DR planning with SOBR and AWS

[tgschulze]

I'm trying to develop a DR plan for recovering to AWS in case of major disaster. Currently, we are running V11, our VMs are backed up to an SOBR that is connected to AWS. We have the SOBR set to copy backups to object storage as soon as they are created so data in AWS is relatively up to date. My question is, is there a recommended or best practice to follow for restoring the VMs to EC2 intances. Specifically, how do I get a backup server up and running in AWS to then import the SOBR data to start restoring directly to EC2 instances.

Thank you for any help.

Tom

[Andreas Neufert]

Rollout an EC2 Server with Windows
Download and Install Backup & Replication
Create a Object Repository (not a SOBR) and point it to the (internal) IP address of the S3 Bucket and let it rescan. Internal IP to avoid egress charges.

Go to the import section under HOME-Backups and start the Direct Restore to EC2.

You can test this out today or create a small bucket with a small VM and document the workflow.

[StevenMeier]

Hi Andreas ,
I know this is a long time since you replied to above so hopefully you are around.
I have questions about you have said.
I have a On Premise VBR Server (non domain joined) that I have replicated to AWS as EC2 Instance and when is on both boxes work fine. The AWS instance is only to be used in case of DR .
Our on Premise VBR Server uses performance local Tier and capacity Tier and cold tier via SOBR in . I export and copy our Veeam SQL database backup to the AWS EC2 veeam Server .
What I want to know is
1) should I just use the SOBR as it is in the AWS DR VBR Server no changes or because the performance tier is based on premise and if it Dies or is unavailable because of DR it wont be available should I use a different method to talk to the AWS s3 buckets that SOBR uses to restore From.

I am struggling to find examples of the best way to handle these scenarios.

2) To test can I import jobs if using the method above you mention by creating a repository and importing and have this not affect the sobr if production is still running as I am worried about breaking the production backups testing this ?

Any other ideas or links you can point me too for established ways of doing this....would be great.

Thanks in advance if you can provide guidance on this

Cheers
Steve

[veremin]

Overall, it’s not necessary for you to replicate the database and set it up within an EC2 instance. To be able to restore backups from the Capacity Tier in a DR scenario, you simply need to add the former Capacity Extent as a standalone object storage repository and then import backups from it.

You can also perform this operation for testing purposes, but in that case, it’s better to stop all activity on the production backup server that interacts with this repository to avoid collisions.

At this time, we do not recommend using the same object storage repository on two different backup servers. In the future, we plan to address this situation and potentially introduce a read-only mode.

Thanks!