Slow Drive Fault

Symptom

1. Global Cache cannot provide services properly in the cluster, and a slow drive fault alarm is queried.
2. During routine cluster health check, the administrator finds a slow drive fault alarm.

Cause Analysis

The drive read/write latency in the cluster is abnormal. As a result, Global Cache times out during request processing, causing service exceptions.

Solution

1. Restart the Global Cache process on the node.
2. Remove and then insert the abnormal drive on the node where the alarm is generated.