Gidday all,
We have a cluster of a pair of DS10s running 4.0F service pack 4. The ASE is
a pair of LSM mirrored disks. They are currently not in production but are
being primed for some serious work. During testing we have installed some
ksh scripts to ensure that some of the processing happens as a daemon.
Prior to starting these daemons, the cluster could easily be failed between
machines. Once the script is started the cluster fails to fail manually. It
fails successfully (is that an oxymoron) if the machine is rebooted.
There is nothing obvious in the script that I can see that has anything on
the shared disks held open which might cause this problem. The only other
thing I can think of is that it is due to using the asemgr line command to
determine the current status of the ASE. The script does this to ensure it
only processes when there are some disks to process to.
The portion of the script I suspect is;
---snip---
clusternode=`asemgr -d -v | grep orion-prod | cut -f 3 -d " " | sed
s/\-alt//`;
---snip---
This line determines the node name of the system the ASE is currently on.
Any thoughts? Any thou shalt nots? Any boy you really missed the boat??
all will be generously accepted
Thanks
Wayne Blom
System Specialist
Technical Development Healthcare
F H Faulding & Co Limited
Ground Floor
1 Station St
Hindmarsh SA 5007
Ph: +61 8 8241 8334
FAX: +61 8 8241 8357
Mobile: +61 0419 808 496
Email: wayne.blom_at_au.faulding.com
Received on Thu Sep 07 2000 - 05:04:33 NZST