topic Re: fail to cmquerycl in Operating System - Linux

fail to cmquerycl

Ricardo Rocha — Tue, 07 Feb 2006 11:40:28 GMT

Hi

I'm having problems in configuring a ServiceGuard cluster with RH EL3. When running cmquerycl i get:
Begin checking the nodes...
Looking for other clusters ... Done
Error: Node ermesinde40 is refusing Serviceguard communication.
Please make sure that the proper security access is configured on node
ermesinde40 through either file-based access (pre-A.11.16 version) or role-based
access (version A.11.16 or higher) and/or that the host name lookup
on node ermesinde40 resolves the IP address correctly.
Failed to gather configuration information.

DNS is ok, and both nodes are in /etc/hosts.
I've also configured cmclnodelist and /root/.rhosts files.

If I run this command with only the local node, it ends ok. It fails to gather the information on the other side.

I'm running out of ideas...surely someone has other ones.

Thanks,

Ricardo

Re: fail to cmquerycl

Steven E. Protter — Tue, 07 Feb 2006 12:00:55 GMT

Shalom Ricardo,

Serviceguard has a cmnodelist file in /etc/cmcluster that supersedes and works better than .rhosts.

I suggest you look into that and test the connectivity with rlogin if you wish to continue using rhosts.

SEP

Re: fail to cmquerycl

Ricardo Rocha — Tue, 07 Feb 2006 12:06:25 GMT

Thanks for the quick reply...

I've removed .rhosts file and I'm currently working with cmclnodelist. The sytax I'm using is:
nodeA root
nodeB root

But the problem remains...

Re: fail to cmquerycl

Jeff_Traigle — Tue, 07 Feb 2006 15:20:35 GMT

Glancing through the cmquerycl man page on HP-UX, I notice the -n option requires the node_name must match the node name returned by the hostname command. Maybe you have a mismatch there that's causing a problem even though your DNS and /etc/hosts is ok... maybe...

Also might be worth checking the permissions on cmclnodelist. Ours are set up as 640. Can't recall if SG cares about this, but I know some other security-conscious applications will misbehave if permissions aren't tight enough.

Re: fail to cmquerycl

melvyn burnard — Tue, 07 Feb 2006 18:00:58 GMT

which version of Serviceguard are you using?
Also, what hostnames are set for each lan on both nodes?
and finally list the contents of your cmclnodelist file on both servers

Re: fail to cmquerycl

Ricardo Rocha — Wed, 08 Feb 2006 05:15:41 GMT

Hello all
Thanks for your help.
My two nodes are called "ermesinde39" and "ermesinde40" (output from hostname command).
The contents of /usr/local/cmcluster/conf/cmclnodelist are:
ermesinde39 root
ermesinde40 root
The file has 640 permissions and owned by root.
The /etc/hosts have both FQDN and hostname.
The ServiceGuard is serviceguard-A.11.16.04-0 for RedHat EL3.

Almost starting to pull my hair out!

Thanks,

Ricardo

Re: fail to cmquerycl

Serviceguard for Linux — Thu, 09 Feb 2006 00:37:51 GMT

Common causes:

# Sometimes this means that the firewall is set up incorrectly. Try turning off the firewall (at least temporarily).

# Also, identd may not have started. "identd" is the service associated with pidentd (see #2). Make sure this is set to start automatically. (I thinnk this is currently most common.

# Check that the file cmclnodelist has been created and is set up correctly. (For test purposes try "+ +"

Double check that you can ssh both ways.

Reboot both nodes.

Re: fail to cmquerycl

Ricardo Rocha — Fri, 10 Feb 2006 07:06:44 GMT

Bingo!!!
Damm firewall! Enabling the connection between the nodes solved it.
Thank you all!

Ricardo