Description of problem: While doing volume set in a loop to trigger graph change, the number of established connections for the glusterfs client process keep increasing exponentially Version-Release number of selected component (if applicable): upstream How reproducible: Consistently Steps to Reproduce: 1. netstat -np
Check if its already fixed.
Still seen in the mainline/3.3.0qa31
After a graph-switch, atleast one request should be sent to /dev/fuse. Without it, fuse-bridge cannot initiate cleanup. Can you check whether this issue exists even with an operation like ls b/w graph switches?
Closing this bug, since with operations on mountpoint, cleanup is done.