Closed Bug 572578 Opened 15 years ago Closed 15 years ago

Expand disk space on khan

Categories

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

All
Other
task
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: clouserw, Assigned: jabba)

Details

(Whiteboard: [onhold, eow])

We talked a bit in IRC and I said I'd file this bug. It sounds like we could add an array to khan or mrz wanted to try SAS via NFS. I don't know what the conclusion was, but it sounded like we should do something. :)
Phong - price out an SB40c with 6x300GB SAS drives and with 6x146GB. Wil's concerned about NFS performance so not sure if that's the best way to go.
Assignee: server-ops → phong
I'm open minded, I've never used SAS+NFS. I think I'd be more comfortable if we had a reliable way to monitor what was slow. Benchmarking current disk vs NFS would be good. Are the NFS mounts in /data SAS?
No, all off SATA backed storage. Aravind might be able to allocate some faster NFS space for you get benchmark.
Did we decide that NFS will work for this or should I order the storage blade?
Assignee: phong → server-ops
mrz says storage blade, punting back to phong.
Assignee: server-ops → phong
sb40 with 6x300 GB drives on order.
the storage has arrived. we will need to schedule a downtime to move this blade to a different chassis that has space for the storage blade.
Flags: needs-downtime+
Whiteboard: [needs khan downtime]
guys, can we get this scheduled?
The main concern of the webdev team is just getting notice - we're flexible on the actual time it happens. If you want a time, I'll throw out Thursday @ 6pm, but we're easy.
This requires us to physically move khan to a different blade chassis that has space for the storage blade to be right next to it. The other alternative is to get shaver's okay for downtime for sm-valgrind01 and move that blade instead.
I can move the blade to the other chassis tonight at 6pm. Once that is done, the rest should be able to be done remotely. If we need to move /home to the new storage blade, we might need an additional downtime to make that move, but should be short and trivial.
Assignee: phong → jdow
Whiteboard: [needs khan downtime] → 8/19 @ 18:00
Flags: colo-trip+
Whiteboard: 8/19 @ 18:00 → 8/19/2010 @ 18:00
=> ctrl slot=3 show config Smart Array P400 in Slot 3 (sn: PAFGL0T9SYP03N) array A (SAS, Unused Space: 0 MB) logicaldrive 1 (1.4 TB, RAID 5, OK) physicaldrive 1I:1:5 (port 1I:box 1:bay 5, SAS, 300 GB, OK) physicaldrive 1I:1:6 (port 1I:box 1:bay 6, SAS, 300 GB, OK) physicaldrive 2I:1:1 (port 2I:box 1:bay 1, SAS, 300 GB, OK) physicaldrive 2I:1:2 (port 2I:box 1:bay 2, SAS, 300 GB, OK) physicaldrive 2I:1:3 (port 2I:box 1:bay 3, SAS, 300 GB, OK) physicaldrive 2I:1:4 (port 2I:box 1:bay 4, SAS, 300 GB, OK) [root@khan ~]# fdisk /dev/cciss/c1d0 Unable to read /dev/cciss/c1d0 [root@khan ~]# fdisk /dev/cciss/c1d0 /var/log/messages shows kernel messages that make me think something broke in the driver. When can khan go down for a quick reboot?
Flags: needs-downtime-
Flags: needs-downtime+
Flags: colo-trip-
Flags: colo-trip+
Whiteboard: 8/19/2010 @ 18:00
6pm tonight?
Missed the reboot and folks are logged in. If everyone can jump offline I can do it over the weekend.
That reboot worked. Working on copying /home to this new drive. Hopefully everyone can refrain from logging back in until this is done.
Ran the large rsync yesterday. Just need a window to move /home over. Should be < 30 mins.
monday @ 6pm? or 7:30am?
I can do 7:30am if it's not too late to make the announcement. Let me know.
Didn't hear anything, so let's do this at 6pm tonight.
This is done now. The old /home is still around under /home.old, but I want to remove that as soon as I get an "ok" from webdev. I did rsync -a on it, so everything should be there.
Just sent to webdev@ > Hey, > > IT finished expanding /home on khan today. The old directory is in /home.old. > It should be the same, but double check if you're missing anything. If you > don't say anything by the end of the week, they're going to whack it. > > Wil
Whiteboard: [onhold, eow]
/home.old has been nuked and there is now plenty of disk space to go around.
Status: NEW → RESOLVED
Closed: 15 years ago
Resolution: --- → FIXED
Product: mozilla.org → mozilla.org Graveyard
You need to log in before you can comment on or make changes to this bug.