LMS 3.1 jrm issue

Unanswered Question
Jun 2nd, 2009

LMS 3.1 jrm process state is "Waiting to initialise". I have tried to stop and start crmdmgtd but no change. All scheduled tasks such as discovery, config archive, inventory collection are not running as a result. Where should I start?

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
mburrell5 Tue, 06/02/2009 - 22:01

the jrm.log file was about 1.23GB in size. I have had to rename it and create a new file. it may not be much use but here it is anyway. I restarted the jrm process after creating the file. the process is now sitting in the "waiting to initialise" state

Joe Clarke Tue, 06/02/2009 - 22:09

Actually, the older log might be more useful. Was the hostname changed on this server? Do you have a firewall on this server? Post the NMSROOT/lib/classpath/jrmuser.properties.

mburrell5 Tue, 06/02/2009 - 22:51

yes the hostname was changed about 6 months ago because it had an underscore in it (at the time of upgrade to LMS 3.1). No firewall on the server. see attached file as requested plus update of jrm.log (still in waiting to initialise state)

Joe Clarke Tue, 06/02/2009 - 23:23

This jrm.log is the same as the first one. Is there a newer log?

Joe Clarke Wed, 06/03/2009 - 15:27

There does appear to be a communication problem with Daemon Manager. Post a screenshot of your Services control panel showing all of the CiscoWorks services as well as the output of the pdshow command.

Joe Clarke Wed, 06/03/2009 - 15:57

Did you run pdshow multiple times? If this is the output of one command, your installation is completely hosed.

Joe Clarke Wed, 06/03/2009 - 18:03

I mean your installation is severely damaged. The pdshow is listing multiple instances of the same daemons. For example, I'm seeing five instances of jrm. This should not by possible.

Try rebooting, then collect the output of a fresh pdshow. If you're redirecting the output to a file, make sure you do not append the output.

mburrell5 Wed, 06/03/2009 - 18:31

Obviously i was appending the output to the file. Sorry about that. the correct pdshow is attached. I haven't rebooted yet

Joe Clarke Wed, 06/03/2009 - 18:47

This looks better. It appears that jrm might be fine, but the server may be overloaded. It's taking 12 minutes for jrm to initialize. Jrm needs to start within five minutes. What are the specs on this server? what is the current CPU load?

mburrell5 Wed, 06/03/2009 - 18:58

I was afraid it might come to this. the server specs are as recommended except it is a single CPU not dual. CPU load varies between 50-100% with an average of about 75%.

Joe Clarke Wed, 06/03/2009 - 19:06

This is most likely the problem. There are not enough CPU cycles to service the daemons in a timely fashion. Adding another CPU, removing other applications, or reducing the number of managed devices may help this.

mburrell5 Wed, 06/03/2009 - 19:13

Ok, fair enough. I am just curious why it was working until about 4-6 weeks ago and then stopped. thanks again for your help

Joe Clarke Wed, 06/03/2009 - 19:29

The number of jobs on the system could have tipped the balance. That assumes that CPU utilization has been constant, and nothing else has changed. If you can connect to the Common Services job browser, you can try deleting some of the old jobs, and see if that helps.


This Discussion