A Core Dump Occurs on the Client Process Due to Faulty BCM ZooKeeper Cluster

Symptom

When the BCM ZooKeeper cluster is faulty, the client invokes RBD to query the pool information. The initialization fails, and a core dump is generated.

Key Process and Cause Analysis

The command output on the client shows that the connection to ZooKeeper fails. Check the gcache.log file:

It shows that the BCM ZooKeeper cluster is faulty and cannot be connected. The rbd command on the client fails to initialize the ZooKeeper connection and does not retry. The initialization fails, and abort is invoked. As a result, a core dump occurs.

Conclusion and Solution

This is a common phenomenon. You need to rectify the fault of the BCM ZooKeeper cluster. Then, RBD can be invoked to query the pool information.