Solved: LMS 2.6: Can no longer find devices in Device Center

Jeff Law · ‎02-02-2010

Hi,

I recently went through the excercise of renaming our 3 CiscoWorks servers to be the same name as the server itself was named. This was because of a Jrm server problem that we started experiancing.

I thought I had completed this successully, because once done, Jrm and its dependant services started running again, and all the scheduled jobs started working.

However, it appears I was wrong. Since making these changes we have lost visibility to most of our devices through the Device Center application. All the devices (2,625 of them) are still available when looking through the Common Services menu options. However, when searching for a device in the Device Center, only 11 of these 2,625 devices are visible. I did a search using the * as a wild character and that was all that appeared in the list.

All that appears for a device not listed is the IP address, and the Device Type fileds (with correct inofrmation), and in the Functions AVAilable section, only the Tools menu.

All processes on the server, which is the master server in our environment, running CS, and RME, are running with no apparent errors EXCEPT for the process EDS-TR, which it says has Never Started.

Does anyone know why the devices no longer appear?

Regards

Jeff

Joe Clarke · ‎02-02-2010

I have a feeling you did not do the PIDM migration correctly. See step 8 at http://www.cisco.com/en/US/docs/net_mgmt/ciscoworks_common_services_software/3.0.5/user/guide/diagnos.html#wp1078582 . You must run this SQL query to update the PIDM records in the cmf database.

View solution in original post

Joe Clarke · ‎02-02-2010

I have a feeling you did not do the PIDM migration correctly. See step 8 at http://www.cisco.com/en/US/docs/net_mgmt/ciscoworks_common_services_software/3.0.5/user/guide/diagnos.html#wp1078582 . You must run this SQL query to update the PIDM records in the cmf database.

Jeff Law · ‎02-02-2010

Oops. You are correct, I didn't do that step.

Trying to do it now though using the following command after running the net stop crmdmgtd command (on all the servers):

D:\CSCOpx\log>dbisqlc -c "uid=cmfDBA;pwd=*******;eng=cmfEng;dbf=D:\CSCOpx\databases\cmf\cmf.db" -q update PIDM_app_device_map SET app_hostname='HP1813' where app_hostname='GEB'

I get a the following pop up error message:

Error at line 1

Specified database file is already in use

It could be a typing error, but I cant see it. D:\CSCOpx is my NMSROOT directory.

Looking at the services still running on the master server where I am running this command, I notice the following Cisco services are still started:

CiscoWorks Web Server

CWCS Cmf database engine

CWCS rsh/rcp service

CWCS syslog service

CWCS tftp service

Do I have to stop any of these?

Regards

Jeff

Jeff Law · ‎02-02-2010

Hmmm. If I manually stop the CWCS Cmf databas

e engine, then run the command, I get a different error message:

Error at line 1

Primary key for table 'PIDM_app_device_map' is not unique

I can Continue, Stop, or Exit. Not sure of the consequences of continuing, so Exit is what I choose...

Jeff Law · ‎02-02-2010

Ok. By poking around a bit more, I have managed to I have found that there are 4 records in this table that seem to be the cause of that error message.

I have managed, by using the isql utility and dredging my memory for SQL commands, updated all the other records so that they now refer to the correct server.

I cant see any reason why these 4 records cannot be changed, they look identical to the other records. The device_id's for these records are: 2270, 2282, 3958, and 3959.

Perhaps if I can somehow delete these through the GUI first, and rediscover them. Is there a way to get the device name?

Regards

Jeff

Jeff Law · ‎02-02-2010

Found the name of the ID by using the dcrcli utility.

Thanks for your help Joe.