From: Greg Rudd (g.rudd@isu.usyd.edu.au)
Date: Thu May 15 2003 - 05:34:30 EDT
Hi all
I am noticing the following strange behavior when trying to move volumes in a
advfs file domain from one disk to another on a hsg80 I seem to notice that
the amount of IO from disks usining monitor will go upto around 30MB/S and
then the machine will just hang up.
Is this a case of the KGPSA trying to drive the IO at a harder rate than what
the HSG80 can deliver which is 20mb persecond on the backplane of the
controler to the disks, hence causing the machine to lock up. If this is the
case is there any kernel/hsg settings that I can change to prevent this from
happening. Having looked at the IO subsystem all the settings there are
still the default. Whould this behavior also have somthing to do with the
aiotask_max_num setting in the RT subsystem
BTW I have also noticed this behavior when moving data from both a hsv110 and
hsg80 to a hsz70 also on a ES40 running v5.1a
Thanks in advance
-greg
Hardware & OS
Alphaserver ES40 833mhz with KGPSA card
OS: Tru64 version 5.1a pk4
HSG config presented to ES40
Raid set consisting of 6x72gb drives configured as a RAID 5 set split into 3
partitions of 50, 150,150
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:49:18 EDT