Crfclust.bdb occupies huge space and causes the disk to be full

Crfclust.bdb occupies huge space and causes the disk to be full

Problem Analysis:

When you get to work, you have a phone call saying a set of RAC database can not access, then landed on the first view of the disk space, found two nodes/dbsoft directory is full, the first suspect is the alarm log files and $oracle_home/dbs generated by the archive log file is not deleted, After the examination the facts denied my judgment.

It was immediately thought that the CHM (Clusterhealth Monitor) service did not cause the CRF file to grow indefinitely resulting in full disk space, and then into the CRF directory check $oracle_home/crf/db; a CRF file causes the disk to fill up, The Crfclust.bdb file has reached 31GB.

Problem handling:

Now that we know the cause of the problem, the next step is to deal with the problem, and there are two ways to deal with it:

1, delete the large CRF file, and close the ORA.CRF. (We are using the first method.) ）

2, delete the large CRF file, and apply patch 10165314 for repair.

Problem-handling process:

We are using the first method above, node one processing process is as follows:

1. View ORA.CRF Service

2. Stop ORA.CRF Service

3. Delete all the crf*.bdb in the $oracle_home/crf/db/$HOMENAME/directory

4. View disk space, free.

The second process of node is the same.

Crfclust.bdb occupies huge space and causes the disk to be full