Closed
Bug 863263
Opened 11 years ago
Closed 11 years ago
Create VMs to replace kvm signing servers
Categories
(Infrastructure & Operations :: Virtualization, task)
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: arich, Assigned: gcox)
References
Details
The following signing servers need to move off of kvm in scl1 and scl3 and onto vmware in scl3. signing1.build.scl1.mozilla.com signing2.build.scl1.mozilla.com signing3.srv.releng.scl3.mozilla.com bhearsum: do we have the capability to cross-datacenter signing since we'll still have a number of slaves in scl1? I thought I remembered someone saying that we didn't do that right now. Also, is three sufficient capacity, and are there signing servers anywhere else for redundancy?
Comment 1•11 years ago
|
||
(In reply to Amy Rich [:arich] [:arr] from comment #0) > The following signing servers need to move off of kvm in scl1 and scl3 and > onto vmware in scl3. > > signing1.build.scl1.mozilla.com > signing2.build.scl1.mozilla.com > signing3.srv.releng.scl3.mozilla.com > > bhearsum: do we have the capability to cross-datacenter signing since we'll > still have a number of slaves in scl1? I thought I remembered someone > saying that we didn't do that right now. Also, is three sufficient > capacity, and are there signing servers anywhere else for redundancy? We avoid it where ever possible right now to reduce the amount of cross-colo bandwidth. It shouldn't be a problem to do cross colo signing though as long as the links are fast enough and hold up to the extra traffic.
Reporter | ||
Comment 2•11 years ago
|
||
After an issue with signing3 today, we talked a bit more about this, and we're already doing cross datacenter signing from AWS. So Ben doesn't think this is a technical limitation, just one of perceived performance. Since the only builders left in scl1 are for windows, I don't think we'd be seeing that much difference in performance by putting the servers in scl3. I'm going to get the ball rolling here and request 3 new VMWare vms with the following specs: 2 CPU 300G of disk (combined) for / and /boot 4G RAM (and 4G swap) VLAN 248 in SCL3 CentOS 6.2 (we'll be hooking these up to releng puppet, so please do NOT puppetize with infra puppet) Hostnames: signing4.srv.releng.scl3.mozilla.com signing5.srv.releng.scl3.mozilla.com signing6.srv.releng.scl3.mozilla.com
Assignee: server-ops-releng → server-ops-virtualization
Component: Server Operations: RelEng → Server Operations: Virtualization
QA Contact: arich → dparsons
Reporter | ||
Comment 3•11 years ago
|
||
Note that these are disk intensive, so SAS over SATA if possible.
Assignee | ||
Comment 4•11 years ago
|
||
DNS was sitting there already (yay). Cloned out 3 copies, bumped CPU and RAM and disk, logged MAC in inventory, gave them their hostnames, updated vmware tools. Did nothing with puppet, per :arr, and did not yum update. Verified root is remotely available via a password in the gpg file.
Assignee: server-ops-virtualization → gcox
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Updated•10 years ago
|
Product: mozilla.org → Infrastructure & Operations
You need to log in
before you can comment on or make changes to this bug.
Description
•