Keep application group running for 1 month?

[champ_]

Hello,

I have a customer that want to have a application running for one month in order to do some upgrade testing.
Is this possible? Will the application group shutdown when the backup will run, since the servers in the application group belongs to that backup job?
What is best practice if a customer wants to have a sandbox environment running for one month and not disturb the backup/surebackup?

Regards

[PTide]

Hi,

Is this possible?

Well, that depends. A VM with an intensive throughput running from a backup may result in growing cache file on the NFSDatastore. So the answer is "yes, it is possible" but I don't think it's a good idea.

Will the application group shutdown when the backup will run, since the servers in the application group belongs to that backup job?

That depends on what kind of backup policy you use, but generally speaking - if a backup job needs to do something (merge, delete etc) with any of files involved in SureBackup's activity, then SureBackup Job will be stopped by that ob because backup job has a higher priority.

What is best practice if a customer wants to have a sandbox environment running for one month and not disturb the backup/surebackup?

I'd recommend you to restore a VM from backup into a separate resource pool behind proxy appliance, so it will work like a "normal" VM, without causing problems to your repo.

Also please refer to this thread - it might be helpful.

Thank you.

[champ_]

Hi,

Thanks for the answers. When you talk about vm restore. Do you mean restoring it to a host and then assigning it to the correct veeam vswitch on that host?

[PTide]

Do you mean restoring it to a host and then assigning it to the correct veeam vswitch on that host?

Yes, you'll need to assign restored VM to specific virtual networks so that the VM does not interfere with its production forerunner.

[dellock6]

Does the customer has any of the supported storage arrays in VBR (HP, NetApp, EMC?). If so I think that the on-demand sandbox from Storage Snapshots may be interesting for him as it was designed exactly for such use cases.

[champ_]

Hi,

No, the customer does not have a supported storage array for storage snapshot.

[champ_]

Hi again,

Will the surebackup job also fail if the backup job is of type Incremental?

Regards

[PTide]

Hi,

Yes, it will stop in case one of the following takes place:

a) Merging the backup surebackup is running from
b) Pruning the backup surebackup is running from

Example 1: if you are running a forever incremental then each job run will merge your oldest full with incremental. That would cause surebackup to stop the job.
Example 2: if you are running incremental with periodic fulls then everything is going to be ok, unless your backup job needs to prune full which is being used at the moment.

So, everything depends on your backup policy and retention options. If you run forever or reverse incremental then you don't have a chance to run a VM for a month in a row (unless you take you backup once a month). If you trigger only fulls manually then the backup which is being used by surebackup will not be affected.

[champ_]

So a restore of all the vm's needed for the dev environmnent.
Then change the virtual switch on the vm's.
And then manually start the Proxy server (virtual lab) and the boot up the servers.

Is this the steps I have to take in order to make it work?

[PTide]

So a restore of all the vm's needed for the dev environmnent.

Yes, restore them with different network settings (you can change them when configuring a restore job). I'd also recommend you to restore VMs to another storage so they do not affect your production performance.

Then change the virtual switch on the vm's.

Yes, place restored VMs into an isolated network behind proxy appliance via replacing each VM's NICs with NICs belonging to a preconfigured virtual lab switch. If you need to setup some custom port forwarding rules then please refer to this guide explaining how to log in to proxy appliance.

And then manually start the Proxy server (virtual lab) and the boot up the servers.

Correct.

With all these steps taken you will basically get a copy of your existing VM infrastructure running on an isolated network.