Closed
Bug 579111
Opened 15 years ago
Closed 15 years ago
Add 5 linux32 build slaves to trybuild pool
Categories
(Release Engineering :: General, defect, P2)
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: joduinn, Assigned: coop)
References
Details
(Whiteboard: [hardware][buildduty])
Attachments
(2 files, 1 obsolete file)
1.37 KB,
patch
|
bhearsum
:
review+
coop
:
checked-in+
|
Details | Diff | Splinter Review |
1.09 KB,
patch
|
bhearsum
:
review+
coop
:
checked-in+
|
Details | Diff | Splinter Review |
After the VMs are created in bug#579110, this bug is to track any manual setup
we need to do in the slaves, and the matching buildbot masters setup.
This should improve our linux32 wait times on production-master02:8011
Reporter | ||
Updated•15 years ago
|
Priority: -- → P5
Comment 1•15 years ago
|
||
That's the try-as-a-branch master, so the hostnames on bug 579110 aren't right. Why don't we recycle some moz2-linux-slaveNN VMs to try ? Besides, I thought ESX was low on storage and resources.
Reporter | ||
Updated•15 years ago
|
Whiteboard: [hardware]
Reporter | ||
Comment 2•15 years ago
|
||
(In reply to comment #1)
> That's the try-as-a-branch master, so the hostnames on bug 579110 aren't right.
Hostnames now fixed in dep.bug. Thanks for spotting that.
> Why don't we recycle some moz2-linux-slaveNN VMs to try ?
I'd rather not delete moz2-linux-slaveNN VMs until we have a clearer picture of load on builders without any unittest jobs involved. Once we have that data, yes, we can juggle as appropriate.
> Besides, I thought
> ESX was low on storage and resources.
From audit with mrz earlier this week, we have space/capacity for these new slaves (combination of remaining space, and space freed up by deleting old/obsolete VMs).
Reporter | ||
Comment 3•15 years ago
|
||
(In reply to comment #2)
> (In reply to comment #1)
> > That's the try-as-a-branch master, so the hostnames on bug 579110 aren't right.
> Hostnames now fixed in dep.bug. Thanks for spotting that.
Tweaked summary to correct pool.
Assignee: nobody → armenzg
OS: Mac OS X → Linux
Summary: Add 5 linux32 build slaves to production pool → Add 5 linux32 build slaves to trybuild pool
Updated•15 years ago
|
Assignee: armenzg → nobody
OS: Linux → Mac OS X
Summary: Add 5 linux32 build slaves to trybuild pool → Add 5 linux32 build slaves to production pool
Whiteboard: [hardware] → [hardware][buildduty]
Comment 4•15 years ago
|
||
Armen, did you mean to undo the summary correction, assignee, and OS that joduinn set ?
Comment 5•15 years ago
|
||
Nope... and back!
Assignee: nobody → armenzg
Status: NEW → ASSIGNED
OS: Mac OS X → Linux
Priority: P5 → P2
Summary: Add 5 linux32 build slaves to production pool → Add 5 linux32 build slaves to trybuild pool
Comment 6•15 years ago
|
||
I had to fix these on these slaves on buildbot.tac:
- buildmaster_host = 'sm-staging-try-master.mozilla.org'
+ buildmaster_host = 'production-master02.build.mozilla.org'
- port = 9982
+ port = 9011
and I had to remove the production keys and scp the try keys:
> rm .ssh/* && cd .ssh
> scp cltbld@try-linux-slave20:/home/cltbld/.ssh/* .
For the first block of changes on buildbot.tac were do I update the 0 -> staging scripts to fix this?
Anything we could wrt the try keys as well?
Comment 7•15 years ago
|
||
Attachment #460901 -
Flags: review?(bhearsum)
Comment 8•15 years ago
|
||
Attachment #460902 -
Flags: review?(bhearsum)
Comment 9•15 years ago
|
||
(In reply to comment #6)
> I had to fix these on these slaves on buildbot.tac:
> - buildmaster_host = 'sm-staging-try-master.mozilla.org'
> + buildmaster_host = 'production-master02.build.mozilla.org'
> - port = 9982
> + port = 9011
> and I had to remove the production keys and scp the try keys:
> > rm .ssh/* && cd .ssh
> > scp cltbld@try-linux-slave20:/home/cltbld/.ssh/* .
>
> For the first block of changes on buildbot.tac were do I update the 0 ->
> staging scripts to fix this?
>
> Anything we could wrt the try keys as well?
I'd prefer to wait until we have a general solution for ssh key management to anything about that. Could you update buildbot/buildbot-tac.py in the tools repo, though?
Comment 10•15 years ago
|
||
Attachment #460906 -
Flags: review?(bhearsum)
Updated•15 years ago
|
Attachment #460902 -
Attachment is obsolete: true
Attachment #460902 -
Flags: review?(bhearsum)
Updated•15 years ago
|
Attachment #460906 -
Flags: review?(bhearsum) → review+
Updated•15 years ago
|
Attachment #460901 -
Flags: review?(bhearsum) → review+
Comment 11•15 years ago
|
||
I won't have time to deploy this and check that everything is OK before EOD.
Putting back into the pool.
All patches are reviewed positively, though.
Assignee: armenzg → nobody
Priority: P2 → --
Assignee | ||
Comment 12•15 years ago
|
||
I'll try to get these deployed today.
Assignee: nobody → ccooper
Priority: -- → P2
Assignee | ||
Comment 13•15 years ago
|
||
Comment on attachment 460906 [details] [diff] [review]
add new set of slaves to our configurations
http://hg.mozilla.org/build/buildbot-configs/rev/b2be0970b7ba
Attachment #460906 -
Flags: checked-in+
Assignee | ||
Comment 14•15 years ago
|
||
Comment on attachment 460901 [details] [diff] [review]
puppet changes for new set of try slaves
http://hg.mozilla.org/build/puppet-manifests/rev/efef728cd3c1
Attachment #460901 -
Flags: checked-in+
Assignee | ||
Comment 15•15 years ago
|
||
These slaves are now connected to try.
Status: ASSIGNED → RESOLVED
Closed: 15 years ago
Resolution: --- → FIXED
Updated•12 years ago
|
Product: mozilla.org → Release Engineering
You need to log in
before you can comment on or make changes to this bug.
Description
•