LMS2.6/RME: ConfigMgmtSever unable to connect

Unanswered Question
Sep 23rd, 2009


Job archive doesn't work. It use to work before.

We have this message:

CM0051 Unable to connect to ConfigMgmtSever process Cause: URN_NOT_FOUND : urn "Archive Mgmt" : Not found !! Action: Please restart ConfigMgmtSever process."

pdshow seems to be ok. But, I am afraid because there are few files on log folder:

bash-2.03$ pwd


bash-2.03$ l

total 3388

2 drwxrwxr-x 5 casuser casusers 512 Apr 10 2008 .

2 drwxrwxr-x 29 casuser casusers 1024 Sep 23 12:44 ..

2 drwxr-x--- 2 casuser casusers 512 Dec 7 2006 conf

1360 -rw-r----- 1 casuser casusers 687588 Sep 21 12:48 ctm.log

2016 -rw-r----- 1 casuser casusers 1024000 Apr 10 2008 ctm.log.1

2 drwxrwxr-x 17 casuser casusers 512 Dec 5 2006 dfmLogs

2 drwxr-xr-x 2 root other 512 Dec 5 2006 dfmdb

2 -rw-r--r-- 1 root other 50 Dec 7 2006 schema.log

0 -rw-r----- 1 casuser casusers 0 Dec 5 2006 socketmessages.log


Operating system is solaris 5.8.

Cu has reboot, no change.

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Joe Clarke Wed, 09/23/2009 - 07:22

The log directory for Solaris is /var/adm/CSCOpx/log. Post the daemons.log and dcmaservice.log.

Joe Clarke Wed, 09/23/2009 - 10:32

I don't see any errors here that would account for ConfigMgmtServer going down. As of 14:30 local time, it was working. Where are you seeing the error?

ebezombes Wed, 09/23/2009 - 10:44

Here you can find screenshot from the issue.

May be information on logs are too recent compare to the last job fail ?

Joe Clarke Wed, 09/23/2009 - 10:46

Yes. The job in question ran 12 days ago. The problem is no longer present now. If it happens again, you need to capture the logs immediately.

ebezombes Thu, 09/24/2009 - 00:40

The daily job has stopped since 09/20. In fact the job is hang since 09/20 see screenshot3.jpg.

How can we stop this job ?

Joe Clarke Thu, 09/24/2009 - 07:09

There is no screenshot attached. However, if the job is really running, pdshow should show a process with the job ID as its name. If you pdterm that process, the job will stop.

ebezombes Thu, 10/01/2009 - 05:54


You can find dcmaservice.log and a screenshot of joblist.

Job 4303 has worked on 09/25, 09/26 but failed on 09/27, 09/28, 09/29.

Joe Clarke Thu, 10/01/2009 - 11:04

I don't see any errors in the dcmaservice.log. What are the failure details of job 4303?

ebezombes Thu, 10/01/2009 - 22:46

4303.1 and 4303.2 are successfull.

But, 4303.3, 4303.4 and 4303.5 failed and took more than one day.

If the customer stop and restart LMS, then jobs work for 2 or 3 days and then fail.

It happens manytime on Sunday, maybe there are a lot of jobs ?

Joe Clarke Fri, 10/02/2009 - 08:39

What are the details of the failure? Yes, the job failed, but why? If you click on the job ID, why does it say the devices within the job failed?

ebezombes Mon, 10/05/2009 - 04:34


You can find error message in file.

Fetch is 120 secondes.

When customer try with sync archive, it works.

Joe Clarke Mon, 10/05/2009 - 07:30

This is just the job list. I don't see any details for the failed jobs.

Joe Clarke Mon, 10/05/2009 - 11:01

I see a huge gap in the dcmaservice.log when this job should have run. It makes me think the server is crashing, but I see no other indications of that. The next time the job fails, before doing anything else with RME, open a TAC service request so this problem can be looked at in real time. A thread dump of ConfigMgmtServer may be required to figure out why it is not running the job.


This Discussion