Hello managers,
we are using a SWXCR raid controller (with a 24GB RAID 5 set) at one of
our servers
(Digital Unix 4.0d) serving the 24GB ADVFS filesystem /nfs with NFS for
Digital Unix and
HP-UX clients.
This morining I got an email from the SWXCR monitor with the following
message:
"The hard disk at channel 0, target 2 had a miscellaneous error."
The filesystem was okay - so I decided to just run a "swxcrcheck -o
/var/log/raid.log -r" to
check for possible problems, the check runs fine with the message at
raid.log:
"No bad blocks were found."
The raid.log was written at 09:50 - but at 9:50 I also had an ADVFS
error and the following
message at uerf:
----- EVENT INFORMATION -----
EVENT CLASS ERROR EVENT
OS EVENT TYPE 198. ASTRO CONTROLLER
SEQUENCE NUMBER 4.
OPERATING SYSTEM DEC OSF/1
OCCURRED ON SYSTEM saar
SYSTEM ID x0005000F
SYSTYPE x00000000
----- UNIT INFORMATION -----
CLASS x0000 DISK
SUBSYSTEM x0000 DISK
BUS # x0000
x0000 LUN x0
TARGET x0
the /nfs was dead - no umount and no mount possible. The real trouble
began - so
decided to shutdown the machine, check the status of the SWXCR (which
was okay)
and booted again. The /nfs was mounted without any errors - but what
happend???
What should I do in the future? Shouldn't I swxcrcheck the mounted raid
set? What does
the ASTRO error message mean?
Anyone an idea???? Probably my mistake but hwo to aviod this in future
???
best regards Dirk
--
Milky Way - Sol System - Earth - Europe - Germany - Thuringia - Erfurt
http://wall.pdv.de/~nerle
Received on Tue Mar 10 1998 - 11:08:03 NZDT