Essay on RAID reliability

Christopher Browne cbbrowne-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org
Tue Feb 7 23:31:04 UTC 2012


http://etbe.coker.com.au/2012/02/06/reliability-raid/

Interesting article describing some of the back-and-forth of risks
relating to whether or not you actually get improved reliability if
you use Really Big Disks to construct RAID arrays.

It seems as though it's challenging to figure out which aspects of the
overall storage system are implicitly and/or explicitly highly
trusted.

I observe that discussions are ongoing on the Postgres hackers list on
the merits of doing per-page checksums, particularly as some of the
failure cases that can take place do not lend themselves to having any
reasonable solution beyond "let me get out a backup tape," and there's
a fair bit of risk of false positives (e.g. - checksum claims failure
of a write, even though what failed was the attempt to flush out the
checksum).  Use of ZFS/BTRFS may be preferable, though it's not
perfect, as it's pushing risks to a different location.
-- 
When confronted by a difficult problem, solve it by reducing it to the
question, "How would the Lone Ranger handle this?"
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists





More information about the Legacy mailing list