I know that when you back up a callmanager, it will back up user and phone databases. Is there a way that you can either save or back up ALL information including route lists, patterns, partitions, DHCP etc?
In the event that you need to completely rebuild your callmanagers, (worst case that both your publisher and subscriber/s get wasted) that you can restore EVERYTHING needed but the OS and normal build? Basicly, everything that is user configued after the initial callmanager build.
Not sure if that's possible, but what we tend to do is periodically swap out one of the scsi disks from the server, assuming that you have a model with multiple disks configured for RAID 1. On the whole, it's usually the contents of the dc directory that change, and these can easily be backed up daily.
Thats what we do too, buy exact same harddrive if you have RAID 1 mirroring. Pull harddrive from slot 0, put the new harddrive in.
Here is the detail step:
thanks for the help. I guess that I could do the same with the subscriber as well. What we are really trying to do is upgrade to 3.3.3 from various other versions. We are looking to remove a hard drive from the publisher and subscribers, put them in our lab/test callmanagers (same servers), upgrade them in the lab then replace them in the production servers. If you follow this, will this work ok? This way, we really should have very little down time on the production network.
I dont think you need to do backup on subscriber, coz all data is in publisher. Subscriber will copy data from Publisher when you install subscriber.
That sounds ok. Just follow the documentation guidelines regarding upgrading to 3.3.3 to make sure that you follow the correct upgrade procedure. Just remember that once you have the newly upgraded 3.3.3 image on the disk, that you power up the production servers with only these disk installed first.
Just bare in mind that an OS-upgrade will also upgrade Array firmware/BIOS....
So, do an upgrade of your OS on the production servers first! (if you are running an old OS..)
And the MCS backupUtility will backup everything to do a complete restore of your callmanager!
Thanks Sascha, very helpfull. Although, when you say upgrade the OS on the production server first, that is half of the upgrade process. I might as well keep going at that rate. If I did it like that, just the OS upgrade on the production servers (publisher and subscriber) can I still go about removing one hard drive out of both the servers and making them primary in my lab servers and complete the rest of the upgrade like that? But then again, the BIOS would not be upgraded on my lab servers to finish the upgrade, do you see where i am going with this.
I appreciate your feedback.
well, the OS-upgrade only takes 45 mins....should have been done to keep the CallManagers protected for MS-vulnerabilities.....
But that is the best way in my opinion, upgrade the OS, and then take the disks out (shut down the servers first!!!)
But....if you are upgrading from 3.2, just copy the MCS.sti and recovery files, because an upgrade to 3.3 will be a complete installation (new image...)
and OS will be installed anyway, so your labserver arrays/BIOS will be upgraded too....!!!
hope it's clear to you..
Thanks again. Last one for you... You say "But....if you are upgrading from 3.2, just copy the MCS.sti and recovery files, because an upgrade to 3.3 will be a complete installation (new image...)
and OS will be installed anyway, so your labserver arrays/BIOS will be upgraded too....!!! "
How will my production BIOS get upgraded if I do it all on the lab servers? - this being that I don't upgrade the os while the hard drives are in the production servers.
Thanks again, again.
Sorry if I wasn't clear...
What I meant was that if you upgrade from 3.2, you don't need to pull a disk and upgrade that one....
but you still need to upgrade the OS/BIOS/Firmware on the production server....
That is if you have enough disks....