Date: Fri, 10 Jun 2011 02:33:18 -0700 From: Jeremy Chadwick <freebsd@jdc.parodius.com> To: Karl Pielorz <kpielorz_lst@tdx.co.uk> Cc: freebsd-fs@freebsd.org Subject: Re: ZFS scrub 'repaired' pool with no chksum or read errors? Message-ID: <20110610093318.GA39276@icarus.home.lan> In-Reply-To: <729A0755FAEF480774EEF4AB@HexaDeca64.dmpriest.net.uk> References: <729A0755FAEF480774EEF4AB@HexaDeca64.dmpriest.net.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Jun 10, 2011 at 09:43:14AM +0100, Karl Pielorz wrote: > I'm running FreeBSD-8.2R amd64 w/4Gb of ECC RAM on a machine used > for 'offsite' backups (that are copied to it using zfs > send/receive). > > I scrub this machine every now and again (about once a month) - > recently this resulted in the following output after the scrub > completed: > > " > # zpool status > pool: vol > state: ONLINE > scrub: scrub completed after 2h49m with 0 errors on Thu Jun 9 > 17:09:31 2011 > config: > > NAME STATE READ WRITE CKSUM > vol ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > ada0 ONLINE 0 0 0 256K repaired > ada1 ONLINE 0 0 0 > ada2 ONLINE 0 0 0 > > errors: No known data errors > " > > Should I be worried there was 256k of 'repairs' done, even though > there were no checksum errors, or read errors detected? > > The console logged no errors - and nothing shows in syslog. > > The machine is always cleanly shut down - and the drives all appear > fine from a SMART point of view - I'm just a bit concerned as to > where the repairs came from - as ZFS doesn't seem to know (or be > able to tell me) either :) ZFS experts please correct me, but my experience with this has shown me that the scrub itself found actual issues while analysing all data on the entire pool -- more specifically, I believe READ/WRITE/CKSUM are counters used for when errors are encountered during normal (read: non-scrub) operations. It's been a while since I've seen this happen, but have seen it on our Solaris 10 machines at my workplace. I've never been sure what it means; possibly signs of "bit rot"? If you're worried about your disk (ada0), please provide output from "smartctl -a /dev/ada0" and I'll be more than happy to review the output and provide you with any insights. I do believe you when you say it looks fine, but every model of disk is different in some regard. -- | Jeremy Chadwick jdc@parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, US | | Making life hard for others since 1977. PGP 4BD6C0CB |
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20110610093318.GA39276>