RE: [dm-devel] scsi errors

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sat, 8 Jan 2005, Guy Coates wrote:

> On Sat, 8 Jan 2005, christophe varoqui wrote:
>
> > On Sat, 2005-01-08 at 14:53 +0000, Guy Coates wrote:
> >
> > > The filesystem is XFS.
> > >
> > > To trigger the problem I simply mount the filesystem and wait. There is no
> > > IO  apart from the filesystem journaling activity.
> > >
> > Not even the checkers ?
> >

Ok, it seems to be the TUR path checker.

When I run the multipath command to create the initial device map I get
the UNDERRUN error reported for each target.

Jan  9 10:36:40 web-4-09 kernel: scsi(3:0:5) UNDERRUN status detected 0x15-0x800.
Jan  9 10:36:40 web-4-09 kernel: scsi(3:0:5) UNDERRUN status detected 0x15-0x800.

However, I can mount the multipath device and do sustained, error free IO
to it, so long as multipathd is not running.

If I start multipathd I start seeing errors within 35 minutes of starting
the deamon.

Jan 9 10:25:08 web-4-09 kernel: scsi(3): Found in Pending queue pid821520, State = 6., fcport state=4 jiffies=37d2967 Jan 9 10:25:08 web-4-09
kernel: scsi(3:0:6) qla2x00_done: did_error = 2, comp-scsi= 0x0-0x0.
etc... <snip>

If I use the readsector 0 path_checker, I still get the UNDERRUN
errors on initial device map creation, but I don't get any further
errors[1] after starting multipathd and it all works correctly.

Guy


[1] (The ghost paths report Device sdX not ready. as expected)

-- 
Dr. Guy Coates,  Informatics System Group
The Wellcome Trust Sanger Institute, Hinxton, Cambridge, CB10 1SA, UK
Tel: +44 (0)1223 834244 ex 7199



[Index of Archives]     [DM Crypt]     [Fedora Desktop]     [ATA RAID]     [Fedora Marketing]     [Fedora Packaging]     [Fedora SELinux]     [Yosemite Discussion]     [KDE Users]     [Fedora Docs]

  Powered by Linux