01-11-2016 10:19 AM - edited 03-19-2019 10:34 AM
Hi,
I have a query regarding Snapshots on CUCM version 10.5, based on article published at:
It states Snapshots are not supported because it causes latency and other issues, however what would be the best solution for performing a system state backup incase of a host failure?
Can snapshots be used to restore or is it just having the service running in the background performing automatic snapshots causing the latency? Or is it because after restoration of a snapshot the service degrades after each restore? I am a little confused on the matter and looking to implement a suitable backup solution for a CUCM 10.5 cluster.
Any help would be much appreciated,
Regards,
Christopher
Solved! Go to Solution.
01-11-2016 11:06 AM
So the official answer to this is that the only supported method of backup is to capture regular and complete DRS backup sets from the cluster using the built-in DRS components. In the event of a cluster or node failure, you would restore the DRS backup set onto a like version/IP UC node/cluster.
With that piece of business aside; NO, snapshot restorations ARE NOT supported as a means of backup / DR. Latency is caused by the monitoring services and as the snapshot grows in size. However, that said, it does usually work BUT you are on your own as far as TAC is concerned in terms of support if you do it.
My recommendation is that you configure a hollow cluster (not configured except IP/NTP/FQDN ... etc) EXACTLY the same as the production cluster (minus device/user data) and then power it off. Take regular DRS backups of the production cluster.
In the event of a complete node or cluster failure, power up your hollow cluster and restore the DRS to it (while powering off your old production cluster). You can also store the hollow VM HDDs on different hardware if it is a want/must.
Here is the administrative guide for CUCM DRS: http://www.cisco.com/c/en/us/td/docs/voice_ip_comm/cucm/drs/10_0_1/CUCM_BK_D220A91D_00_disaster-recovery-system-cucm-100/CUCM_BK_D220A91D_00_disaster-recovery-system-cucm-100_chapter_011.html
Here is a guide on how to setup a Linux server as a DRS storage server: http://ryanthomashuff.com/2015/11/why-uc-engineers-need-to-become-linux-aficionados-part-2/
Here is a Cisco guide on configuring a DRS storage server: http://www.cisco.com/c/en/us/support/docs/voice-unified-communications/unified-communications-manager-callmanager/110309-callmanager-backup.html
Thanks,
Ryan
(: ... Please rate helpful posts ... :)
01-11-2016 11:06 AM
So the official answer to this is that the only supported method of backup is to capture regular and complete DRS backup sets from the cluster using the built-in DRS components. In the event of a cluster or node failure, you would restore the DRS backup set onto a like version/IP UC node/cluster.
With that piece of business aside; NO, snapshot restorations ARE NOT supported as a means of backup / DR. Latency is caused by the monitoring services and as the snapshot grows in size. However, that said, it does usually work BUT you are on your own as far as TAC is concerned in terms of support if you do it.
My recommendation is that you configure a hollow cluster (not configured except IP/NTP/FQDN ... etc) EXACTLY the same as the production cluster (minus device/user data) and then power it off. Take regular DRS backups of the production cluster.
In the event of a complete node or cluster failure, power up your hollow cluster and restore the DRS to it (while powering off your old production cluster). You can also store the hollow VM HDDs on different hardware if it is a want/must.
Here is the administrative guide for CUCM DRS: http://www.cisco.com/c/en/us/td/docs/voice_ip_comm/cucm/drs/10_0_1/CUCM_BK_D220A91D_00_disaster-recovery-system-cucm-100/CUCM_BK_D220A91D_00_disaster-recovery-system-cucm-100_chapter_011.html
Here is a guide on how to setup a Linux server as a DRS storage server: http://ryanthomashuff.com/2015/11/why-uc-engineers-need-to-become-linux-aficionados-part-2/
Here is a Cisco guide on configuring a DRS storage server: http://www.cisco.com/c/en/us/support/docs/voice-unified-communications/unified-communications-manager-callmanager/110309-callmanager-backup.html
Thanks,
Ryan
(: ... Please rate helpful posts ... :)
01-11-2016 11:06 AM
Thank you very much for your response, this will definitely be useful.
I will review the links and see what suits best.
Regards,
Christopher
01-11-2016 11:29 AM
Please let me know if I can help further.
Thanks,
Ryan
07-27-2016 10:07 AM
A slightly different question related to snapshots for backups. Would it be supported to do a temporary snapshot for purposes of a backup software package to back it up and then remove the snapshot? It's different in that the snapshot is removed as soon as the backup is taken.
Thanks,
-Tom
07-27-2016 10:36 AM
Snapshots are supported for some apps, if the table says it's not supported, it means not supported under any circumstance.
http://docwiki.cisco.com/wiki/Unified_Communications_VMware_Requirements#VMware_Feature_Support_for_Unified_Communications_8.0.282.29_through_10.0
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide