On Wed, 4 Sep 2013, Paul Heinlein wrote:

> On Tue, 3 Sep 2013, Paul Heinlein wrote:
>
>> Drive /dev/sdb on abbot has begun its failure spiral. All of abbot's
>> filesystems are on RAID-1 arrays, so there's no immediate danger of
>> system failure. I realize that abbot is in the process of being
>> replaced, but I've ordered a replacement drive anyway.
>
> The replacement drive is scheduled to arrive on Friday, so I plan to 
> power down abbot to swap drives sometime between 7:30 and 9:30 a.m. 
> (PDT) this Saturday, September 7.
>
> Barring unforeseen problems, abbot should be offline for roughly 20 
> to 30 minutes.

It took me a couple minutes to figure out how to remove the failed 
drive from abbot's case, which I'd never opened, but otherwise the 
drive swap went according to plan.

Of course, when I rebooted abbot it insisted on fsck-ing its disk 
(not because of the drive swap but because it'd be too long since the 
last fsck) -- so abbot was offline for closer an hour than 30 minutes.

Once online, reassembling the RAID-1 mirrors was quick and easy. The 
main mirror, for the / filesystem, is still resync-ing itself; I'm 
guessing the process won't be complete until mid-evening PDT.

In the meantime, hackage and other abbot-based services are humming 
along just fine.

Since the old drive was failing, I can't reliably zero it out, so I'll 
physically destroy it.

-- 
Paul Heinlein                 Galois, Inc.
Systems Administrator         421 SW Sixth Avenue, Suite 300
[email protected]           Portland, Oregon 97204
+1 503 626-6616 x140          http://corp.galois.com/
_______________________________________________
haskell-infrastructure mailing list
[email protected]
http://community.galois.com/mailman/listinfo/haskell-infrastructure

Reply via email to