DC Server Failed

Naveen Raj Janardhan · ‎03-30-2009

Hi All,

We use CallManager 4.2 cluster in our organization. A few days back, I got several emails generated by RTMT suggesting "DC Server Down". I fixed this by restarting the DC Sevice on the Publisher server.

My question however is, what could have possibly led this issue to have caused in the first place??

Will rate all helpful posts.

Thanks,

Naveen

Tommer Catlin · ‎03-30-2009

I have seen a lot of CUCM 4.x servers that go running for months and months with little to know attention. Like any Windows Server, you need to do maintenance. This includes rebooting your Publisher/Subscriber I would say every other month or so. It all depends on the CPU cycles and physically memory being used on the system. Windows has inherent memory leaks. The system will eventually run out of memory if nothing is done for months on end. Odd things will pop up such as you have seen.

Being proactive is better than reactive.

Naveen Raj Janardhan · ‎03-30-2009

Hi Tommer,

It is interesting you should mention that. We had our Publisher/subscribers rebooted just Tuesday last week after having the DMA tool installed on the Publisher. Do you think this breakdown has anything to do with the installation?

Thanks,

Naveen

Tommer Catlin · ‎03-30-2009

Yeah, then it's something else. DMA should not cause anything, but when it does run, it eats up alot of processing cycles and memory. Unless something leaked on the application? You can probably turn on traces for that service and if it happens again, you can send the logs to TAC. But it sounds like if you are using DMA, it will be a mute point anyways soon enough.

Naveen Raj Janardhan · ‎03-30-2009

Thanks a lot for the info Tommer. I did find a lot of errors in the event logs pointing out to something very similar to a lot of CPU cycles. 5 points from me for the detailed explaination!!!

Tommer Catlin · ‎03-30-2009

Possibly if the DMA tool was "grinding" trying to gather the data in the SQL database, it could have cause some sort of outage. I have not personally seen it, but anything is possible I guess. If you have had a history of database problems with 4.x, the DMA tool may have caused it to run longer than it should trying to extract the information to it's "TAR" file.

Anyways, glad I could help! Helps us all in the forums!