On Fri, Aug 14, 2020 at 19:32:38 -0400, Digimer wrote:
Hi all,
I've got a (pacemaker + drbd) cluster running VMs. When the VM is
running on another node, the local storage is Secondary, meaning that it
can't be read or otherwise touched. As such, virsh thinks there is no
storage and constantly prints errors like:
====
Aug 14 19:25:24 mk-a02n01.digimer.ca libvirtd[1655]: 2020-08-14
23:25:24.812+0000: 1679: error : qemuOpenFileAs:3119 : Failed to open
file '/dev/drbd/by-res/srv01-sql/0': No such file or directory
Aug 14 19:25:24 mk-a02n01.digimer.ca libvirtd[1655]: 2020-08-14
23:25:24.812+0000: 1679: error : qemuOpenFileAs:3119 : Failed to open
file '/dev/drbd/by-res/srv01-sql/1': No such file or directory
Aug 14 19:25:24 mk-a02n01.digimer.ca libvirtd[1655]: 2020-08-14
23:25:24.812+0000: 1679: error : qemuOpenFileAs:3119 : Failed to open
file '/dev/drbd/by-res/srv07-el6/0': Wrong medium type
===
To the system logs (journald), about once every second or two.
Is there any way to tell libvirtd not to complain about this?
Please capture also debug logs [1] around the time the message is logged
so that we can see when it's happening. Unfortunately 'qemuOpenFileAs'
is a somewhat universal helper.
[1]
https://libvirt.org/kbase/debuglogs.html