ApsaraMQ for RocketMQ provides a fault drill feature that simulates an availability zone-level failover. You can use this feature to test and improve the high availability and stability of your service.
Background
An availability zone (AZ) failure can render service instances in that zone unavailable, leading to partial or widespread service disruptions. ApsaraMQ for RocketMQ supports multi-zone deployment and provides a fault drill feature. By simulating a zone failure—stopping and then resuming the service in a selected AZ of your ApsaraMQ for RocketMQ instance—you can proactively identify and fix vulnerabilities in your system. This practice helps ensure your services remain stable and can withstand real-world failures.
Limitations
This feature is available only for Dedicated Edition instances and Platinum Edition instances that are billed on a subscription or pay-as-you-go basis.
Usage notes
During a fault drill, the following behaviors are expected:
Client connections are temporarily interrupted and then automatically re-established.
The delivery of stacked messages is delayed.
Duplicate messages may be delivered after the service is resumed.
Messages in a sequential topic may be delivered out of order for a short period.
Ensure that your cluster has sufficient capacity. This prevents the remaining AZs from being overloaded and causing service interruptions when one zone is stopped.
While a fault drill task is running, you cannot resize or edit the target instance.
How it works
Procedure
Step 1: Create a drill task
Log on to the ApsaraMQ for RocketMQ console. In the top navigation bar, select a region, such as China (Hangzhou).
In the left-side navigation pane, choose .
On the Fault Drill page, click Create Task.
In the Create Task panel, configure the following parameters, and then click OK.
Task Name: Enter a name for the drill task.
Instance: Select the instance that you want to test.
Step 2: Stop the service
On the Fault Drill page, click the name of the target task.
On the Fault Drill details page, select an AZ and click Stop Services.
Step 3: Verify your application
Verify application availability, identify issues, and fix them.
Step 4: Resume the service
On the Fault Drill page, click the name of the target task.
On the Fault Drill details page, click Recover Services.
Step 5: End the drill
On the Fault Drill page, click the name of the target task.
On the Fault Drill details page, click End Drill.