880784 - (t-xp32-ix-004) t-xp32-ix-004 problem tracking

Status: RESOLVED WONTFIX

(Infrastructure & Operations Graveyard :: CIDuty, task)

Description

[Armen [:armenzg]]

Loaning to Mossop.

Comment 1

[Armen [:armenzg]]

Waiting on re-imaging.

Comment 2

[Armen [:armenzg]]

Waiting to be re-imaged.
After re-imaging they can do directly into the pool.

Comment 3

[Armen [:armenzg]]

Back into production.

Comment 4

[Phil Ringnalda (:philor)]

According to the screenshots in the logs in bug 845134, back into production with a permanent dialog wanting to know whether we want to set up a new Dropbox account, or use an existing one.

Disabled in slavealloc.

Comment 5

[Armen [:armenzg]]

The referenced screenshot:
https://bug845134.bugzilla.mozilla.org/attachment.cgi?id=730256

I don't know exactly what you mean by Dropbox.

I thought we had something in place to prevent dialogs like this.

Q, did we have anything in place to prevent the XP machines to present dialogues like this?

Comment 6

[Phil Ringnalda (:philor)]

Attachment: dropbox.png

By Dropbox I mean this particular slave is installing what appears to be a Dropbox client (and has a file named tpaint.html on the desktop which isn't part of the usual desktop litter, and generally looks like it didn't actually get reimaged post-loan); I meant the screenshots in the logs for this slave, not the stale screenshot from some other slave.

Comment 7

[[e-mail] (:bhearsum)]

IIRC, we recently discovered that "move back to original ou" doesn't mean "re-image". I requested a re-image explicitly.

Comment 8

[[e-mail] (:bhearsum)]

Back in production.

Comment 9

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

Re-imaged and returned to production.

Comment 10

[Phil Ringnalda (:philor)]

Somehow wound up with the cursor over where the reftest window appears, which breaks running reftests since random form controls are displayed as hovered. Disabled.

Comment 11

[Phil Ringnalda (:philor)]

Interesting that it's still running jobs, three days after I disabled it.

Comment 12

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

Forcibly rebooted.

Comment 13

[Phil Ringnalda (:philor)]

Somebody apparently disabled it in slavealloc and rebooted it in the middle of a job, last Friday afternoon, and when I reenabled it (since they failed to say why they disabled it), it now has the never-get-overs and isn't taking jobs. Redisabled.

Comment 14

[Alin Selagea [:aselagea]]

Re-imaged and returned to production.

Comment 15

[Phil Ringnalda (:philor)]

Mouse pointer, disabled.

Comment 16

[Phil Ringnalda (:philor)]

Reenabled.

Comment 17

[Phil Ringnalda (:philor)]

It's lost interest in taking jobs, rather a lot of reboots had no effect.

Comment 18

[Firefox Bug Husbandry Bot]

Bulk change of QA Contact to :jlund, per https://bugzilla.mozilla.org/show_bug.cgi?id=1428483

Comment 19

[Adrian Pop]

is this machine still needed ? I have checked the slave alloc and slave health >> the machine seems to be disabled.
can we close the ticket ?

Comment 20

[Phil Ringnalda (:philor)]

The answer, which you already got elsewhere, is really funny in a pathetic way. This machine is *desperately* needed, it's one of only 8 WinXP machines that have to do all the WinXP tests on esr52 until sometime in August, and this and one other are broken, so it takes 15 hours (triple what we pretend is the total time it will take us to ship a chemspill release) to finish running tests on a push, but reimaging WinXP is broken, and fixing it is wontfix, so fixing this desperately needed machine is, too.

Comment 21

[Jordan Lund (:jlund)]

yeah, re-imaging XP is broken as philor mentioned: https://bugzilla.mozilla.org/show_bug.cgi?id=1413924#c9

Pending is already painful when we need to push to esr. If we lose any more than the 8 (8!) left, we will need to figure out a different plan than crossing our fingers until Aug and Buildbot EOL