View Incident:
http://co-op.engr.sgi.com/BugWorks/code/bwxquery.cgi?search=Search&wlong=1&view_type=Bug&wi=821124
Status : closed Priority : 2
Assigned Engineer : lord Submitter : ananth
*Modified User : lord *Modified User Domain : sgi.com
*Description :
This problem seen with XFS 1.0 bits on a 64 M, 2CPU system with
1 scsi disk. Tests being run in parallel:
while true; do holes -l 300000000 -b 4099 -i 7 -c 7 /xfs/holes; date; done
and AIM 7, run to crossover.
It took about 3-4 hours to hit this problem, by then
AIM had gone to about 170 processes. In the end,
console became unresponsive ... looking at time-stamps
.....
==========================
ADDITIONAL INFORMATION (ADD)
From: lord@xxxxxxx (BugWorks)
Date: Apr 13 2001 01:09:27PM
==========================
Ananth's latest hang is actually a different form of deadlock,
the fix is valid and does remove the deadlock. This particular
hang is a case of over agressive metadata readahead - in the
case of low memory the readahead should be giving up rather than
attempting to do the readahead. I have a fix running, but not
sure yet if it is the correct thing to do, dbench load is really
clogging up the system now.
|