topic Re: Abort message when we run cmruncl command in Operating System - HP-UX

Abort message when we run cmruncl command

shameemsoft — Sun, 16 Jan 2011 05:17:11 GMT

Hi,

I tried to start cluster in our HP UNIX 11iv3 host. it is metro cluster. but it is failed and find the error message below.

#cmruncl -v
cmruncl: Validating network configuration...
Gathering network information
Beginning network probing (this may take a while)
Completed network probing
Abort(coredump)

it was working before. server was crashed due to network issue. boot area is correpted. we booted from DRD and recovered the disk. now OS is working fine. but we are facing issue with startup the cluster.

please help me on this.

Regards
Shameem

Re: Abort message when we run cmruncl command

Manix — Sun, 16 Jan 2011 09:22:16 GMT

>Completed network probing
>Abort(coredump)

Looks "cmcld" is giving core dump which can be analyzed by HP.

As you said there was network panic & it`s stopping at the same point it`s better to
check the inter node connectivity manually
between the lan cards mentioned in the cluster.conf.ascii file.

You may use linkloop for the same.

See the detailed errors in syslog.log file are
there some reasons of failures

Re: Abort message when we run cmruncl command

nijokj — Mon, 17 Jan 2011 07:44:32 GMT

Hi Shameem,

Have you tried to start the cluster on single node, if not please try to start the cluster on single node after that you can try to join other systems to cluster.

try to start cluster in single node

cmruncl -v node1

if it is started run,

cmrunnode -v node2

also see syslog.log for further logs and information.can you post cmviewcl -v output.

Re: Abort message when we run cmruncl command

Stephen Doud — Mon, 17 Jan 2011 11:54:50 GMT

A SG abort may be an indication that an unexpected configuration caused Serviceguard to fail to accept the content of the cluster binary file. Depending on the version of Serviceguard installed, use either of the two commands to inspect and compare the network configuration stored in the cluster binary file against the output of 'ioscan -fn | grep lan', 'lanscan' and 'netstat -in':
pre-A.11.17: cmviewconf
A.11.17 and later: cmviewcl -v -f line

I suspect that the network configuration may have changed as a result of the recovery, and the different configuration may be blowing up Serviceguard.