390493 - Create "production" linux VM for Mozilla1.8/FF2.0.0x by copying prometheus-vm

Status: RESOLVED FIXED

(mozilla.org Graveyard :: Server Operations, task, P2)

Description

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

To further test build automation, we need a linux buildbot slave to be used by a "productions" buildbot master. This needs to run in a VM which contains the exact same bits as we currently use in prometheus-vm.

Once the new VM is copied and running, we need to setup a buildbot slave.

Comment 1

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

This vm should be setup the same as the staging linux buildbot slave (see bug#388333 for details).

Comment 2

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

Could you please create this new "production-prometheus-vm" from the recent backup to avoid downtime?

Just talked with mrz about this.

Comment 3

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

From offline discussions, this VM does need to use local storage until the performance questions are resolved...

Comment 4

[Jeremy Orem [:oremj]]

mrz, will you look at these?

Comment 5

[matthew zeier [:mrz]]

To make this easier, I added the backup copy to the inventory and am doing a clone.  I'll leave it inventoried so you can do this in the future as well.

Comment 6

[matthew zeier [:mrz]]

Box up at 10.2.71.17.

Comment 7

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

1) this has a DNS entry of production-prometheus-vm.build.mozilla.org.

2) This bug originally talks about two things:
- "IT please clone a VM". This is now done.
- "Build please install a buildbot slave". This is now tracked in bug#391786
Summary updated to clarify.

Comment 8

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

Unable to use this vm because the file-system is read-only, hence reopening bug. Help??


[buildmaster@build-console Automation]$ whoami
buildmaster
[buildmaster@build-console Automation]$ cd
[buildmaster@build-console ~]$ pwd
/home/buildmaster
[buildmaster@build-console ~]$ touch john
touch: cannot touch `john': Read-only file system
[buildmaster@build-console ~]$
[buildmaster@build-console Automation]$ cd Automation
[buildmaster@build-console Automation]$ pwd
/home/buildmaster/Automation
[buildmaster@build-console Automation]$ ls -la
total 88
drwxrwxr-x  15 buildmaster buildmaster  4096 Aug  8 12:32 .
drwx------   6 buildmaster buildmaster  4096 Aug  9 09:48 ..
drwxrwxr-x   2 buildmaster buildmaster  4096 Aug  8 12:31 buildbot-configs
-rw-rw-r--   1 buildmaster buildmaster   265 Aug  8 12:30 buildbot.tac
-rw-------   1 buildmaster buildmaster   172 Aug  8 12:31 changes.pck
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 l10nverify
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 linux_build
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 linux_update_verify
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 macosx_build
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 macosx_update_verify
lrwxrwxrwx   1 buildmaster buildmaster    27 Aug  8 12:31 master.cfg -> buildbot-configs/master.cfg
lrwxrwxrwx   1 buildmaster buildmaster    28 Aug  8 12:31 mozilla.css -> buildbot-configs/mozilla.css
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 sign
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 source
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 stage
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 tag
-rw-rw-r--   1 buildmaster buildmaster 13795 Aug  9 22:19 twistd.log
-rw-------   1 buildmaster buildmaster     5 Aug  8 12:31 twistd.pid
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 update
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 win32_build
drwx------   2 buildmaster buildmaster  4096 Aug  8 12:31 win32_update_verify
[buildmaster@build-console Automation]$ cp master.cfg master.cfg-backup2007aug10_01
cp: cannot create regular file `master.cfg-backup2007aug10_01': Read-only file system
[buildmaster@build-console Automation]$ 

Comment 9

[Nick Thomas [:nthomas] (UTC+12)]

I hit also found read-only file system over the weekend, with l10n-linux-tbox (which is new last week). Both the /var (/dev/sda2) and /builds (/dev/sdb1) disks had the problem, but I've rebooted and fsck'd since.

Comment 10

[Nick Thomas [:nthomas] (UTC+12)]

urgh, how about:

I also found read-only file system over the weekend, on l10n-linux-tbox
(which was new last week). Both the /var (/dev/sda2) and /builds (/dev/sdb1)
disks had the problem, but I've rebooted and fsck'd since. It's based on the new CentOS5 ref platform, unlike John's box.

Comment 11

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

ugh, sorry, wrong bug. Reopened in error.

The problem I hit was with build-console.build.mozilla.org, not with production-prometheus-vm.build.console.org.