Run Failover

Unlike test failover jobs which can be run multiple times and are available to be reset, failover is a one-time job for each replication policy. After a replication policy is failed over, all relevant replication jobs will stop running and no future failover jobs can be run for this policy. Resources created during your test failover jobs will be reset as well. To protect the same resource scope, you’ll need to create a new replication policy. Therefore, for testing purposes, run failover test jobs instead of the failover job.

To run failover jobs, complete the following steps:

  1. On the Replication page, select your target replication policy.

  2. Click Failover.

  3. In the confirmation window, you can check the time of your latest recovery point. If you click OK here, a failover job will run based on this recovery point.

    To fail over based on another recovery point, click see more details in the confirmation window, and you will be redirected to Policy details > Recovery points. Select a successfully created recovery point in the left section, and click Failover on the upper-right. The failover job will run based on this selected recovery point.

Failover Details

Refer to the following tables for the details of failover for each resource type.

Microsoft Azure

ResourceFailover Details
Virtual machineCreate virtual machine in the policy's target environment .
SQL databaseApply metadata of the selected recovery point to the database replicated in the policy's target environment.
Storage accountApply metadata of the selected recovery point to the storage account replicated in the policy's target environment.
Delete the replication rule created by the policy's replication job from the Azure object replication page.
Kubernetes ServiceCreate Kubernetes clusters and disk storage in the policy's target environment.
My SQL flexible serviceApply metadata of the latest recovery point to the database replicated in the policy's target environment.
Delete the read replica created by the policy's replication job from the source database.
PostgreSQL flexible serverApply metadata of the latest recovery point to the database replicated in the policy's target environment.
Delete the read replica created by the policy's replication job from the source database.

Amazon Web Services

ResourceFailover Details
EC2 instanceCreate EC2 instances, IAM roles, policies in the policy's target environment and apply the saved settings.
S3 bucketApply metadata of the selected recovery point to the S3 bucket replicated in the policy's target environment.
Delete the replication rule created by the policy's replication job from your S3 bucket.
Aurora and RDS databaseApply metadata of the selected recovery point to the database replicated in the policy's target environment.
Delete the read replica created by the policy's replication job from the source database.
DynamoDBApply metadata of the selected recovery point to the database replicated in the policy's target environment.
Delete the read replica created by the policy's replication job from the source database.
Elastic Kubernetes ServiceCreate Kubernetes cluster, disk storage, IAM role, and IAM policy related to the cluster in the target environment.
*Note: If there are IAM role and policy with the same names, new ones will not be created

Google Cloud

ResourceFailover Details
Cloud storageApply metadata of the latest recovery point to the bucket replicated in the policy's target environment.
Delete replication destination in source bucket and cross-bucket replication.
Kubernetes engineCreate Kubernetes cluster, disk storage, IAM role, and IAM principle and role of related instance.
*Note: If there are IAM role and policy with the same names, new ones will not be created.
SQL instanceApply metadata of the latest recovery point to the instance replicated in the policy's target environment.
Delete the read replica created by the policy's replication job from the source database.
Virtual Machine (VM) instanceCreate VM instance, IAM roles, and IAM principles related to the instance.
*Note: If there are IAM role and policy with the same names, new ones will not be created.

Webhook URL Request

If you have configured a Webhook URL in the replication policy, the failover job details will be sent to the configured URL in HTTP/HTTPS POST requests. An example of the details sent is as follows:

{ "PolicyName": "policy01", "ResourceMapping": [ { "ResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Compute/virtualMachines/vm01", "ResourceType": "VirtualMachine", "ResourceName": "vm01", "TargetResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Compute/virtualMachines/vm01" }, { "ResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Compute/disks/vm01_OsDisk_1_94794b06b4df4cdf868f6d1cd692dccc", "ResourceType": "Disk", "ResourceName": "vm01_OsDisk_1_94794b06b4df********6d1cd692dccc", "TargetResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Compute/disks/vm01_OsDisk_1_94794b06b4df********6d1cd692dccc" }, { "ResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Network/networkInterfaces/vm01898_**", "ResourceType": "NetworkInterface", "ResourceName": "vm01898_**", "TargetResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Network/networkInterfaces/vm01898_z1" }, { "ResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Network/virtualNetworks/vnet01", "ResourceType": "VirtualNetwork", "ResourceName": "vnet01", "TargetResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Network/virtualNetworks/vnet01" }, { "ResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Network/publicIPAddresses/vm01-ip", "ResourceType": "PublicIPAddress", "ResourceName": "vm01-ip", "TargetResourceId": "/subscriptions/xxxxxx/resourceGroups/xxxxxx/providers/Microsoft.Network/publicIPAddresses/vm01-ip" } ] }