Symptom
When a disk or riser card upgrade is performed using the CLI, the primary controller is switched to another controller during the upgrade (due to an unexpected reset of the primary controller or other exceptions). Afterwards, the upgrade result is not displayed on the CLI, as shown in the following figure.
Alarm Information
N/A
Possible Causes
- The primary controller is unexpectedly reset (for example, an Oops or OOM error occurs).
- Network connections are normal before the upgrade, but the connection between the primary controller and external gateway becomes abnormal during the upgrade. Therefore, the storage array switches the primary controller to another controller that can connect to the gateway.
Fault Diagnosis
- Verify that the upgrade result is not displayed on the CLI.
- Log in to a controller using the CLI. On the CLI, go to the developer mode and then minisystem mode.
- Run showsysstatus to find the primary controller. For example, controller 0 is the primary controller, as shown in the following figure.
- Log in to the primary controller using the CLI. On the CLI, go to the developer mode and then minisystem mode.
- Run showsystrace to check whether a flow whose ID is 4 exists. If the flow exists, query the Date Time information. If the time is later than the upgrade start time, this fault occurs.
Solution
Collect storage array logs to find out the reason why the primary controller is switched.
- If the controller switchover is irrelevant to the upgrade, run the upgrade command again.
- If the controller switchover is triggered by the upgrade, stop the upgrade.
Check After Recovery
N/A
Applicable Versions
OceanStor V3 series V300R001
OceanStor V3 series V300R002
