Perform Aiven for PostgreSQL® switchover to the recovery region Limited availability

Perform a planned promotion of your recovery service while the primary service is healthy.

Switch over to your Aiven for PostgreSQL® recovery service for planned maintenance, simulating a disaster, or testing the resilience of your infrastructure.

Prerequisites

CRDR setup up and running
One of the following tools for operating CRDR:

Switch over

CLI
API
Terraform

Run avn service update to promote the recovery service to active:

avn service update RECOVERY_SERVICE_NAME \
   --disaster-recovery-role active

Replace RECOVERY_SERVICE_NAME with the name of the recovery service, for example, pg-demo-recovery.

Verify the switchover by checking both services:

Recovery service status:

avn service get RECOVERY_SERVICE_NAME \
  --json | jq '{state: .state, disaster_recovery_role: .disaster_recovery_role}'

Expect the following output:

{
  "state": "RUNNING",
  "disaster_recovery_role": "active"
}

Primary service status:

avn service get PRIMARY_SERVICE_NAME \
  --json | jq '{state: .state, disaster_recovery_role: .disaster_recovery_role}'

Expect the following output:

{
  "state": "RUNNING",
  "disaster_recovery_role": "passive"
}

Call the ServiceUpdate endpoint to change the disaster_recovery_role of the recovery service to active:

curl --request PUT \
  --url https://api.aiven.io/v1/project/PROJECT_NAME/service/RECOVERY_SERVICE_NAME \
  -H 'Authorization: Bearer BEARER_TOKEN' \
  -H 'content-type: application/json' \
  --data '{"disaster_recovery_role": "active"}'

Replace the following:

PROJECT_NAME, for example crdr-test
RECOVERY_SERVICE_NAME, for example pg-demo-recovery
BEARER_TOKEN

After sending the request, verify the status of each service:

Recovery service status:

avn service get RECOVERY_SERVICE_NAME \
  --json | jq '{state: .state, disaster_recovery_role: .disaster_recovery_role}'

Expect the following output:

{
  "state": "RUNNING",
  "disaster_recovery_role": "active"
}

Primary service status:

avn service get PRIMARY_SERVICE_NAME \
  --json | jq '{state: .state, disaster_recovery_role: .disaster_recovery_role}'

Expect the following output:

{
  "state": "RUNNING",
  "disaster_recovery_role": "passive"
}

The Terraform provider manages the disaster_recovery integration as a static topology declaration. All integration fields are immutable, so a switchover requires destroying the existing integration and creating a new one with the source and destination reversed. This approach has a window during which the DR integration is offline. For a lower-risk switchover, use the Aiven CLI or API instead.

To use Terraform:

Remove the existing disaster recovery integration from Terraform state.
```
terraform destroy -target=aiven_service_integration.disaster_recovery
```

Create an integration with the recovery service as the source.

resource "aiven_service_integration" "disaster_recovery_switched" {
  project                  = var.project_name
  integration_type         = "disaster_recovery"
  source_service_name      = aiven_postgresql.recovery.service_name
  destination_service_name = aiven_postgresql.primary.service_name
}

Apply the new configuration:
```
terraform apply
```

After the switchover completes, your primary service is Passive, and the recovery service is Active, which means the recovery service is in control over your workloads.

Prerequisites​

Switch over​

Prerequisites

Switch over