PERC3/Di failure workaround hypothesis

Hooft ing. P.J.G. van hooft at
Tue May 25 06:52:00 CDT 2004


Here is another datapoint.
We have some pe2560s running RedHat 7.3, 9 or RHEL 3, using
raid1 and raid5.
We didn't have any problems on these machines except for a
server (running apache/php/postgresql/mailman) that uses raid1.

When we put this server into production it started to hang with
scsi errors at random sector numbers.

Controller details (they're the same on all of our systems):
                CLI: 2.8-0 (Build #6076)
                API: 2.8-0 (Build #6076)
    Miniport Driver: 2.7-1 (Build #3170)
Controller Software: 2.7-1 (Build #3170)
    Controller BIOS: 2.7-1 (Build #3170)
Controller Firmware: (Build #3170)

I disabled the read and write caches May 24th per Matt Domsch's
instructions and the system has not hung itself since.


Peter van Hooft
Philips Research
Eindhoven, The Netherlands

More information about the Linux-PowerEdge mailing list