From mboxrd@z Thu Jan 1 00:00:00 1970 From: "David Lethe" Subject: Re: Distributed spares Date: Tue, 14 Oct 2008 8:30:00 -0500 Message-ID: <2cda01c92e01$12827a4c$3e01a8c0@exchange.rackspace.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Sender: linux-raid-owner@vger.kernel.org To: =?iso-8859-1?Q?Keld_J=F8rn_Simonsen?= , "Martin K. Petersen" Cc: Billy Crook , Justin Piszcz , Bill Davidsen , Neil Brown , Linux RAID List-Id: linux-raid.ids -----Original Message----- =46rom: Keld J=F8rn Simonsen Subj: Re: Distributed spares Date: Tue Oct 14, 2008 8:06 am Size: 1K To: "Martin K. Petersen" cc: "Billy Crook" ; "Justin Piszcz" ; "Bill Davidsen" ; "Neil Brown" ; "Linux RAID" On Tue, Oct 14, 2008 at 06:12:29AM -0400, Martin K. Petersen wrote:=20 > >>>>> "Keld" =3D=3D Keld J=F8rn Simonsen writes:=20 > =20 > Keld> I have also been thinking a little on this. My idea is that if=20 > Keld> bit errors develop on disks, then there is first maybe one bit=20 > Keld> error, and the crc check on the disk sectors then finds and=20 > Keld> corrects these.=20 > =20 > Keld> If you rewrite such bit errors, then that bit error will be=20 > Keld> corrected, and you prevent the one-bit error from developing to= =20 > Keld> a two-bit error that is not correctable by the CRC.=20 > =20 > I think you are assuming that disks are much simpler than they=20 > actually are.=20 > =20 > A modern disk drive protects a 512-byte sector with a pretty strong=20 > ECC that's capable of correcting errors up to ~50 bytes. Yes, that's= =20 > bytes.=20 > =20 > Also, many drive firmwares will internally keep track of problematic=20 > media areas and rewrite or reallocate affected blocks. That includes= =20 > stuff like rewriting sectors that are susceptible to bleed due to=20 > being adjacent to write hot spots.=20 =20 Good to know. Could yo tell me if this is actually true for normal=20 state-of-the art SATA disks, or only true for more expensive disks?=20 Do you have a good reference for it.=20 =20 best regards=20 keld=20 --=20 To unsubscribe from this list: send the line "unsubscribe linux-raid" i= n=20 the body of a message to majordomo@vger.kernel.org=20 More majordomo info at http://vger.kernel.org/majordomo-info.html=20 =20 read the manual for any disk drive... They go into error detection, co= rrection recovery algorithms and capability in great detail. =20 -- To unsubscribe from this list: send the line "unsubscribe linux-raid" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html