Hide Forgot
Description of problem: Concurrent migration will hang on libvirt-0.9.10-6.el6.x86_64.rpm Version-Release number of selected component (if applicable): kernel-2.6.32-250.el6.x86_64 qemu-kvm-0.12.1.2-2.262.el6.x86_64 libvirt-0.9.10-6.el6.x86_64 How reproducible: 90% Steps to Reproduce: 1. Start 15 guests with empty disk 2. Do migration with # for i in {1..15}; do virsh migrate --live mig-$i qemu+ssh://{target ip}/system --unsafe & sleep 1; done 3. Do virsh list Actual results: virsh list will hang, and migration will not finished Expected results: virsh will not hang, migration succeed without error Additional info:
Created attachment 572163 [details] libvirtd.log
Created attachment 572164 [details] backtrace
How is this different from bug 803186, is the only difference in removing virsh migrate-setmaxdowntime command from the loop? This is very likely a dup of that bug but I'll wait for the additional info before closing it.
I report it because the bug 803186 is reported on rhel 6.2 but I can only reproduce this bug on rhel6.3, and on rhel6.2 concurrent migration will not cause virsh command hang
It's a race so you were not probably lucky enough to hit it on rhel6.2. I wasn't able to reproduce this bug even on rhel6.3 (until I saw the backtrace and logs). *** This bug has been marked as a duplicate of bug 803186 ***
Please have a look at bug 803186 comment 17