Customer of mine has CallManger 4.1 and this morning their publisher crashed. During investigation he discovered that one of the IDE channel's failed and when he moved the hard drive to another channel the server came up and was functioning fine. Having said that, the customer was concerned that the server had only one hard drive. He inquired of me if he could install another hard drive and do a RAID 1 configuration to provide redundancy in the event one of the hard drives fail. I have done research for quite some time now and have been unable to find documentation that speaks to this. Any suggestions or direction is greatly appreciated!!!
It sounds like your customer is interested in upgrading and thus purchasing new hardware. If that is the case, this would be my advice - leave the server as-is and make sure you have backups going to an off-box location (tape drive, network file share, etc). It is likely that 1) your server is already unsupported or nearing end of support if it is is an MCS H1/I1 and 2) 4.1 is definitely End of Life. So, while adding and configuring RAID may sound like a good stop gap - you may be shooting yourself in the foot if something goes wrong.
Please rate helpful posts!
Since you are on a Windows install (presumably on a 7815 or other low end server without a RAID controller) there is nothing to stop you putting another IDE disk in and setting up software RAID.
However - the Cisco line would be that if you want a RAID setup, then you should get a server that is built with support for RAID.
If you move to a linux-based release of CCM, you won't have access to do things like add software RAID.
You would also incur additional performance overhead that you wouldn't get with hardware RAID... and if you ever get a disk failure, you might need manual intervention to get it to boot of the second disk in the software RAID pair.
So - the official line would be:
Upgrade to bigger servers with RAID controllers..
Or just ensure you are taking backups, and rely on your server-level resilience to get you through (assuming you have two CallManagers in your cluster).
Please rate helpful posts...