Disaster Recovery to vCloud Air is now a tested and reliable solution and helps customer to replicate their critical workloads to vCloud Air so that in event of Disaster in on-premise, customer can recover their workloads and continue their operation without much impact to business.
In this post we will learn how to recover VM’s in vCloud Air in event of disaster or in case when scheduled maintenace is planned in on-premise. For those who are not much familiar with what is vCloud Air Disaster Recovery solution, I urge you to read about that from below links:
Lets get started with recovering VM’s.
There are two type of recovery available with vCloud Air/vSphere Replication i.e.
- Test Recovery
- Full Recovery.(Planned Migration)
Test recoveries allow you to verify that source data is replicated correctly on the target site. When you initiate a replication task to cloud, Disaster Recovery to Cloud creates a placeholder virtual machine on the target virtual data center. The placeholder virtual machine is not visible on the network and is not accessible until you recover it or run a test recovery.
When you run a planned migration operation, the replication source virtual machine is powered off. The placeholder virtual machine that is created in the cloud during replication is configured to run as a fully functional virtual machine. When the recovered virtual machine is powered on in the target cloud site, the replication task on the source is no longer active.
To perform a test recovery login to your vCenter Web-Client and navigate to vSphere Replication > Monitor > Outgoing Replications and right click on VM which you want to recover in cloud.
1: For doing test Recovery, select the option from the dropdown menu and click on it.
2: Under Test recovery options, you get 2 options. Let me explain a bit about these:
Synchronize recent changes: This option allows you to start the recovered VM with the very latest data available on the source VM. Since every VM is configured with a pre-defined RPO value and changes are synched from source VM to destination VM when VM met its RPO interval. This option is helpful to push the changes to destination VM if next sync has not happened.
This option seems nice but its a bit unrealistic as in case of disaster, you would not have source VM in working condition to push those changes which might have occured in your VM after last successful sync.
Use latest available data: in this case the VM is recovered on the basis of the last successful synchronization. This is probably what you would use in the case of a real disaster.
For Test recovery you can go with option 1 and hit Next.
3: On Ready to complete page ensure that Power on test VM is selected. Hit finish.
4: You will see that prior to test recovery, a sync of data change will happen between source and destination VM.
At the same time in VCA side you will see DR test started.
And a few second later test recovery will complete.
6: At this tiem recovered VM will be in power on state. You can login to VM and verify if the replication is happening fine and data is present in your VM.
Under VM tab in VCA portal, you will see VM status as “Test Complete”
And in on-prem VC, VM test status will chage to Ready
Post performing a test recovery and verifying everything is in OK state, the next task is to perform a test cleanup because you can run another test recovery or a planned migration for a replication only after the results of its previous test recovery are cleaned up.
Test cleanup can be done from both on-prem and from VCA portal. In on-prem VC, select the VM which is test recovered and hit button to start cleanup process.
In VCA side you will observe Test cleanup started message.
In on-prem VC “cleanup in progress” message can be observed.
And few seconds later test cleanup will be completed.
And the Cloud VM will be again Powered-off.
Planned migration is the actual recovery of VM in cloud. You have 2 options for recovering a VM.
- From on-prem (Planned Migration)
- Cloud side recovery (from VCA portal by selecting VM and clicking on Recover).
Question is when to use what option. Answer is very simple.Planned Migration is useful for 2 cases:
1: When doing any scheduled maintenance in on-prem. At this time both source and cloud VM are in healthy state and Cloud VM can be used for continuing business operations as source VM may be shutdown during maintenance process.
2: When you want to make changes to your on-prem replicated VM like modifying disk size or adding other resources to VM. At this time you can full recover the cloud VM and power it off and make the same changes in cloud VM which you have made in on-prem VM and again re-establish replication between the two.
Recovering VM from cloud side is used when you dont have source VM available i.e a disater has happened in your on-prem. At that time you can login to VCA portal and recover the VM to minimize the downtime of your critical business application.
In this post I am demonstrating Planned Migration only.
1: Select the VM in question and right click on it and from drop down menu, select Planned Migration.
2: You can use “synchronize recent changes” option to ensure all new data changes are being pushed to your cloud VM and hit Next.
3: If your on-prem VM has VMware tools installed, then you can select guest shutdown option which will gracefully power off your on-prem VM, else you can select Power-Off which will perform a hard shutdown of VM. Hit Next to continue.
4: Select “Power on the recovered VM” and hit finish.
VCA portal will let you know that DR failover has been initiated.
VM state in on-prem VC will chage to “Recovering”
And after waiting for few minutes you will see staus chaging to “Recovered”
VCA portal will confirm that failover task have been completed and VM is now recovered in Cloud.
Thats it for this post. I guess replicating and recovering from failure wa snever so simple but vCloud AIR DR solution has made it possible.
I hope you enjoyed reading this post. Feel free to share this on social media if it is worth sharing. Be sociable