Enabling lockd

For image-based storage pools which are POSIX compliant, you can enable it easily by uncommenting lock_manager = "lockd" in /etc/libvirt/qemu.conf or on both hypervisors:

Now, enable and start the virtlockd service on both the hypervisors. Also, restart libvirtd on both the hypervisors.

# systemctl enable virtlockd; systemctl start virtlockd
# systemctl restart libvirtd
# systemctl status virtlockd

Starting vm1 on f22-02:

[root@f22-02]# virsh start vm1
Domain vm1 started

Starting the same vm1 on f22-01:

[root@f22-01]# virsh start vm1
error: Failed to start domain vm1
error: resource busy: Lockspace resource '/var/lib/libvirt/images/testvms/vm1.qcow2' is locked

Another method to enable lockd is to use a hash of the disk's file path. Locks are saved in a shared directory that is exported through the NFS, or similar sharing, to the hypervisors. This is very useful when you have virtual disks, which are created and attached using multipath LUN. fcntl() cannot be used in these cases...