Skip to main content

Hi,

Below is my configuration:

Commserve is VM machine and version is 11.24.103.

Media servers - 1 is Physical and another is VM - Both are configured as Disk’s library (Disk’s are coming from Storage Lun’s).

Proxy Node (Access Node) - Both media servers are assigned as access node.

We have around 100 VM servers under a vCenter, so I have split 10 sub clients in each VM.

I have selected the Auto mode and backups are running nbd most jobs.

Currently, I can see the backup/restores are running very slow. Is there anything I can do improve the speed. 

 

Thanks in advance.

  

 

Hello @Thava_91 

When you say slow, what exactly is the bottle neck is something we have to figure out. If you double click on a running job, you will be able to see how much percentage is for reading, network and writing and accordingly we can proceed further to check which area to look further. Also, what is the current throughput with which the backups are running and what’s the expected throughput?

Best,

Rajiv Singal


@Thava_91 

You may configure hotadd transport mode as one of the VSA is VM
https://documentation.commvault.com/11.24/expert/32048_hotadd_transport_for_vmware.html


Hello @Thava_91 

When you say slow, what exactly is the bottle neck is something we have to figure out. If you double click on a running job, you will be able to see how much percentage is for reading, network and writing and accordingly we can proceed further to check which area to look further. Also, what is the current throughput with which the backups are running and what’s the expected throughput?

Best,

Rajiv Singal

I have attached snapshots of current throughput and load.


@Thava_91 

You may configure hotadd transport mode as one of the VSA is VM
https://documentation.commvault.com/11.24/expert/32048_hotadd_transport_for_vmware.html

I can see few of the hosts transport mode showing hotadd and which is running on the VM VSA.

I have not configured hotadd transport mode.


Hi @Thava_91,

 

The vsbkp.log and the cvperfmgr.log on the Media Agent (vsbkp.log on the VSA Proxy) will show you some statistics around the VSA side of the Jobs, These might help you identify any possible bottlenecks for the Job. - In vsbkp.log look for lines starting with string “stat”, You will see these for Readdisk, WritePLBuffer, etc. 

Also check the CVD.log on the Media Agent, this will show media write speeds too.

 

Let us know what you find, If you get stuck then I’d suggest raising a case and our team can help.

 

Best Regards,

Michael

 


Reply