It is not possible to provide a quick fix by reclaiming space in the variable directory.
The best option is to move it to another device/location.
The following are the steps involved:
1. Shutdown SharePlex and verify all SharePlex process are shutdown normally (ps -ef | grep sp_ from the UNIX prompt).
2. As root, tar up the entire variable directory (preferably cd to the $vardir and tar the entire contents as root from that location).
3. As root at new location, untar the tar file from the original location of the variable directory to assure correct ownerships and permissions are retained.
4. Modify SP_SYS_VARDIR environment setting and any environment or startup scripts to reflect the change in vardir location if utilizing scripts.
4a. If applicable, modify the value of the environment variable defaults.yaml file to reflect the location of vardir.
5. Now correctly export the new SP_SYS_VARDIR to assure vardir location is now set correctly and verify all scripts have been modified to reflect new $vardir location.
6. Start SharePlex and resume replication.
It is preferable that this procedure is tested in a QA or test environment first.
Since disk full situations can cause queue corruption and other outages, it is strongly recommended that this be addressed as soon as possible.
