Controller Failure in a Non-UltraPath Environment

Symptom

• All indicators on controller B are off. Indicators on controller A are normal.
• Read and write requests from the application servers connected to controller B cannot be sent to the storage system, causing service interruption. On the Performance page of the DeviceManager, the write or read I/O traffic on the front-end port of controller B becomes 0.

Possible Causes

The controller is faulty.

Impact

If a controller is faulty and host services are interrupted when UltraPath is not installed, you can manually switch the host services to another functional controller.

Fault Diagnosis

Figure 5-27 Flowchart for handling service interruption caused by a fault in controller B in a non-UltraPath environment

Procedure

1. Switch services from controller B to controller A.
   1. Remove the cable between controller B and the application server.
   2. Connect the application server to controller A.
   3. Reconfigure the initiator of the host where services are interrupted.

2. Power off controller B and restart it.

   Check whether the indicator of controller B is steady green.

   • If yes, go to 3.
   • If no, keep the fault environment intact and contact technical support engineers.

3. Switch services from controller A to controller B.
4. Observe the service processing.

   Enable the performance monitoring function before you check the service processing on the DeviceManager.

   1. Log in to DeviceManager and choose Insight > Performance > Analysis.
   2. Click Create.

      The Create Chart page is displayed.

   3. Specify the value of Chart Name.
   4. Select LUN and Custom for Object Type. In Object Instance, select the LUN that provides services for the server through controller B.
   5. Select Total IOPS (IO/s) for Statistical Metric.
   6. Click OK.

      Observe the I/O performance monitoring data and check whether the performance line is smooth.

      • If yes, the service processing is normal and the procedure is complete.
      • If no, keep the fault environment intact and contact technical support engineers.

Suggestion and Summary

To rectify the fault completely:

1. Install UltraPath on the application server.
2. Replace the faulty controller.
3. Upgrade the storage system.
4. Send the collected logs to technical support engineers for subsequent handling.