Update 04th July 2017 11:00 p.m.: Today Microsoft ASR team allow replicating Server 2016 VM’s (Azure-to-Azure DR) scenario as well. These VM’s can support Storage space technology. Can check my short video here.
Kindly note this feature still in preview mode. Being said that I believe this is very important option for some customers. Based on customer feedback Microsoft has identified following points to justify this feature.
- You need to meet compliance guidelines for specific apps and workloads that require a business continuity and disaster recovery (BCDR) strategy.
- You want the ability to protect and recover Azure VMs based on your business decisions, and not only based on inbuilt Azure functionality.
- You need to test failover and recovery in accordance with your business and compliance needs, with no impact on production.
- You need to fail over to the recovery region in the event of a disaster and fail back to the original source region seamlessly.
So being said that below are my observations on ASR for Azure IaaS VM’s.
- Setup and configuration is very much easy (Of course careful planning is required)
- VM’s with Managed disks are not supported (This option will be coming soon)
- You Site Recovery Resource Group has to be created on different region and cannot be on the same region where you production VM’s exists.
- Automated replication. Site Recovery provides automated continuous replication. Failover and failback can be triggered with a single click via GUI.
- Minimum replication time interval is 5 min (Wish this will be improved soon)
- Just like protecting and testing on-premise VM’s to Azure, you can run disaster-recovery drills with on-demand test failovers, as and when needed, without affecting your production workloads or ongoing replication.
- You can use recovery plans to orchestrate failover and failback of the entire application running on multiple VMs. This can be controlled via runbooks (very nice feature)
Ok now let’s get back to action
To make things easier I’ve went ahead and created two RG (Resource Groups) in advance in two regions. I hope name convention is easy to understand it’s purpose.
Inside the ASR-PROD I already created single Server 2012 R2 VM.
So now we have a production VM ready to b protected. Next step is to create Recovery Vault on destination RG.
Select the VMs you want to replicate, and then click OK.
if you want you can override the default target settings and specify the settings you like by clicking Customize.
Once given command to execute Azure Recovery service will go ahead and do the job
Initial replication might take some time. It all depend on how many number of disk you have in your Iaas VM and their size. But I am pretty sure it’s lot faster than uploading your on-premise datacenter VM to Azure scenario. I have experience 3-4 days to upload single VM to Azure
Finally the success results would be as follows,
Nice GUI work from Azure ASR team visually showing which to which region VM getting replicated to,
Experience the DR drill. For this under the Site Recovery click the “Test Failover” option. This will create VM on the ASR RG. Once the test is complete you can select the option called “Cleanup test failover” This will delete the VMs that were created during the test failover
During my demo lab creation came-up with below mentioned error. Problem is newly added disk is not be initialized inside the guest OS. Due to that reason ASR unable to replicate that disk to DR site.