25 Apr
2008
25 Apr
'08
1:02 a.m.
On Thu, Apr 24, 2008 at 12:57:12PM -0600, M. Warner Losh wrote:
In message: <4810D417.8040005(a)bitsavers.org>
Al Kossow <aek(a)bitsavers.org> writes:
: The critical point to add to this is that the data integrity needs to be
: constantly verified, even on presumed stable storage, and it is migrated
: to what at that time is easily to deal with storage, so you CAN easily
: verify it.
I keep my archives on a series of disks. There's always at least 2
copies, often times more, and the underlying disks get swapped out on
a round-robin basis. Helps limit my exposure to one or two
failures... I've had horrible luck with all other methods...
We do the same thing here. For /home which has all the stuff we really
care about we have
/nightly - last night's copy of the data
/nightly2 - same thing, night before
/weekly - last Sunday's copy of the data
/weekly2 - same thing, week before
and we do it so that
diff foo /nightly/$PWD
works. Handy, that.
--
---
Larry McVoy lm at bitmover.com http://www.bitkeeper.com