Date: Fri, 1 Feb 2008 21:10:57 +0100 From: Erik Trulsson <ertr1013@student.uu.se> To: Kurt Buff <kurt.buff@gmail.com> Cc: FreeBSD Questions <freebsd-questions@freebsd.org> Subject: Re: This has begun to annoy me... Message-ID: <20080201201057.GA38220@owl.midgard.homeip.net> In-Reply-To: <a9f4a3860802011140x5d50663bgfadcbaf351c95f6d@mail.gmail.com> References: <a9f4a3860802011140x5d50663bgfadcbaf351c95f6d@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Feb 01, 2008 at 11:40:32AM -0800, Kurt Buff wrote: > I've been getting this in my daily security email from one of my boxes > for quite a while, and have been ignoring it, because of workload. > > However, it's finally annoyed me enough to pursue it. > > What would the significance of the following section be? > > zsquid.mycompany.com kernel log messages: > +++ /tmp/security.4blejPLW Fri Feb 1 03:02:08 2008 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=4662143 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=113995359 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=112970015 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=4668319 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=4849151 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=115527359 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=113714335 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=113715199 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=4570911 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=42958719 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=113343327 > +ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) LBA=117315327 > > It looks like disk write errors to me, but I'm not sure. > > Thoughts? Not so much write *errors* as write *problems*. What seems to have happened is that some write requests do not finish fast enough so the ATA driver gets a timeout, but retries the operation and it works on the second try. (If it hadn't worked on the second try either then there should have been much scarier messages in the log.) It might be because of a disk going bad where the disk has to do remapping of bad blocks when it detects write problems which take extra time, causing timeouts. It might be because of sub-standard cabling which cause errors (but then there ought to have been problems with reading as well as writing.) It might be because FreeBSD's ATA driver uses a timeout value which is a bit too aggressive (too low) for your controller/disk combination. You could changing the disk for another. If you get similar messages with the other disk too, then it is probably not the disk which is at fault. -- <Insert your favourite quote here.> Erik Trulsson ertr1013@student.uu.se
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080201201057.GA38220>