Cisco Prime 2.2 Infrastructure: Connectivity Problem

[adam.kirton1]

Hello,

We are having a bit of trouble with our Prime 2.2 (Note: our Prime is installed on a Virtual Server). In our Network inventory, it claims we have over 31000 devices when we actually only have around 500 physical devices. The other 30500 devices are stated to be unmanaged in the administration tab. Because of the high number of the devices, the server starts to become unresponsive and uses all the CPU power we assigned to it. So we tried to delete them from the list all in one go and from here, we are getting tonnes of issues. Our GUI started to crash and timeout when we want to connect to it.

We have restarted our services with the cli command "ncs stop" then "ncs start". When we run these commands, prime becomes responsive for 5-10 min, then instantly crashes and unable to connect. We have restarted the whole server from powered down, still not working. Our last alternative is to remake cisco prime from scratch and we do not want to do this. 

I have attached a screenshot of the version we have currently installed. Information been taken from a SSH connection to the server.

If you would like to help and need more screenshots, reply back!

Thanks,

Adam Kirton

[Marvin Rhoads · Accepted solution]

It sounds like your database is corrupted. (PI uses Oracle "under the covers".)

A TAC case would be the best route to resolve this.

[AFROJ AHMAD · Accepted solution]

Hi Adam,

If it was working at a point  and you have a backup of the prime around the same time then I would advise

to reinitialize the database and restore the good known backup.

Thanks-

Afroz

[adam.kirton1]

Thank you for the help. Sadly we had to restart the Prime from scratch. We all came to the conclusion and now starting to re-add all the devices. We did think about the back-up's but if we used it, it could probably cause faults in the future. 

And again, Thanks for the help!

Adam Kirton

RCBC

[AFROJ AHMAD]

Hi Adam,

you have already rebuild the sever  but I would advised to keep the prime backup job scheduled (weekly or at least monthly  ) and it is always wroth trying to restore the backup in case of database issues.

also I would suggest to export the devices list ,once you have all the devices in place  that will save a lot of task in such scenarios where you need to manually add the device or do a network discovery.

Thanks-