david.macias Tue, 09/02/2008 - 11:55
User Badges:
  • Blue, 1500 points or more

Personally, this doesn't seem like a good idea, since you will have to take the server offline to do this. Why don't you buy a few extra hard drives and just use RAID as your back up. Gives you pretty much the same result.


david

PETER NEGUS Wed, 09/03/2008 - 05:59
User Badges:

The RAID solution is not going to give us fast recovery in the event of total hardware disaster. Besides, any non-standard hardware configuration on the MCS boxes is going to invalidate your TAC support. I was thinking of using this prior to system upgrades etc when we would be taking the box off line anyway.

mauricej74 Fri, 09/05/2008 - 09:49
User Badges:

We are in a similar situation -- we need to make 'fixes' to both our Call Manager and CRA servers and I'd like to do an image of the disks. I know TAC frowns upon imaging, but, incase of a major problem with our updates we can't afford the downtime reloading the OS, software, restoring from BARS, etc.


I was curious as to how others (who have successfully imaged their systems) do it. What I know of True Image Server, is that the software has (?) to be installed under Windows, then it creates a live image.


What I would prefer, is being able to boot off a CD, and image with the system down to a USB drive (either to an image file, or direct disk copy).

Michael Owuor Fri, 09/05/2008 - 11:41
User Badges:
  • Cisco Employee,

Hi Maurice,


For CallManager running pre-appliance versions, we have for a long time recommended removing a drive before performing a major change. Because the UCCX applications have been built on the same OS as the pre-appliance versions of CallManager, the same step of removing a drive has worked well for that application.


Because the steps do not involve changing the type of RAID configuration on the server, and because a replacement drive should be inserted, this is indeed a TAC supported process.


Here's an example of documentation of the steps for removing a drive on a CallManager server.

http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/upgrade/4_2_3/upgrfrom.html#wp1076513


And here are the steps for reverting back to the last known working configuration using the drive that had previously been removed:

http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/upgrade/4_2_3/revert.html#wp1040178


This process often takes from 15 to 30 minutes, depending on the size of database


Third party tools may work, but of course the burden would be on you to fully test the procedures in a lab environment. TAC may indeed push back if you encounter problems and it is discovered that the restoration was done using a non-supported third party tool and procedure. I would highly recommend following the supported steps of BARS backup or pulling a drive.


Hope this helps.


Regards,

Michael.

mauricej74 Fri, 09/05/2008 - 11:51
User Badges:

Hi Michael,


Thank you for the valuable advice. The only thing is, that both of our Call Managers (on MCS 7825H-3.0) only have one drive installed in them. There is space inside each for a second drive but I haven't attempted to add anything to them yet, due to possible support issues (or worse)



Michael Owuor Fri, 09/05/2008 - 12:08
User Badges:
  • Cisco Employee,

Ahh...you certainly are in a delicate situation, Maurice. Don't add a drive to the MCS7825 as they certainly don't support RAID 1 like their MCS7835/45 counterparts. Whichever route you chose, make sure your preventive actions for potential problems includes multiple dry runs in the lab or test environment until you are very comfortable with it.


Regards,

Michael.

Actions

This Discussion