Pulling drives for CallManager upgrades

[bpalmquist]

Do you need to pull drives from subscriber servers when doing upgrades? A lot of documentation says to do so for the publisher, but doesn't mention the subs. How would you fall back the subs if you haven't pulled a drive?

Also, the recommended practice is to shut down, pull drive 0 (and 2 if there is one) and power back up.

Anyone ever have problems with the drive pulling process?

[djscott]

We have done backup and restore on seven different VoIP servers (4 CCMs and 3 VMs) for almost five years, without trouble. We always use Slot 1.

Just be careful to never mix up drives or slots.

[bpalmquist]

You are shutting down, then pulling slot 1 and rebooting? Do you pull Subscriber drives too?

We've been doing it for years too, but have had some issues lately...

[heathrw]

I always pull drives before upgrades. I do remove publisher and subscribers so if I have to revert back they are at both the same level. Removing from Pub and Sub has saved us on an occasion.

[bhattacharya.s]

This is not a Cisco recommended and supported solution. Well, you can try it and it works too.

Make sure that you do not mix up the drives and slots. Before an upgrade, you can pull out the drive from slot 0. If the upgrade fails, then you can put in this drive and delete the other logical drives and create the logical drives again and then extend the array.

That way you will have the system back up and running much faster.