On Mon, Jul 28, 2008 at 04:06:48PM -0500, Andrew M. Hettinger wrote:
> 
> additionally, when you DO replace a volume, a resilver occurs (which has
> all the effect of a scrub as well). Latent errors might be noticed in
> rarely used data during this process... If that's the case a raidz1 might
> not be able to recover them (as you already have one drive pulled being
> replaced), while a raidz2 likely would.

That's a valid concern, but in our configuration it's probably OK.
Clients will mirror LUNs across a pair of thumpers, so in the case above
they can resync from the other one.

A very good point though, and one which I'll factor in to the figures,
thanks.

Ceri
-- 
That must be wonderful!  I don't understand it at all.
                                                  -- Moliere

Attachment: pgpDcjchXptTL.pgp
Description: PGP signature

_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to