659118 - [tracking bug] move w764 talos machines to other silos

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P4)

Description

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

Plan:

 a. move 45 of the non-functional w764 testers to other silos

 b. move the other 5 and officially stop testing on w764; we can start again when we get a windows admin to set up the cluster.

Armen, how should we distribute these among the other silos?

Comment 1

[Armen [:armenzg]]

Dustin how does this distribution look like?

I added more to XP, Win7 & Fedora because they have more jobs to run and/or are at the end of the tail (long builds for Windows).

Platform            | Initial   | Repurpose | New total
-------------------------------------------------------
talos-r3-fed-*:     | 50 slaves |    10     |    60
talos-r3-fed64-*:   | 52 slaves |     4     |    56
talos-r3-xp-*:      | 50 slaves |     9     |    59
talos-r3-w7-*:      | 48 slaves |    12     |    60
talos-r3-leopard-*: | 50 slaves |     6     |    56
talos-r3-snow-*:    | 52 slaves |     4     |    56
-------------------------------------------------------
                           SUM ->    45

[1] From https://bugzilla.mozilla.org/show_bug.cgi?id=625474#c8

Comment 2

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

lgtm

Comment 3

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

Armen: can you break this out into specific machines with specific OS and hostnames?  That way we can track the hardware through the change if necessary.  Also, please make sure the three currently-working machines are among the 5 that get left behind as w764 testers.  Then reassign to relops.  This will still be blocked on the releng-shapshots bug, but that should be done soonish.

Comment 4

[Armen [:armenzg]]

Sounds good.
What are the 5 slaves that are still running?

Comment 5

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

According to http://build.mozilla.org/builds/last-job-per-slave.txt, only 005, 006, and 009 are still running.  The other two are up to you - if you can kick two back into operation, that'd probably be best.

Comment 6

[Armen [:armenzg]]

I managed to reboot and reconnect t-r3-w764-0{03,04}

This means that we have now 5 machines up and running.

Comment 7

[Armen [:armenzg]]

It seems few slaves need to be checked:
* talos-r3-w7-040          No jobs yet
* talos-r3-fed-060         No jobs yet

Besides that every other slaves seems to have taken a job.

Comment 8

[Armen [:armenzg]]

(In reply to comment #7)
> It seems few slaves need to be checked:
> * talos-r3-w7-040          No jobs yet
Re-enabled on slave alloc.
> * talos-r3-fed-060         No jobs yet
It thinks to be ref machine. Made a comment on reboots-scl1 bug.

Comment 9

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

talos-r3-fed-060's resuscitated - it was down at the time I brought the others up.

talos-r3-w7-040's up and running as of 20m ago, but hasn't gotten a job yet.  Noted to check back in the slave spreadsheet.