Description of problem: We have 19 IMAP servers supporting 86000 faculty and
staff. We are getting intermitteng hangs that require a reboot.


Version-Release number of selected component (if applicable):
 Red Hat Enterprise Linux AS release 3 (Taroon Update 7)

How reproducible: We have a test system where we induce disk io load and
quota updates, which has an MTBF under 20 minutes; under 10 minutes is
common.


Steps to Reproduce:
1.Run program simulating high multiprocess disk load: 'stress_jnl.py'
2.Run program updating quotas: 'qtwiddle2_jnl -n 20000 -s 1'
3.Wait for hang
  
Actual results:
The system hangs, i.e. accepts no console input other than SysRq.


Expected results:
The system would continue to run.


Additional info:
We have tried the patch mentioned in bug # 122252 but still hang.
We can provide account on test box if needed.