682408 - setup mw32-ix-slave02, mw32-ix-slave20, w32-ix-slave03, w32-ix-slave23, w32-ix-slave26, w32-ix-slave29, w32-ix-slave34 and w32-ix-slave41

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P2)

Description

[Armen [:armenzg]]

These got reimaged:
mw32-ix-slave02.build.mtv1
w32-ix-slave03.build.scl1
w32-ix-slave23.build.scl1
w32-ix-slave26.build.scl1 
w32-ix-slave29.build.scl1
w32-ix-slave41.build.scl1

Comment 1

[Armen [:armenzg]]

production-opsi:~# opsi-admin -d method deleteClient mw32-ix-slave02.uib.local
production-opsi:~# opsi-admin -d method deleteClient w32-ix-slave03.uib.local
production-opsi:~# opsi-admin -d method deleteClient w32-ix-slave23.uib.local
production-opsi:~# opsi-admin -d method deleteClient w32-ix-slave26.uib.local
production-opsi:~# opsi-admin -d method deleteClient w32-ix-slave29.uib.local
production-opsi:~# opsi-admin -d method deleteClient w32-ix-slave41.uib.local

* fixed hostname and rebooted
* I have marked the "nagios" package to be re-deployed (see bug 677888 for details)
* I have marked them to get buildbot-production installed (ftr win32-ix-ref has it marked as installed)

Could these slaves have been reimaged from win2k3-ref-img instead of win32-ix-ref? (I see the package buildbot-production not marked as "installed" in the former which matches the state of the slaves cloned).

w32-ix-slave03 is down - see bug 682574.

Comment 3

[Armen [:armenzg]]

w32-ix-slave34 will be reimaged in bug 682931 from a new snapshot and hopefully the problems of the first batch won't happen to it. I will do the setup of this slave in this bug as well.

Comment 6

[Armen [:armenzg]]

w32-ix-slave34 has had disk errors. arr needed an mtv slave so I decided to give her mw32-ix-slave20. mw32-ix-slave20 was allocated to the buildbot-0.7 pool so it wasn't taking any jobs.

production-opsi:~# opsi-admin -d method deleteClient w32-ix-slave34.uib.local
production-opsi:~# opsi-admin -d method deleteClient mw32-ix-slave20.uib.local

[3:58pm] arr: armenzg: hm, so when I boot the new snapshot, I get a message saying that c: needs to be checked
[3:58pm] arr: and then it proceeds to spit out a LOT of errors
[3:58pm] arr: is there a machine in mtv1 I could reimage?

Comment 7

[Armen [:armenzg]]

Attachment: allow mw32-ix-slave20 and w32-ix-slave24 to be part of the production pool

I will be moving w32-ix-slave24 into the buildpool and w32-ix-slave23 into the try pool to make the configs less jumpy.

bhearsum gave me r+ on IRC.

Comment 8

[Armen [:armenzg]]

Comment on attachment 557168 [details] [diff] [review]
allow mw32-ix-slave20 and w32-ix-slave24 to be part of the production pool

I will have to wait until the releases go through so I can add the win32 slaves to the buildpool.
I would not like to take a chance to affect any release build because I missed something on my testing.

Comment 9

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

(In reply to Armen Zambrano G. [:armenzg] - Release Engineer from comment #7)
> I will be moving w32-ix-slave24 into the buildpool and w32-ix-slave23 into
> the try pool to make the configs less jumpy.

Has this happened yet? I ask because both slaves are currently spamming the twistd.log on bm14 with unauthorized login exceptions.

On logging in, first thing I noticed was that the date was set to to Sept 5th on w32-ix-slave23. I fixed that, and have disabled the 2 slaves in slavealloc until you have a chance to look at this.

Comment 10

[Armen [:armenzg]]

I have put the following slaves back into production:
mw32-ix-slave02 - connected to build-sjc1 (it could give time out issues)
mw32-ix-slave20 - connected to build-sjc1 (it could give time out issues)
w32-ix-slave08 - as a try slave
w32-ix-slave24 - as a build slave (swapped with slave23)
w32-ix-slave26
w32-ix-slave29
w32-ix-slave34
w32-ix-slave41

I will watch over them.

w32-ix-slave23 - as a try slave - waiting on reconfig

Comment 11

[Mike Taylor [:bear]]

landed with 2011-09-06 reconfig

Comment 12

[Armen [:armenzg]]

All of these slaves have been taking several jobs and have been successful.