

Comment on attachment 8729107 [details] [diff] [review]
bug1186617tools.patch

Review of attachment 8729107 [details] [diff] [review]:
-----------------------------------------------------------------

Can I get you to tag the tools repo *before* you land this patch with PANDA_FOOPY_SUTLIB (or something better, naming things is hard) so we can easily find this code again in the future if we need to?

Nit: you have trailing whitespace on some of your patched files. Might want to add something like https://github.com/ccooper/dotfiles/blob/master/.vimrc#L56 to your .vimrc file.

Comment 20

[Kim Moir [:kmoir] ET]

Attachment: bug1186617tools-2.patch

patch with whitespace fixed

Comment 21

[Kim Moir [:kmoir] ET]

tagged as PANDA_FOOPY_SUTLIB_20160314 before landing patch

Comment 22

[Kim Moir [:kmoir] ET]

Attachment: bug1186617tools-2.patch

Comment 23

[Kim Moir [:kmoir] ET]

Attachment: bug1186617puppet.patch

I ran a machine in staging against this patched puppet env and it worked fine.

Comment 24

[Dustin J. Mitchell [:dustin] (he/him)]

Comment on attachment 8730320 [details] [diff] [review]
bug1186617puppet.patch

Review of attachment 8730320 [details] [diff] [review]:
-----------------------------------------------------------------

What a sea of `-`.. I love it

The /data/bmm directory should also be deleted (maybe check with Jake to see if we should back it up somewhere first), and the mozpool-related secrets removed, too.

::: modules/slaveapi/manifests/instance.pp
@@ -63,5 @@
>                  "furl==0.3.5",
>                  "orderedmultidict==0.7.1",
>                  "pytz==2013.7",
>                  "slaveapi==${version}",
> -                "mozpoolclient==0.1.5",

Does slaveapi need to be updated first?  Without mozpoolclient, will we get ImportErrors?

Comment 25

[Kim Moir [:kmoir] ET]

dividehex: should the /data/bmm directory be backed up before we delete it?

Comment 26

[Kim Moir [:kmoir] ET]

Comment on attachment 8729131 [details] [diff] [review]
bug1186617sh.patch

due to bug 1256375

Comment 27

[Kim Moir [:kmoir] ET]

changes to slave api to remove mozpoollclient are in bug 1256375

Comment 28

[Jake Watkins [:dividehex]]

(In reply to Kim Moir [:kmoir] from comment #25)
> dividehex: should the /data/bmm directory be backed up before we delete it?

Well, I don't ever see this project being resurrected from the dead but just for nostalgia I've taken a tarball of bmm.  You can go ahead and remove it from /data

Comment 29

[Kim Moir [:kmoir] ET]

closing - I think I have removed all the panda code.

Comment 30

[Kim Moir [:kmoir] ET]

Attachment: bug1186617sh-2.patch

fix slave_health crontab to not copy devices.json

Comment 31

[Kim Moir [:kmoir] ET]

Comment on attachment 8732351 [details] [diff] [review]
bug1186617sh-2.patch

and updated slave_health on cruncher manually and now the worker health web page is updating again