clcomd Problem! 1

[khalidaaa]

Hi AIXers,

I have a problem with my test cluster! I've been away form work for sometime and when i came back i found out that i can't start my cluster!

That's the output of smitty clstart

Command: failed        stdout: yes           stderr: no

Before command completion, additional instructions may appear below.

connect: Connection refused
clhaver[109]: cl_socket(s1test1): Connection refused
connect: Connection refused
clhaver[109]: cl_socket(s1test2): Connection refused

Verifying Cluster Configuration Prior to Starting Cluster Services.

ERROR: Cannot find node(s):  -C.

Please ensure all requested nodes are online.
There are no active cluster nodes to verify against.
Verifying node(s):  requested to start

When i tried to view the the cluster processes (clcomd and clstrmgr) It appears that only clstrmgr is active!

I tired alot of things from rebooting the system to changing the /usr/es/sbin/cluster/etc/rhosts but with no luck!

the two node-cluster is running on a VIO.

Any idea how can i get my cluster back?

Thanks in advance

Regards,
Khalid

 

[khalidaaa]

and by the way, this clcomd problem appears in one node! the other node is running the clcomd with no problems!

 

[cts123]

I would suggest, check all your setting with respect to HA for both the nodes again and co-relate.

And do "Verify and Synchronize HACMP Configuration"

If you see you are set, then do smitty clstart. It should work. If you see any error in your hacmp.out file, take action accordingly.

Regards,

-Sam

 

[khalidaaa]

Hi Sam,

I tried to sync the cluster but It doesn't synchronize for some reason!

That's what i get when i sync:

Command: failed        stdout: yes           stderr: no

Before command completion, additional instructions may appear below.

ERROR: You must start the clcomdES subsystem.
Please execute /usr/bin/startsrc -s clcomdES

Oh My God! With the above error message, I used to do this previously but it didn't work but now it is working for some reason! I guess that happened when i copied /usr/es/sbin/cluster/etc/config/clvg_config from the other node to here!

I managed to synchronize now but i got so many error messages that i don't know where to start to correct!

I will try to figure out the errors and i will let you know. (I managed to start the cluster on one node any way!)

Thanks for tip though A star deserved.

Regards,
Khalid

 

[cts123]

Just for your reference ..I am sending the output of my cluster process status.
I have HACMP 5.4

/usr/es/sbin/cluster/etc > ps -ef | grep clu

root 204936 176258 0 Mar 14 - 3:16 /usr/es/sbin/cluster/clcomd -d
root 1110236 176258 0 Apr 13 - 8:34 /usr/es/sbin/cluster/clstrmgr
root 1167484 176258 0 Apr 13 - 2:18 /usr/es/sbin/cluster/clinfo


/usr/es/sbin/cluster/etc > lssrc -a | grep -i cl

clcomdES clcomdES 204936 active
clstrmgrES cluster 1110236 active
gsclvmd 864258 active
clinfoES cluster 1167484 active

Regards,

-Sam

 

[khalidaaa]

Just to close the problem, I had to recreate the persistence IPs on the cluster and now every thing back to normal!

Thanks Sam for your comments

Regards,
Khalid