High Availability and Disaster Recovery v1
Use this guide with the deployment details confirmed for your subscription. See the shared infrastructure overview before planning a recovery exercise.
Review DropPoint endpoints
Subscriptions with a continuously running secondary region can let DropPoints follow the highly available Tenant hostname or configure them to connect continuously to both regional endpoints. Continuous connections reduce reconnection time because a DropPoint is already available to service secondary-region work.
For a Tenant named testcorp, replace testcorp.flowgear.net with these endpoints when your subscription supports continuous regional connections:
p-testcorp.flowgear.net
s-testcorp.flowgear.net
Click Update & Restart, sign in to the secondary Tenant, and confirm that the DropPoints are online there.
Warning: Do not explicitly configure both regional endpoints for a cold-standby subscription. Its secondary machines do not run until failover, and DropPoint exponential backoff can delay its secondary connection. Use the highly available hostname unless Flowgear confirms another configuration for the subscription.
Test Connections and HTTP invokes
Confirm that every cloud-executed Connection accepts the secondary region's outbound address. Use a DropPoint when a provider requires a stable private network path.
Test published HTTP endpoints through the secondary region while preserving the Environment hostname in the Host header. For example:
GET https://s-testcorp.flowgear.net/someapicall
Host: testcorp-api.flowgear.net
Authorization: Bearer yourapikey
You can automate this check with Web Request 2 by changing the request domain and adding the Host header.
Prepare and test Workflows
Always On Workflows are brought online automatically only by the region currently recognized as the active primary. If the primary region goes offline, the secondary can take over this responsibility once it is available and recognized as primary. Allow time for the automatic restart checks and confirm that the expected Workflows are running.
Signing in explicitly to the secondary region lets you inspect and manage it; it does not cause a second set of Always On Workflows to start there while the primary remains active. Follow the regional sign-in guidance when checking either region.
Use idempotent operations, bounded retries, Environment-specific Connection values, revision history, structured logs, and external reconciliation. Test designer and published starts, schedules, listeners, HTTP and MCP interfaces, DropPoint loss and recovery, revision rollback, and downstream side effects.
Do not change DNS, host headers, production launch controls, or provider allowlists without an approved plan and rollback path.