PERC3/Di failure workaround hypothesis
pzero at hotpop.com
Fri May 21 04:08:01 CDT 2004
On 20-05-2004 23:06, Matt Domsch wrote:
> We believe we have root cause to the SCSI command timeouts seen with
> the PERC3/Di, related to how the RAID controller firmware handles
> read and write caching. To be certain that this is the *only*
> problem, we ask you to do the following, and to report back to the
> list success (i.e. you no longer see the SCSI mid-layer time out
> commands, the system runs as expected), or failure (you still see
> command timeouts). If the latter occurs, then there is a secondary
> problem that we'll need additional details from you to help debug.
what servers and Linux versions are affected by this problem?
Is there any way to reproduce the problem in a consistent way?
More information about the Linux-PowerEdge