Fault Description
This fault action will restart the TencentDB for the MySQL instance specified by the user to simulate fault scenes such as momentary disconnection of database services.
Impact of Faults
1. During a restart, the instance will be unable to properly provide services, and existing connections may be interrupted. Make preparations to avoid impact.
2. During a restart, existing connections may occur flash interruptions.
3. During a restart, instance change and other operations are not allowed.
4. During a restart, if the size of the business write-in is large, there will be a lot of dirty pages. To reduce the duration of service unavailability, restart may fail.
5. During a read-only instance restart, weights and elimination policies will be invalid. Please stay tuned.
Experiment Implementation
Step 1: Experiment Preparation
A TencentDB for MySQL Instance.
Step 2: Experiment Orchestration
2. Click Skip and create a blank experiment at the lower left quarter.
3. Fill in experiment information, select a TencentDB for MySQL instance for the experiment objective, and click Add Now to add fault actions.
4. Select MySQL restart fault action, and click Next.
5. There is no fault parameter to be filled in for this fault action. Click Confirm, and then click Next.
6. Submit to create an experiment.
Step 3: Experiment Execution
1. Go to Experiment Details, click Execute to start executing an experiment.
2. After a fault occurs, click Details to check execution results. Restart ID, asynchronous request ID, and estimated restart duration can be obtained from the log. You can click Task Details List link in the log to go to TencentDB for MySQL Task List and check task details. Task Details: Restart task initiated can be checked in the Task List. For more information about the execution, click Task Details on the right side.
Instance Monitoring: During a restart, the fracture can be seen in Monitoring Metrics such as CPU.