08-30-2016 12:02 AM
Hi,
In a test case where there are 3 VM's sharing a hard drive. The client agent is installed on the VM which is directly to attached to drive. To perform a backup operation of the shared drive, should the services of the VM has to be stopped for performing backup operation?
Netbackup version - 7.7.3
The 3 Vm's are in an ESXi host.
A hard disk is created and shared with 3 VM's.
08-30-2016 12:24 AM - edited 08-30-2016 12:25 AM
Hi Koundinyad,
That's an interesting test scenario
I can't think of a reason why you would need to stop any services. I am guessing the other two VM's are accessing the shared drive using a Windows share?
The only possible issues I could think of are:
I would think any of the above issues would should in the job details in Activity Monitor, and should be reasonably easy to resolve.
Have you actually tried backing up the shared drive and you have specific problems, or are you just looking for thoughts as to whether this will work?
Steve
08-30-2016 12:44 AM
I am trying to understand more about the 'shared drive'.
Does each VM have its own vmdk residing on the shared drive used as datastore on ESX server?
Are you trying to test with VMware policy or do you want to backup the VMs with normal MS-Windows policy?
08-30-2016 01:14 AM
I have tested for a smaller amount of data nearing about 50 GB and it was successful. I observed that any operation being performed on this disk is causing backup in a frozen state and is resuming only after completion of the operations on the disk. And also the date rate after the operation seems to slow down the backup operation.
I want to test backup for data of 700 GB and is concerned if the services has to be stopped for un interrupted operation.
Yes the other 2 Vm's are using Windows share.
08-30-2016 01:19 AM
I am using MS-Windows policy.
No each VM doesn't have vmdk residing. vmdk is only present on the VM which is directly attached to the drive.
08-30-2016 10:23 PM
Hi Koundinyad,
I'm interested as to what sort of activity is happening on this disk. Is it a database of some sort that the other two VM's are accessing, or is it just standard file system access? If it is a database there are some things you can do to improve it. I'll go through those once I've seen your answer to the question above.
Also, do you have a time window where activity on this disk is very minimal? I'm wondering if you might be backing this up during a time when there is heavy I/O and whether you can can try scheduling the backup during a period where there is minimal activity to test this.
Steve
08-30-2016 10:38 PM
It's a standard file system access that is being ahred by other 2 VM's.
The time window for minimal I/O disk operations on this disk is good enough. What I'm worried is, what if the disk is being accesseb by any of the other VM's that is sharing the shared storage?