NFS hangs - Help!

[Ionut Grigorescu_2]

Hi,

I have a system of 2 server running HPUX11.00, MC/SG with 3 packages, one db package running normaly on one server, the other 2 packages on the other one. Both server configured as nodes in the cluster. I had a test case to prove if High Availability Cluster runs OK - it didn't :-( After cmhaltnode node1, both system and osspkg were halted on node1 and started on node2. The problem is they are still starting... and one of them was the NFS server. I can work now on node1 OK but I don't have neither telnet access nor console login to node2. With cmviewcl I can see that the osspkg and system package are still starting on node2, cmhaltpkg and cmhaltcl don't running anymore. What can I do - I don't want to reboot node2, being the database server?

thank you in advance,
ionut

[Geoff Wild]

Rebooting might be the only way...

Things to try:

Can you remsh over? probably not...

Do you have "ServiceGuard Manager" installed?
Requires MC/SG 11.13 or higher on your cluster...

You could try a cmhaltnode node2 (from node1).

If NFS is hung - does node1 mount and/or export to node2? You could try stopping NFS on node1:

/sbin/init.d/nfs.client stop
/sbin/init.d/nfs.server stop
/sbin/init.d/nfs.core stop

Otherwise - reboot - sorry.

[Michael Steele_2]

Can you attached the results of 'cmscancl' from either node.

Also check for errors in /etc/opt/resmon/log/api.log, especially if you've installed the special EMS SG module.

Also attach your ioscan -fnk report.

Since your testing why not TOC the box?

[Ionut Grigorescu_2]

Hi,

I would TOC it but I have no backup for the database and since I don't have experience with Oracle, I don't know if the database will be OK after a TOC

[Ionut Grigorescu_2]

Thank you for your help - I had to TOC the box and now is OK. Case closed.