How often to restart UCCX services?

[Sergio Fare]

Hi,

Is there Cisco documentation on how often services should be restarted, or the entire cluster be restarted? The reason why I ask is that we have services running for 588 days since the deployment of our UCCX cluster. I have restarted various services here and there to resolve certain issues, but have not restarted the whole cluster and I think it's due time. I'm a little nervous about it, to be quite honest. Any insight?

Thank you very much.

[Chris Deren]

Just like every software, the idea is to never restart it unless you have issues. If you feel something is not running well, you can restart it to see if it clears, but no software vendor will tell you to proactively reboot their system as that would imply unresolved bugs and scare customers away :-)

Chris

[david.macias]

Wow, that's an impressive uptime, I'm of the mentality that a restart every few months does good for the soul.

david

[Tanner Ezell]

I'd agree, especially if you are running software earlier than 8.5SU4; a periodic reboot helps clear out any memory issues.

Tanner Ezell
www.ctilogic.com

[Arundeep Nagaraj]

Hi Sergio,

I would again stand by the fact that a reboot is not necessary until a problem has been identified and the root cause is known to be resolved by the reboot.

Well, if you ask me the procedure to reboot a cluster in a HA environment, then

(Assuming your UCCX Version is 8.x and above)

++ Reboot Node 1

++ Wait for all the services to come up. You can monitor by the CLI "utils service list".

++ Once they are STARTED, reboot node 2.

++ Wait for the all the services to come up

++ Check for the services under Cisco Unified CCX Serviceability -- Tools -- Control Center Network Services

++ Also good idea, to check the replication from Cisco Unified CCX Serviceability -- Tools -- Datastore Control Center -- Replication Servers. The status for both the nodes should be 'Connected' and 'Active'

Also, a good idea would be to use the release notes of the UCCX version that you are running on and find if there are any known issues which involved a reboot as the solution.

Regards,

Arundeep