HUM Processes Down

Answered Question
Aug 31st, 2008

Dear All

Our HUM application ahs crashed by itself, the following two processes are down and I try to restart them manually they don't come back up:

ProcessName ProcessState ProcessId ProcessRC ProcessSigNo ProcessStartTime ProcessStopTime

3. UPMDbMonitor Administrator has shut down this server 0 15 0 8/31/2008 11:53:24 AM 8/31/2008 11:56:39 AM

4. UPMProcess Administrator has shut down this server 0 15 0 8/31/2008 11:53:24 AM 8/31/2008 11:56:24 AM

All the HUM portlets read:

"Unable to get Database Connection. Either UPMDbEngine process is down or Connection Pool is exhausted."

How to proceed with this issue?

CiscoWorks Common Services 3.1.1

Health and Utilization Monitor 1.0.0



I have this problem too.
0 votes
Correct Answer by Joe Clarke about 8 years 1 month ago

This command will only reinitialize the HUM data. None of the other apps will be affected.

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
Farrukh Haroon Mon, 09/01/2008 - 01:01

Thank you for your help. I tried to locate this file but its virtually empty with only this in it:


UPMDbMonitor {Sun Aug 31 11:53:24 2008} Running DbMonitor version 2.1



Joe Clarke Mon, 09/01/2008 - 04:08

Please post a list of services and their startup parameters as well as the list of contents under NMSROOT/databases/up and the output of netstat -a -n -o -b.

Joe Clarke Mon, 09/01/2008 - 18:47

You did not include the service startup parameters. Based on the rest of the rest of the output, everything looks okay. Therefore< I need to see the service startup parameters.

Farrukh Haroon Mon, 09/01/2008 - 19:03

Sorry for my stupidity, but how can I capture the service startup parameters? I'm assuming you want the status Manual/Automatic/Disabled thing available in services.msc



Joe Clarke Wed, 09/03/2008 - 04:23

This looks fine, and since there is no errors in the DbMonitor log, I cannot say what the problem with the database is. Typically, issues like this are either a problem with the database password, or the database is corrupted. This could also be an issue caused by a DLL conflict, but we typically see the CmfDbMonitor process also go down. However, if you have multiple copies of libeay32.dll and ssleay32.dll on the system (especially in C:\WINDOWS or C:\WINDOWS\system32), this can cause crashes in the DbMonitor processes.

If you rule out the DLL problem, and what to look at the other possible causes, you could first try resetting the HUM DB password:

NMSROOT\bin\perl NMSROOT\bin\ dsn=upm

If that does not work, you could try reinitializing the HUM database:

NMSROOT\bin\perl NMSROOT\bin\ dsn=upm dmprefix=UPM

Of you could contact the TAC to have them walk you through enabling DbMonitor debugging.

Farrukh Haroon Sun, 09/07/2008 - 03:49

Thank you for your help. I did a search for these two DLLs on the CiscoWorks server, and I found both in the following locations:


C:\Program Files\CSCOpx\bin

They are not present in the Windows directory etc. File size of both the above locations are same.

I also tried to reset the HUM DB password, but the same issue persists. In fact now there is a new issue, the UPM process have gone down completely. The HUM DB process available in windows service manager is also stopped now, I can't start it from either Windows or CiscoWork Process Manager (UPM processes). Please find attached new pdshow.

Please le me know how to proceed with this now. We are seriously doubting HUM's stability to monitor our network now. If it can't stay up itself, we don't except it to tell us the availablity of our network devices :)



Joe Clarke Sun, 09/07/2008 - 07:02

Now it appears the database is corrupt. If you have a good LMS backup, you can restore from that. Else, you will need to reinitialize the UPM database:

NMSROOT\bin\perl NMSROOT\bin\ dsn=upm dmprefix=UPM

Farrukh Haroon Sun, 09/07/2008 - 08:05

Thank you for your reply. Will this affect only the HUM component or all? I really don't care about the HUM data right now and am willing to restore a 20-25 day old backup, but I want all th RME/DFM/CM data to stay there, is that possible?



Correct Answer
Joe Clarke Sun, 09/07/2008 - 08:08

This command will only reinitialize the HUM data. None of the other apps will be affected.

Farrukh Haroon Wed, 09/10/2008 - 05:15

Dear Clark

Thank you for all your help. We were able to restore HUM to a fresh DB. Is there anyway to restore HUM (and not the whole CiscoWorks) to a last working bakcup of UPM DB instead of a fresh DB?




This Discussion