I'm exploring Azure Site Recovery (ASR) and would love some tips on automating disaster recovery processes. While I get the concept of Recovery Plans, I'm curious if there's a way to fully automate the disaster detection and recovery procedures. Here are some specific points I'm struggling with:
1. Can the restoration process be automated so that it doesn't require any manual intervention? I understand that Recovery Plans allow for grouped VM restorations, but I'm looking for a way to trigger these processes automatically upon disaster detection.
2. How do you manage service dependencies when restoring Azure VMs? For instance, the database must be online before the application servers come up.
3. When restoring a VM in a secondary region, it usually lacks a Public IP address or Network Security Group (NSG). How can I automate the reconfiguration of these settings, especially when dealing with numerous VMs?
4. Finally, what's the best approach to handle DNS when we restore services in another region and thus get new private IPs?
I'm eager to learn how others manage their disaster recovery processes!
1 Answer
Based on my experience with ASR projects, automatic failover isn't common among clients because it requires very reliable signals to avoid false alarms. If you want to go that route, setting up Azure Monitor to trigger a runbook or function in a separate area to kick off failover plans could help.
Thanks for the insight! When you mentioned scripts, what languages do you recommend? Are you referring to Bicep, Terraform, or something else? Also, how does a "failover group" fit into this?