Troubleshooting system storage problems

In the previous chapters, we introduced three different storage systems, namely GlusterFS, CEPH, and DRBD volumes. In this section, I will list out the problems I have encountered while building the system for each separate filesystem.

DRBD volume not in synchronization

During a normal operation, data synchronization should be automatically performed. However, there are different scenarios that will cause a synchronization problem. The following problems will be covered in this section:

• Need access to up-to-date data

• Node with Diskless status

• Node with Unknown status

Need access to up-to-date data during service initialization

Root cause: Your system might indicate that the data inside the current node (in the following example, vmsrv02) is not up to date, and the resynchronization does not take place probably because of the timeout.

Solution: So, we need to tell the DRBD service to copy data from a peer node by overwriting the data in the current node....