From: jbarratt (jbarratt@COMPSAT.COM)
Date: Mon Jan 19 2004 - 20:08:32 EST
I had these symptoms in an HACMP cluster. check all of these things
below. the cause was #2/#3
1. are all of the hosts in seperate storage partitions? they should be
if this is not a cluster.
2. are the code levels current? the FAStTs seem to be really sensitive
to fw/SM version conflicts, verify that the levels you are running are
supported. I start with the FAStT HACMP compatability matrix and work
down from there.
3. is rdac working? is it installed? use fget_config -l dar0 to make
sure that both dac0 and dac1 look like you expect them to.
4. make sure your cabling is correct.
Hey everyone.
We are having trouble with FAStT500 controller fail over. I have
several pSeries systems SAN attached via McData ED5000 systems. These
FAST controllers seem over react when a system is rebooted that has an
array on the FAST. What happens is the pSeries system would be rebooted
and the FAST losing its connection to that system would fail Controller
A (preferred) to Controller B and do this for every system on Controller
A. Does anyone have a clue why that is happening? IBM documentation
makes some mention of this behavior, but doesn't go into much detail.
Any experience or reference on how these controllers really work would
be helpful. I have the Redbooks and other docs.
Thanks,
Ken
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:17:31 EDT