Background: LMS 3.1 ( CS 3.2,CM 5.1.3,RME 4.2,portal 1.1 )
OS: Windows 2003 SP2.
Had windows patches installed one week earlier.
Remoted (RDP) into server for daily checks and found CPU at 100%
This has happened before when our discovery took to long. Assuming the same issue, I restarted the services.
As they came up, they took several minutes longer than normal and seemed to be stalled when Tomcat started.
All eventually show as started except for apache.
I checked the event viewer for CASUSER which had successful audits.
My current sll dll's are dated 1/18/2008
I checked all the user rights according to a previous post.
I added CASUSER to the administrators group.
and finally thinking it may have bee a corrupted databse I did a restore and still have the same symptom.
Symptom: It looks like Tomcat isn't coming up all the way which never starts Tomcat monitor and therefore no apache...
I will appologize in advance for slow responses as I am dealing with a completely isolated network and may have to recreate
sections of output that you ask for. I attached pdreg-l appache and pdshow....
Request: Any suggestions are appreciated...... I also recently learned about the points system so I will try to vote on my previous threads.
You did not post your pdshow or pdreg output. You should also post screenshots of the Services control panel showing all of the CiscoWorks services.
Post the NMSROOT/MDC/tomcat/logs/stdout.log and stderr.log. What other software do you have installed on this server?
The files are attached...
stdout was empty 0k
Applications on the server beside CiscoWorks:
BMC patrol (AGENT)
Symantec Client Security (Antivirus)
Arcsight smart agent
When first troubleshooting, I did notice that the antivirus was scanning the CSCOpx directories. This had previously been
excluded and is again but I don't know if it could have done any damage.
I appologize for not being able to provide the file you need.
I am not able to move files off the system. Would it benefit me to compare it to the same file on an existing working server?
or provide a particular section of the file?
I understand if you are unable to go any further without this data. Given that TAC will probably want the same files,
do you think I am better off uninstalling LMS, reinstalling and trying to restore a backup?
Thanks for all the help!
It is important to see the entire regdaemon.xml file as it exists on the server. If that file is corrupt, then restoring a backup which contains it would simply reintroduce the same problem.
Given what I have seen thus far, I cannot say what is causing your problem. A reinstall may or may not fix it. It could be due to a corrupt regdaemon.xml, or it might be some problem with anti-virus interaction. There could also be some corruption of the Tomcat config files. It would be easy to verify the former and the latter if you could provide the files. The issue with AV could be corrected by a reininstall provided AV was disabled, then on-access scanning remained disabled for NMSROOT.
Of course, this may be something else entirely. The only thing that's clear is that Tomcat is not starting properly, and appears to be hanging (or maybe looping) trying to load the Cisco components of LMS. It may boil down to requiring remote access to look at the problem in real time.
I sooo wish I could allow remote access....
I will see what options I have for providing the files... I can tell you that the regdeamon file appears different from
a June regdaemon.xml.bak in the same folder as the order of the
Would it hurt to copy the older regdaemon.xml.bak of the existing one ?
As far as the Anti Virus, are you saying I would need to reinstall CiscoWorks?
Without validating the current regdaemon.xml, I cannot recommend doing anything. The current file may be fine, in which case restoring the old file could break something else. If the problem is that the AV software has somehow broken the LMS jar files, or messed with the Tomcat config, then a reinstall would correct that.
It sounds like circumstances are going to force me to reinstall. I guess at that point I could try the old regdaemon.xml file first.
Will I have to do a complete uninstall / install ? And if so will I be able to restore a backup after that?
I cannot answer these questions with certainty. If you're going to reinstall, you should reinstall from scratch. To be 100% safe, you should obtain the clean_system.exe tool from TAC which will make sure the system is pristine before doing the reinstall.
The backup MAY work provided there is no corruption in the LMS data (and I cannot say for certain if corruption exists now). My advice is to do the clean reinstall. When you have it installed, test it before doing any restore. Make sure you can login. Try adding one device, and make sure things work. Then, take a backup of LMS at that state. Then try the restore. If things are still working, go forth in peace. If not, you should be able to restore the pristine backup, and start over (without doing another reinstall).
I did try taking a regdaemon.xml file form a similarly configured box (paths/os..etc), updated the unique fields, (server name, ACS key..etc)
and starting the services with the same result. I will take one last shot at getting the regdaemon.xml file to you but will also start planning
for a reinstall. again appreciate all the help....I'll keep you posted.
You can't do this. The regdaemon.xml file is VERY server-specific. Each one is unique to the server from which it came. Even if the new file was not corrupt, restoring it to a new server would break other things.
Well, I did not win the file battle so I ended up doing a clean install.
Everything installed fine except I can not seem to point it back to the ACS server.
I can see packets passing between LMS and ACS on ports 49 and 2002 but
it just hangs at "verifying credentials".
didn't know if you had any thoughts before I started a new thread.
I don't know all of the specific details that go on when Integrating with ACS but it appears
that the when LMS is integrated with ACS, TOMCAT needs to negotiate with it somehow
when it starts up.
The syptom was that the LMS Apache web service would not start.
The TOMCAT services started but did not seem to complete everything as it only used a fraction of the memory.
The fix action was the disabling of three Windows Local Policy security settings under
------ Administrative Templates
--------- Windows Components
----------Internet Explorer: Security Zones: Use only Machine settings
Security Zones: Do not allow users to change policies
Security Zones: Do not allow users to add/delete sites
I don't know the underlying cause/fix action but that did it.