Welcome to this EMC Support Community Ask the Expert conversation. This discussion covers Avamar with VMware backup.
Your hosts for this conversation were:
Being familiar with best practices both specific and general for these environments has enabled him to quickly identify inefficiencies and possible improvements that can be made in technology and process. Experienced in supporting both disk and tape technologies including StorageTek, NetApp, EMC and Data Domain. He writes at backupbuddha.ca about backup recovery for both the enterprise as well as the home consumer.
He hopes someday to retire to his secret underground base surrounded by an army of robot ninjas. You can follow him on twitter @backupbuddha.
Topics covered were :
The discussion was live from the 11th June until the 29th June, 2012 and hosted by Vijay and Dan . Thanks to them both for facilitating a great discussion.
This discussion is now open for questions. Vjay and Dan will be looking in on your questions from tomorrow.
We look forward to a lively and interesting technical discussion.
Hey everyone. Thanks for stopping by! I want to thank Mark and Ian for the invite. Last year I had the opportunity to build a new Avamar grid and configure it to protect my companies cloud computing solution. http://www.mcloud.ca/
I learned a lot and hopefully I can remember some of those learnings and "gotchas" to help you in your current environment. You can follow me @backupbuddha on twitter and check out my website backupbuddha.ca.
Thankyou for this forums.
We are currently using Avamar v18.104.22.168 to do image backups or our ESXi 5 environment but don't seem to be able to get the change block tracking working. A 121 gb virtual machine takes about 4 hours to backup each day. What should I look out to fine out why this is not processing faster and using CBT?
That is far to slow. Even without change block tracking the backup should be much faster. In some cases CBT backups can be slow if there are previous VM snapshots that were taken prior. Delete any previous snapshot and try it again.
The transport mechanism usually determines the backup rate.
Under "Normal" operation, the Avamar proxy "mounts" the snapshot to it's SCSI bus, using the "SCSI Hot Add" feature of vmware. This results in fast disk-speed I/O.
If, for some reason, the "SCSI Hot add" feature does not work, the vmware API fails-over to the slower Network-based transport (Network Block Device or NBD based transport) which is considerably slower.
Look in your avvcbimage log for a message that looks like:
avvcbimage Info <9675>: Connected with hotadd transport to virtual disk ...
If you don't see "hotadd", you will need to diagnose why that mode is not being selected.
There is a long list of steps to diagnose that, so I will not go there - you can request EMC support for assistance with that.
Looking through the log I can see the following,
2012-06-12 12:28:08 avvcbimage Info <9625>: Connected with hotadd transport to virtual disk [SAN-A_R10_VP2_157] sadoc3/sadoc3_1.vmdk
I assume this is what you mean.