Date: Wed, 11 Jul 2012 00:49:34 -0700 (PDT) From: Jason Usher <jusher71@yahoo.com> To: Bob Friesenhahn <bfriesen@simple.dallas.tx.us> Cc: freebsd-fs@freebsd.org Subject: Re: vdev/pool math with combined raidzX vdevs... Message-ID: <1341992974.53118.YahooMailClassic@web122503.mail.ne1.yahoo.com> In-Reply-To: <alpine.GSO.2.01.1207101905120.27589@freddy.simplesystems.org>
next in thread | previous in thread | raw e-mail | index | archive | help
=0AHello Bob,=0A=0A--- On Tue, 7/10/12, Bob Friesenhahn <bfriesen@simple.da= llas.tx.us> wrote:=0A=0A> > Am I really the only person worrying about the= =0A> interactive failure properties of combining vdevs into a=0A> pool ?=0A= > =0A> Yes.=A0 You are the only one.=A0 The strength of the=0A> individual = vdev is the primary determining factor of the=0A> strength of the pool.=0A= =0A=0AThanks for responding. So I must be mistaken, and the failure probab= ility of each vdev is not additive ? As I mentioned earlier in the thread,= I am not a probability person, nor would I trust my own calculations if I = tried.=0A=0ABecause if it is additive, combining vdevs erases about half of= the difference between raidz2 and raidz3, which I think is fairly signific= ant.=0A=0ACan we at least agree that it's not the same as a lone vdev ? If= I can destroy a vdev by blowing 4 disks in it, OR by blowing 4 disks in so= me other vdev, that's a higher risk than if it were alone...
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1341992974.53118.YahooMailClassic>