Hide Forgot
Description of problem: VM's transition to not-responding state then back to UP state. This problem is seen on end-user system when taking VM snapshots. The VM's it-self doesn't have any problems and continue to operate. The VDSM logs contains "monitor become unresponsive (command timeout, age=89.0200000005)" messages for the VM's. Version-Release number of selected component (if applicable): rhevm-3.6.9.2-0.1.el6.noarch vdsm-4.17.35-1.el7ev.noarch libvirt-daemon-1.2.17-13.el7_2.5.x86_64 How reproducible: 1. 100% reproducible on end-user system. Steps to Reproduce: 1. Take and snapshot of the VM and observe the VM's momentarily transition to not responding state then back to Up state. 2. 3. Actual results: VM's momentarily transition to not responding state then back to Up state. Expected results: VM's shouldn't transition to not responding state. Additional info:
Francesco, os it the same as the race on shutdown?
(In reply to Michal Skrivanek from comment #4) > Francesco, os it the same as the race on shutdown? I doubt that - just by looking at the logs - if it takes you >1m to take a snapshot, something's wrong (storage-side performance?). I'd look at VDSM log for storage issues.