Date: Mon, 9 Jul 2012 12:58:14 -0700 (PDT) From: Jason Usher <jusher71@yahoo.com> To: Zaphod Beeblebrox <zbeeble@gmail.com> Cc: freebsd-fs@freebsd.org Subject: Re: vdev/pool math with combined raidzX vdevs... Message-ID: <1341863894.36655.YahooMailClassic@web122501.mail.ne1.yahoo.com>
next in thread | raw e-mail | index | archive | help
Hello again,=0A=0A=0A--- On Fri, 7/6/12, Zaphod Beeblebrox <zbeeble@gmail.c= om> wrote:=0A=0A=0A> ... so, again with simplistic assumptions,=0A> =0A> p(= 36drz3 --- 12 drives, 3 groups) =3D p(12drz3) * 3=0A> =0A> A "vanilla" RAID= -Z2 (if I make an assumption to what you're=0A> saying) is:=0A> =0A> p(36dr= z2) =3D 36 * p(f) * 35 * p(f)=0A> =0A> ... but I can't directly answer you = question without knowing=0A> a) the=0A> structure of the RAID-Z2 array and = p(f).=A0 If we use a=0A> 1% figure for=0A> p(f), then P(36drz3,12,3) =3D 0.= 035% and p(36drz2) =3D 4.3%=0A=0A=0A(snip)=0A=0A=0A> Put simply, you add th= e probabilities of things where any=0A> can cause=0A> the failure (either d= rive of R0 failing, any one of the 3=0A> plexes of a=0A> complex array fail= ing) and you multiply things where all=0A> must fail to=0A> produce failure= .=0A=0A=0AOk. So let's start with those numbers from that hardforum link I= posted:=0A=0A(probability of data loss during a rebuild)=0A=0ARAID-10: =0A= F =3D 5%=0A=0ARAID-Z1:=0A1 - (1 - F)^(9 - 1) =3D 33.7%=0AF=3D 33.7%=0A=0ARA= ID-Z2:=0A1 - (1 - F)^(10 - 1) - (10 - 1) F (1 - F)^(10 - 2) =3D 7.1%=0AF=3D= 7.1%=0A=0ARAID-Z3:=0A1 - (1 - F)^(11 - 1) - (11 - 1) F (1 - F)^(11 - 2) - (= 11 - 1)(11 - 2) F^2 (1 - F)^(11 - 3) / 2=0AF =3D 1.15%=0A=0AAgain, doesn't = really matter what F is, since we are only interested in the comparison...= =0A=0AFrom what you said, above, striping 3 different raidz3 arrays togethe= r into one pool is ADDITIVE ... so the 1.15% rises to 3.45%.=0A=0AYes ?=0A= =0ASo we triple our risk by running all three raidz3 arrays in one pool, bu= t we still have less than half the risk of a single raidz2 vdev (with no st= riping) which is 7.1%.=0A=0AAm I on the right track here ? I think I'm mis= sing something because with one raidz3, I have a 1.15% chance of "losing a = drive during rebuild" but I am thinking about competely healthy arrays who = have a larger chance of blowing up because ONE OF THE OTHER vdevs blows fou= r drives simultaneously. =0A=0ASo I am really comparing 0% probability (if= they aren't combined in a zpool, I can take one vdev out and run over it w= ith a train and the other vdev is unharmed) with X% probability, because no= w something happening in the other vdev can ruin the healthy one...=0A=0AAm= I really the only person worrying about the interactive failure properties= of combining vdevs into a pool ?
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1341863894.36655.YahooMailClassic>