This article describes how the Automated Disaster Recovery testing functionality works for Amazon EC2 resources in CloudRanger. With the existing backup system in place, the disaster recovery testing functionality allows you to test that the Amazon Machine Image (AMI) backups you are taking actually work successfully in practice.
You can define your Recovery Point Objective (RPO). If the backup policy associated with the instances in your disaster recovery policy is frequent enough to meet your requested RPO, then those instances will be restored together to an alternative region. This process can run on demand or based on a defined frequency. For more information on best practices for AWS disaster recovery planning, refer to this CloudRanger blog post.
Step 1: Create a backup policy
Create a backup policy for AMIs.
Note: This step is important as disaster recovery is dependent on a valid AMI being available in both the source and target regions.
Go to the Policies page by clicking Backup Policies on the left pane and click Create Policy.
Step 2: Add AMI backup policy details
Fill in the backup policy details. Refer to the field description for more information. When you are happy with your backup policy settings, click Save Policy.
Step 3: Choose an execution option
After saving the policy, you can either wait for it to run on schedule or click EXECUTE NOW on the backup policies page.
This will run the policy, create the backups and copy them across to your chosen region. The logs can be found by opening the policy again and clicking on the logs button at the top of the page. The steps are shown as they execute.
Step 4: Create a Disaster Recovery Plan (DRP)
You are now ready to create a disaster recovery policy. Start by clicking Disaster Recovery on the left-hand side of your dashboard. Then, click Create Disaster Recovery Plan.
Step 5: Enter your Disaster Recovery Plan values
CloudRanger can recover instances from AMIs taken with the backup policy that you built above. Choose the region that the instances are currently running in and then select the target region as the region that you are copying to in the backup policy.
Using the search by Instance ID and Instance Tag options, you can select the servers you wish to restore.
Choose the time of the AMIs you want to recover with. The system will look for matching AMIs based on the restore region, selected servers and a time of least as old as the desired RPO specified.
If there are no AMIs matching the criteria, CloudRanger will show you what is missing.
Step 6: Clone your environment
If this is the first time you are running this restore policy, you should clone the source networking to the destination region in the Environment Mappings section.
Network and access settings require new values for the destination location of the restore. You can map the values from the Source Region to the Target Region. VPCs can be cloned and mapped automatically but Instance SSH Keys cannot be cloned and must be mapped manually.
You do this by clicking on the destination region drop-down list and then clicking Clone Environment. This will show a pop-up window with each of the source VPCs.
Clicking Start Clone will start the Cloud Formation process, which can take some time depending on the size. If you look in your AWS environment, you will see the Cloud Formation stack being created. This process may take some time. For more information about cloning, refer to VPC Cloning Process.
Step 7: Confirm mappings
Once this process is complete, CloudRanger will then populate all the destination network infrastructure with the mappings from the source that were created.
Step 8: Set the scheduling frequency
In the Disaster Recovery Testing section, you can choose the frequency of the schedules where you would like to automatically test created backups by temporarily restoring them to an instance.
If you are happy with your DR Plan, then click SAVE POLICY.
Step 9: Executing manually
Finally, if required, you can execute your disaster recovery plan manually. You can do this by clicking Execute Now on the Disaster Recovery Plans page and selecting the RPO. It defaults to get the latest backups you have.
You will then be presented with the screen where you can choose your Point In Time Restore. By clicking Restore, you can then recover your specific Amazon EC2 resources to the target region.