IP Office Server Edition

[desphoneguy]

I know this is a very general question.

We recently had a strange thing happen, both our primary and secondary servers on a network connected to the same switch lost network connection at the same time for a couple of minutes, then came back up.

The obvious thing would be a network issue, but I was wondering if there are any routines (such as backups) or options within IP Office Server Edition that could cause this.

 

[critchey]

Server Editions do not shut down or reboot on their own unless told too. We even had a user trying to figure out how to get his Server Edition to reboot regularly and couldn't figure out... its just not something it is meant to do. It should not lose network connection either. It would not be in the Server Edition end and I would not expect the Server BIOS to have anything causing it either but I guess anything is possible. The most likely cause is, as you said, the network switch.

As a side note it seems odd you would have the Primary Server and Secondary Server connected through a single switch meaning a single point of failure. You want it so if the Primary goes down the Secondary takes over. If this switch goes down both are down.

The truth is just an excuse for lack of imagination.

 

[sizbut]

Server Editions do not shut down* or reboot** on their own" - yes they can. But yes, it would be extremely extremely odd for an existing pair of operational servers to both do it at the same time.

* Virtualized SE servers that detect an address conflict when starting, shut down until the address conflict is resolved (okay, that's more a refusal to start rather than a shutdown).

** All IP Office systems run a watchdog function. If it suspects that the system has locked up for any reason, it will restart the server.

Stuck in a never ending cycle of file copying.

 

[critchey]

Well I guess when I said "told too" I should of been more specific. If you schedule a shut down or watchdog triggers then yes it can happen. Under "normal circumstances" it won't "just happen". Both doing it at the same time and being tied to the same switch seems fairly obvious the cause.

The Virtualized SE that detect an IP conflict will actually boot but that interface will not come up (had a customer have a device on 192.168.43.1). The SE boots but the WAN port would not populate it's default IP. The device was not responding to pings and was not showing on network scans but was eventually found.

The truth is just an excuse for lack of imagination.

 

[TheMitelGuy]

What do the logs say?