Originally posted by: SystemAdmin

Hi,

I have just reconfigured LVM and deleted original VGs and added new VGs.
At my site we have implemented HACMP 5.3 cluster with 2 nodes.
I updated VG info in resource group of both nodes. But when I try to
sync resource group I get following error:
smit hacmp: Extended Conf - HACMP Verification and Synchronization (Active Cluster on a Local Node)
<hr />
cldare: Unable to execute a command remotely on node ctydbs01,
check clcomd.log file for more information.
Any configuration changes must be propagated to the node before the node
will be allowed to join the cluster.

<hr />

I'm new to HACMP and don't know where to find manpage for cldare command.

Pl. help.

-Santosh
#AIX-Forum

Originally posted by: kappa

AH, not a problem with your LVM modifications, you have changed something that is security related? => no remote exec possible!
Is clcomd running on both nodes?

And? What kind of communication problem?

--jk
#AIX-Forum

Originally posted by: SystemAdmin

Here is o/p from clcomd.log file, but not getting any clue from it.

<hr />
Sat May 3 17:18:03 2008: Connection to node ctydbs01: closed
Sat May 3 17:22:35 2008: Activating timer = file_collections_auto_prop
Sat May 3 17:32:35 2008: Activating timer = file_collections_auto_prop
Sat May 3 17:42:35 2008: Activating timer = file_collections_auto_prop
Sat May 3 17:52:35 2008: Activating timer = file_collections_auto_prop
Sat May 3 18:02:35 2008: Activating timer = file_collections_auto_prop
Sat May 3 18:12:35 2008: Activating timer = file_collections_auto_prop
Sat May 3 18:22:35 2008: Activating timer = file_collections_auto_prop
Sat May 3 18:24:28 2008: Connection Secmode = Standard
Sat May 3 18:24:28 2008: Trying to establish connection to node ctydbs01
Sat May 3 18:24:28 2008: Connection to node ctydbs01, success, 10.10.10.2->10.1
0.10.1
Sat May 3 18:24:31 2008: Connection Secmode = Standard
Sat May 3 18:24:31 2008: Connection to node ctydbs01: closed
Sat May 3 18:24:31 2008: Connection Secmode = Standard
Sat May 3 18:24:31 2008: Trying to establish connection to node ctydbs01
Sat May 3 18:24:31 2008: Connection to node ctydbs01, success, 10.10.10.2->10.1
0.10.1
Sat May 3 18:25:02 2008: Connection to node ctydbs01: closed
<hr />

Can u point out what is problem.

-Santosh
#AIX-Forum

Originally posted by: kappa

Well, clcomd seems to work.

As a starting point:
1) cllog shows you where to find various cluster logs.
2) PVs must have a PVID before HACMP can use them.
3) try to get a scheduled downtime, stop cluster services on both nodes
4) Each Cluster VG must look identical on both nodes, sync it manually!

VGs need:
Quorum off if Cross-Site, Auto off, all FS auto off, VG Major Device Nr identical (if NFS exports)
LV names must be unique on both nodes, mount points avail?, do you have Disk Heartbeat?
#AIX-Forum

Originally posted by: SystemAdmin

Hi,

Issue resolved.

The clstrmgrES process was not stopped properly
(it's status was shutting instead of inoperative),
hence killed the process from OS then only
the option of Automatically detect and correct Errors
option become visible. Then corrected errors
(resource group VG's timestamp giving error)
Then started cluster successfully.

Thanks for ur response.

-Santosh
#AIX-Forum

Originally posted by: SystemAdmin

Hi,
I have the same issue where by a file system was accidentaly created by a user without using CSPOC. I have tried to sysnch the cluster even after removing the file system but getting the error that it couldn't synchronize. Am not authorised to bring the two nodes on cluster down since they house critical applications. Is it a must that after such a case the two nodes need to be rebooted?
#AIX-Forum