I'm trying to setup live migration in Openstack using Ceph RBD backed volumes.
From what I understand I also need to put the libvirt folder
/var/lib/nova/instances on shared storage for it to work as Nova tests for this
as part of the migration process. I decided to look at using CephFS for this
purpose.
I've created a CephFS file system and mounted it on my two compute nodes as
/var/lib/nova/instances but I'm getting some strange results. It seems like
after one of the hosts accesses a file then it becomes locked and the other
host can't access it. For example, I complete the following steps:
1\ Create new Openstack instance using the boot from image (create new volume)
option. New instance is created with RBD backed volume as expected. On the
relevant compute host I see the instance folder created under
/var/lib/nova/instances/{instance id} with two files inside libvirt.xml and
console.log. If I cat the libvirt.xml file it works as expected.
2\ Live migrate the instance to other host. Appears to work as expected
although instance status in Horizon stays as migrating forever. I can see
instance has moved to second host by running virsh list on both hosts.
4\ Now, if I attempt to cat the libvirt.xml file on the new host I get
"Operation not permitted".
I'm assuming this isn't what's expected?
To test this further I tried the following basic tests:
On Host 2:
root@devops-kvm02:/var/lib/nova/instances# echo hello > test
root@devops-kvm02:/var/lib/nova/instances# cat test
hello
root@devops-kvm02:/var/lib/nova/instances#
Then from Host 1:
root@devops-kvm01:/var/lib/nova/instances# cat test
cat: test: Operation not permitted
root@devops-kvm01:/var/lib/nova/instances#
Then back on Host 2:
root@devops-kvm02:/var/lib/nova/instances# cat test
cat: test: Operation not permitted
root@devops-kvm02:/var/lib/nova/instances#
Should this even work? My understanding is CephFS allows concurrent access but
I'm not sure if there is some file locking going on that I need to understand.
Thanks,
Neville
_______________________________________________
ceph-users mailing list
[email protected]
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
