Date: Tue, 5 Feb 2008 13:06:28 -0600 From: Brooks Davis <brooks@freebsd.org> To: Joe Peterson <joe@skyrush.com> Cc: freebsd-fs@freebsd.org, Dag-Erling Sm??rgrav <des@des.no>, Brooks Davis <brooks@freebsd.org> Subject: Re: Forcing full file read in ZFS even when checksum error encountered Message-ID: <20080205190628.GA86728@lor.one-eyed-alien.net> In-Reply-To: <47A8B24D.9050904@skyrush.com> References: <47A73C8D.3000107@skyrush.com> <86prvby5o1.fsf@ds4.des.no> <47A864D9.4060504@skyrush.com> <864pcnxz8f.fsf@ds4.des.no> <47A88ADE.7050503@skyrush.com> <86abmfwc6h.fsf@ds4.des.no> <20080205173102.GA85735@lor.one-eyed-alien.net> <47A8B24D.9050904@skyrush.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --] On Tue, Feb 05, 2008 at 12:00:29PM -0700, Joe Peterson wrote: > Brooks Davis wrote: > > We've also experienced several situations were zfs was detecting > > corruption caused by bad cabling or bad controller firmware so SMART had > > nothing to report. > > Were the errors sporadic/intermittent? I'd think that if it was cabling, the > checksum error could happen some times and not others. Mine seems to be > predictable and always there... Also, no error was reported when the file was > written, and I think ZFS checks after writing. Strange. This is one reason > why I'm very curious which bits are bad. I'm not sure if they were consistent. We didn't really care since either the data one the disk was platter wasn't the data we put there or we couldn't read reliably. We've not seen any checksum errors (in our somewhat limited experience) that weren't a symptom of a real, underlying problem. -- Brooks [-- Attachment #2 --] -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (FreeBSD) iD8DBQFHqLOzXY6L6fI4GtQRAnNnAJoCZqqBdXS6DF3H4N4+QxXgF42UxACg3anr W9puMMUQ98Rq9hDD57Ay8u0= =o34K -----END PGP SIGNATURE-----
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080205190628.GA86728>