Hide Forgot
Description of problem: If you kill rgmanager at just the right time during event processing during startup, it can throw errors in to the logs: Oct 18 15:26:30 snap rgmanager[15382]: Shutting down Oct 18 15:26:30 snap rgmanager[15382]: Stopping service service:b Oct 18 15:26:30 snap rgmanager[15382]: Service service:b is stopped Oct 18 15:26:30 snap rgmanager[15382]: #45: Unable to obtain cluster lock: Bad file descriptor Oct 18 15:26:30 snap rgmanager[15382]: #50: Unable to obtain cluster lock: Bad file descriptor Oct 18 15:26:30 snap rgmanager[15382]: #13: Service service:wtf failed to stop cleanly Oct 18 15:26:30 snap rgmanager[15382]: #55: Unable to obtain cluster lock: Bad file descriptor Version-Release number of selected component (if applicable): 3.0.12.1-4.el6 How reproducible: 10% Steps to Reproduce: 1. service rgmanager start 2. sleep 8 3. service rgmanager stop Actual results: Errors in logs. This occurs because we close the lockspace while other threads are trying to acquire locks. Expected results: No errors, clean shutdown. Additional info: This does not look like a high-risk condition, but it is a lot of noise.
Development Management has reviewed and declined this request. You may appeal this decision by reopening this request.