how to Perform periodic maintenance of the CUCM as per the Industry best practices.

[Chowdary PSRK]

hi, 

I would like to know how Perform periodic maintenance of the CUCM 10.5.1 as per the Industry  best practices. 

Please, advise.

TIA

Regards,

Chowdary

[Manish Gogna]

Hi Chowdary,

There is no special rule or periodic maintenance manual for cucm. You need to understand the existing setup of cucm and its integration / interaction with other applications and endpoints and ensure that there are no compatibility issues to start with.

http://www.cisco.com/c/dam/en/us/td/docs/voice_ip_comm/uc_system/unified/communications/system/Compatibility/CSR-Compatibility-Matrix-InteractiveHTML.html

Check for the latest firmware release notes for endpoints and see if any of the bugs fixed in that release might be impacting your setup, if yes upgrade to that firmware.

Same goes with any Service update release for the main version that you are on. If there are no compatibility issues with upgrading to 10.5.2 release of cucm then you should consider doing it  as well.

Keep monitoring the RTMT for any high severity alerts as well.

Apart from these regular checks there are some customers who perform PEN testing for vulnerabilities, you may engage your network team for the same.

HTH

Manish

[Bill Mungaven]

I like to reboot my entire 9.1.2 CUCM cluster (1 Publisher, 6 Subscribers, 2 music on hold, 2 TFTP) every 3 months or so. Although I didn't see this in any formal Cisco documentation, in talking with a TAC engineer on a problem I was having, he suggested the 3-month reboot cycle to clean up temporary files, etc. If I don't reboot every 3 months or so, I start to notice unusual, unexplained strange behaviors of my CUCM. After the reboot, things start working properly again. The strange behaviors I saw before the reboot are cleared up after the reboot.

Just my opinion, but it works for my environment.

Bill

[Manish Gogna]

Hi Bill,

If your setup allows a cluster reboot on weekends or in a maintenance windows without impacting production then this approach is okay. However, there is no recommendation as such for a periodic cluster reboot after a certain time has elapsed. I would suggest restarting/resetting  the respective services / devices / trunks to start with, like for issues with DRS backup restart the DRF MA/LA services, for webpages issue restart Cisco Tomcat service and so on. 

HTH

Manish