RE: OT: cluster size=0

[chu tes-mail jpl nasa gov (Eugene Chu)]

Hi Rich,

Thanks for your reply.

We rebooted the machine today, and for some reason, the RAID is working
fine today.  However, the directory and the test file I attempted to 
create are gone.  Re-running the I/O tests, I see about 50 MB/s going
to the RAID.  Going to the boot disk is still around 3.5 MB/s.  I found
out the boot disk is an IBM 9GB drive, and peak transfer is spec'd at
only 9 MB/s, so mayb 3.5 is what I can expect in the space I have left.

The controller is listed as an adaptec 2940U2/w /7890, and the driver
listed is aic7xx.  The disks are connected to the SCSI port on the MoBo
with a short ribbon, terminated at the end.  The errors recorded in the
log file are:

Mar 21 12:44:36 137 kernel: scsi0 channel 0 : resetting for second half of retri
es. 
Mar 21 12:44:36 137 kernel: SCSI bus is being reset for host 0 channel 0. 
Mar 21 12:44:38 137 kernel: (scsi0:0:0:0) Synchronous at 80.0 Mbyte/sec, offset 
63. 
Mar 21 12:44:38 137 kernel: SCSI disk error : host 0 channel 0 id 2 lun 0 return
 code = 26030000 
Mar 21 12:44:38 137 kernel: scsidisk I/O error: dev 08:21, sector 32 
Mar 21 12:44:46 137 kernel: scsi0 channel 0 : resetting for second half of retri
es. 
Mar 21 12:44:46 137 kernel: SCSI bus is being reset for host 0 channel 0. 
Mar 21 12:44:48 137 kernel: (scsi0:0:0:0) Synchronous at 80.0 Mbyte/sec, offset 
63. 
Mar 21 12:44:48 137 kernel: SCSI disk error : host 0 channel 0 id 2 lun 0 return
 code = 26030000 

...

Mar 21 21:37:25 137 kernel: EXT2-fs error (device md(9,0)): ext2_write_inode: un
able to read inode block - inode=458754, block=917516 
Mar 21 21:37:25 137 kernel: EXT2-fs error (device md(9,0)): ext2_write_inode: un
able to read inode block - inode=458754, block=917516 


The system is running 2.2.14-6.0smp.  It may be old enough to have disk
driver patches already.  I'll have to check at RH.

In the mean time, we will run more tests, and keep an eye on it.

eugene