Hide Forgot
Description of problem: Poweroff (halt) hangs the host if an iscsi pool is used on libvirt. Version-Release number of selected component (if applicable): libvirt-0.8.7-3.el6.x86_64 kernel-2.6.32-94.el6.x86_64 It also happens with rhel6.0 & Fedora14 packages. How reproducible: Always. Steps to Reproduce: 1. create a pool from iscsi 2. poweroff machine 3. Hangs with attached error (a timeout) Actual results: Machine don't end poweroff, stays hanged on halt printing the attached bug once each 120 seconds Expected results: Machine powersoff correctly. Additional info: Why do I think the problem is with libvirt? because if I do: virsh pool-destroy <iscsi pools> And now I poweroff, it poweroff correctly.
You've not actually attached the details of the bug you mentioned you would... How are you triggering the shutdown ? Are you doing a controlled shutdown using 'shutdown -h' ? Can you confirm whether the iscsi initscripts are running during shutdown ? This is unlikely to be a libvirt bug. It simply calls out to the iscsiadm toolset.
Created attachment 476383 [details] Oops during halt
Retried, so far: * RHEL5.6 works * Fedora 14 work (at some point got fixed) * RHEL6.0 and RHEL6.1 fails
I'm not sure this is a libvirt issue - see https://bugzilla.redhat.com/show_bug.cgi?id=672530
Since RHEL 6.1 External Beta has begun, and this bug remains unresolved, it has been rejected as it is not proposed as exception or blocker. Red Hat invites you to ask your support representative to propose this request, if appropriate and relevant, in the next release of Red Hat Enterprise Linux.
I'm pretty sure this is not a libvirt problem. I'm reassigning to iscsi for further investigation. It smells a lot like BZ 566772 to me, but I don't have any real evidence to believe that's the case.
(In reply to comment #2) > 'shutdown -h' ? Can you confirm whether the iscsi initscripts are running > during shutdown ? Screen shot in comment #3 shows iscsi sessions still running when the network has been shutdown. We have got to the final part of the shutdown, kernel shutdown. iscsi really wants to send the sync cache command to make sure data is synced up and we are hung there. Normally iscsi init scripts run before the kernel and network is shutdown. Those init scripts stop the sessions and send the sync cache command. In 6.0 there were some shutdown bugs where if service iscsi start had not started the session then they would not get shutdown cleanly like in this bz. So, if you just used iscsiadm to libiscsi to create the session then they might not get shutdown cleanly. In 6.1 these should be fixed. What is the iscsi initiator-utils tools version? Is it iscsi-initiator-utils-6.2.0.872-21.el6? Are you doing iscsi root/boot?
Since RHEL 6.2 External Beta has begun, and this bug remains unresolved, it has been rejected as it is not proposed as exception or blocker. Red Hat invites you to ask your support representative to propose this request, if appropriate and relevant, in the next release of Red Hat Enterprise Linux.
I still use rm /etc/rc[06].d/*network as a workaround. It's a problem in 6.3 as well
This request was not resolved in time for the current release. Red Hat invites you to ask your support representative to propose this request, if still desired, for consideration in the next release of Red Hat Enterprise Linux.
I've played around a bit with a few different 6.x installs, and can't reproduce this just by having a libvirt iscsi pool active. I'm not sure it's still an issue or not.
Hi Juan, Is this still reproducible on Your side?
(In reply to Martin Hoyer from comment #17) > Hi Juan, > Is this still reproducible on Your side? No reply since Nov.. Closing.