From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Brown Subject: Re: FailSpare event? Date: Fri, 12 Jan 2007 09:59:15 +1100 Message-ID: <17830.49475.108136.376489@notabene.brown> References: <17830.47341.560158.521091@notabene.brown> <20070111223628.GU32386@mikee.ath.cx> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: message from Mike on Thursday January 11 Sender: linux-raid-owner@vger.kernel.org To: Mike Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids On Thursday January 11, mikee@mikee.ath.cx wrote: > > So I'm ok for the moment? Yes, I need to find the error and fix everything > back to the (S) state. Yes, OK for the moment. > > The messages in $HOST:/var/log/messages for the time of the email are: > > Jan 11 16:04:25 elo kernel: sd 2:0:4:0: SCSI error: return code = 0x8000002 > Jan 11 16:04:25 elo kernel: sde: Current: sense key: Hardware Error > Jan 11 16:04:25 elo kernel: Additional sense: Internal target failure > Jan 11 16:04:25 elo kernel: Info fld=0x10b93c4d > Jan 11 16:04:25 elo kernel: end_request: I/O error, dev sde, sector 280575053 > Jan 11 16:04:25 elo kernel: raid5: Disk failure on sde2, disabling device. Operation continuing on 5 devices Given the sector number it looks likely that it was a superblock update. No idea how bad an 'internal target failure' is. Maybe powercycling the drive would 'fix' it, maybe not. > > On AIX boxes I can blink the drives to identify a bad/failing device. Is there > a way to blink the drives in linux? Unfortunately not. NeilBrown