Error locking on node node1: Internal lvm error, check syslog Unable to deactivate logical volume "backup00"
I've got two servers at work running GFS and DRBD in primary / primary mode.
I added two drives to one of the nodes, to add extra storage for backups.
I created a new Logical Group (LG) called backup2 and a Logical Volume (LV) called backup00.
There were two 200Gb drives of Physical Volumes (PVs).
I created the LG using the normal lvcreate commands on the console.
For some reason though, because I was using GFS and clvmd on another share called /share, it decided to make the backup2 LG as clustered instead of a plain vanilla one.
Using the system-config-lvm tool I confirmed this.
Further more when I tried to remove /backup2 it I got and error:
Error locking on node node1: Internal lvm error, check syslog
Unable to deactivate logical volume "backup00"
Well node1 doesn't have a clustered lvm anyway. Very odd.
I'm using GFS and DRBD not the normal nmbd stuff.
I fixed this doing the following:
[root@node1 log]# lvmconf --disable-cluster
[root@node1 log]# lvremove -f /dev/backup2/backup00
I added two drives to one of the nodes, to add extra storage for backups.
I created a new Logical Group (LG) called backup2 and a Logical Volume (LV) called backup00.
There were two 200Gb drives of Physical Volumes (PVs).
I created the LG using the normal lvcreate commands on the console.
For some reason though, because I was using GFS and clvmd on another share called /share, it decided to make the backup2 LG as clustered instead of a plain vanilla one.
Using the system-config-lvm tool I confirmed this.
Further more when I tried to remove /backup2 it I got and error:
Error locking on node node1: Internal lvm error, check syslog
Unable to deactivate logical volume "backup00"
Well node1 doesn't have a clustered lvm anyway. Very odd.
I'm using GFS and DRBD not the normal nmbd stuff.
I fixed this doing the following:
[root@node1 log]# lvmconf --disable-cluster
[root@node1 log]# lvremove -f /dev/backup2/backup00
posted by crash at
7:31 AM
2 Comments:
What does GFS and DRDB mean?
Why is your server called shirley?
GFS is a clustering file system that supports cluster level file locking and various other features. It's a Redhat toy.
DRDB is used for file replication between servers in the cluster. It works better than rsync in my case.
The server was called that name before I arrived :p
Post a Comment
Subscribe to Post Comments [Atom]
<< Home