CmfDbMonitor Waiting to initialize with Ciscoworks 3.0.1

Answered Question
Jan 17th, 2008
User Badges:

I am trying to get Ciscoworks 3.0.1 running on an HP Proliant server running Server 2003 R2 Enterprise with SP1. I managed to get 3.0 running on it before I did the upgrade, but had to manually start some of the processes before it would run. I have attached the output of pdshow. Is there a list of Windows updates that either have or have not to be installed, and is the package known to clash with systems, as there are a number of HP tools running on the server?

.NET 2 is not installed - I saw one post that mentioned the service pack for it.

We do have a TAC case currently running for which I have sent the content of several log files. The server only has 1 nic enabled apart from the one used for ILO. Any suggestions please!



Attachment: 
Correct Answer by Joe Clarke about 9 years 4 months ago

Yes, the bug is requesting the same fix that was applied to Apache. That is, dbmonitor.exe will be linked to cwlibeay32.dll to avoid the conflict.

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (2 ratings)
Loading.
Joe Clarke Thu, 01/17/2008 - 13:16
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

A database monitor process in a waiting to initialize state typically points to a bad database password. Please post the CmfDbMonitor.log and UPMDbMonitor.log

Joe Clarke Thu, 01/17/2008 - 14:11
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

Try running:


NMSROOT\bin\perl NMSROOT\bin\dbpasswd.pl dsn=cmf


See if you can successfully change the cmf database password.

derek.sizeland Thu, 01/17/2008 - 14:56
User Badges:

Joe,


It let me change the password. I set it to what I chose for all the system passwords.

I have attached the output.

CmfDbMonitor is still showing as waiting to initialize!



Attachment: 
Joe Clarke Thu, 01/17/2008 - 15:14
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

Once pdshow returns, if you pdterm CmfDbMonitor, then pdexec it, does it come up?

derek.sizeland Thu, 01/17/2008 - 15:28
User Badges:

No change:


D:\Program Files\CSCOpx>pdterm CmfDbMonitor


D:\Program Files\CSCOpx>pdexec CmfDbMonitor


D:\Program Files\CSCOpx>pdshow CmfDbMonitor

Process= CmfDbMonitor

State = Waiting to initialize.

Pid = 7848

RC = 0

Signo = 0

Start = 17/01/2008 23:22:00

Stop = Not applicable

Core = Not applicable

Info = Server started by admin request


It is now 23:30 here, if you can post any more suggestions, I will pick them up in the morning.

Joe Clarke Thu, 01/17/2008 - 22:36
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

What are your service startup parameters for all of the LMS services? What other applications and utilities do you have installed on this server? What processes did you have to start manually when you had LMS 3.0 installed?

derek.sizeland Fri, 01/18/2008 - 05:26
User Badges:

All LMS Services are set as manual, only the Daemon is set to automatic.

I have attached a list of all the packages and patches installed on the server along with another pdshow.

Can't remember the finer details of how I got 3.0 running, I tried lots of different things, but I seem to remember manually starting DCR and jrm.



Attachment: 
Joe Clarke Fri, 01/18/2008 - 10:21
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

Try disabling Symantec AV, then restart dmgtd. Also, please post the outputs of netstat -a -n -o -b when dmgtd is running and when it is not.

derek.sizeland Fri, 01/18/2008 - 14:07
User Badges:

Already turned the Symantec services off.

I managed to get the server up for a while this afternoon by starting processes in CS/Process Management, but very flakey.

Currently, I have got it almost up, but UPMDbMonitor and DCRServer won't start. CmfDbMonitor ran for a while and then stopped.

Contents of 2 netstats attached along with another pdshow.



Attachment: 
Joe Clarke Fri, 01/18/2008 - 14:13
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

Since the processes have now failed, post the new DbMonitor.log files.

derek.sizeland Fri, 01/18/2008 - 14:56
User Badges:

no change in what is in the log file. However, there is now an entry in the system application event log:


Faulting application dbmonitor.exe, version 0.0.0.0, faulting module libeay32.dll, version 0.9.7.3, fault address 0x0001501f.


Source is: Application Error, Event ID: 1000


There is also an error in the system event log:


Generate Activation Context failed for D:\PROGRA~1\CSCOpx\lib\vbroker\bin\MFC80.DLL. Reference error message: The referenced assembly is not installed on your system.


Source is: SidebySide, Event ID is:59


Joe Clarke Fri, 01/18/2008 - 15:07
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

Wow. I have never seen this error with dbmonitor. Do a search on your system for libeay32.dll and ssleay32.dll. Where are your copies? What is the value of your Path environment variable?

Joe Clarke Sat, 01/19/2008 - 11:30
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

The problems are the two libraries under C:\WINDOWS\system32. If you move them out of the way, this problem should go away, but it may break another application that depends on them. Unfortunately, that will be the only workaround for now.

Joe Clarke Sat, 01/19/2008 - 11:31
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

I should also add that if the Legato backup software is running, you may see problems with the Legato versions. You would need to keep Legato from running as well. If after moving the system32 DLLs out of the way you still see a problem, disable the Legato client, then reboot.

derek.sizeland Sat, 01/19/2008 - 12:44
User Badges:

I have renamed the offending .dlls to .dll.old, I didn't get any errors, so I presume they weren't in use. I now have 64 processes running!

All processes that should start at boot now have done so.

I will speak to my server guys on Monday and see if there any potential issues.

We have an SR running with EMEA TAC, I will get it updated on Monday as well.

Thanks for your help, will a long term fix for this issue be looked at?

Correct Answer
Joe Clarke Sat, 01/19/2008 - 12:49
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

Yes, the bug is requesting the same fix that was applied to Apache. That is, dbmonitor.exe will be linked to cwlibeay32.dll to avoid the conflict.

Actions

This Discussion