Chapter 7. GFS2 file systems in a cluster

This section provides:

  • A procedure to set up a Pacemaker cluster that includes GFS2 file systems
  • A procedure to migrate RHEL 7 logical volumes that contain GFS2 file systems to a RHEL 8 cluster

7.1. Configuring a GFS2 file system in a cluster

This procedure is an outline of the steps required to set up a Pacemaker cluster that includes GFS2 file systems. This example creates three GFS2 file systems on three logical volumes.

As a prerequisite for this procedure, you must install and start the cluster software on all nodes and create a basic two-node cluster. You must also configure fencing for the cluster. For information on creating a Pacemaker cluster and configuring fencing for the cluster, see Creating a Red Hat High-Availability cluster with Pacemaker.

  1. On both nodes of the cluster, install the lvm2-lockd, gfs2-utils, and dlm packages. The lvm2-lockd package is part of the AppStream channel and the gfs2-utils and dlm packages are part of the Resilient Storage channel.

    # yum install lvm2-lockd gfs2-utils dlm
  2. Set the global Pacemaker parameter no_quorum_policy to freeze.

    Note

    By default, the value of no-quorum-policy is set to stop, indicating that once quorum is lost, all the resources on the remaining partition will immediately be stopped. Typically this default is the safest and most optimal option, but unlike most resources, GFS2 requires quorum to function. When quorum is lost both the applications using the GFS2 mounts and the GFS2 mount itself cannot be correctly stopped. Any attempts to stop these resources without quorum will fail which will ultimately result in the entire cluster being fenced every time quorum is lost.

    To address this situation, set no-quorum-policy to freeze when GFS2 is in use. This means that when quorum is lost, the remaining partition will do nothing until quorum is regained.

    # pcs property set no-quorum-policy=freeze
  3. Set up a dlm resource. This is a required dependency for configuring a GFS2 file system in a cluster. This example creates the dlm resource as part of a resource group named locking.

    [root@z1 ~]# pcs resource create dlm --group locking ocf:pacemaker:controld op monitor interval=30s on-fail=fence
  4. Clone the locking resource group so that the resource group can be active on both nodes of the cluster.

    [root@z1 ~]# pcs resource clone locking interleave=true
  5. Set up an lvmlockd resource as part of the group locking.

    [root@z1 ~]# pcs resource create lvmlockd --group locking ocf:heartbeat:lvmlockd op monitor interval=30s on-fail=fence
  6. Check the status of the cluster to ensure that the locking resource group has started on both nodes of the cluster.

    [root@z1 ~]# pcs status --full
    Cluster name: my_cluster
    [...]
    
    Online: [ z1.example.com (1) z2.example.com (2) ]
    
    Full list of resources:
    
     smoke-apc      (stonith:fence_apc):    Started z1.example.com
     Clone Set: locking-clone [locking]
         Resource Group: locking:0
             dlm    (ocf::pacemaker:controld):      Started z1.example.com
             lvmlockd       (ocf::heartbeat:lvmlockd):      Started z1.example.com
         Resource Group: locking:1
             dlm    (ocf::pacemaker:controld):      Started z2.example.com
             lvmlockd       (ocf::heartbeat:lvmlockd):      Started z2.example.com
         Started: [ z1.example.com z2.example.com ]
  7. Verify that the lvmlockd daemon is running on both nodes of the cluster.

    [root@z1 ~]# ps -ef | grep lvmlockd
    root     12257     1  0 17:45 ?        00:00:00 lvmlockd -p /run/lvmlockd.pid -A 1 -g dlm
    [root@z2 ~]# ps -ef | grep lvmlockd
    root     12270     1  0 17:45 ?        00:00:00 lvmlockd -p /run/lvmlockd.pid -A 1 -g dlm
  8. On one node of the cluster, create two shared volume groups. One volume group will contain two GFS2 file systems, and the other volume group will contain one GFS2 file system.

    The following command creates the shared volume group shared_vg1 on /dev/vdb.

    [root@z1 ~]# vgcreate --shared shared_vg1 /dev/vdb
      Physical volume "/dev/vdb" successfully created.
      Volume group "shared_vg1" successfully created
      VG shared_vg1 starting dlm lockspace
      Starting locking.  Waiting until locks are ready...

    The following command creates the shared volume group shared_vg2 on /dev/vdc.

    [root@z1 ~]# vgcreate --shared shared_vg2 /dev/vdc
      Physical volume "/dev/vdc" successfully created.
      Volume group "shared_vg2" successfully created
      VG shared_vg2 starting dlm lockspace
      Starting locking.  Waiting until locks are ready...
  9. On the second node in the cluster, start the lock manager for each of the shared volume groups.

    [root@z2 ~]# vgchange --lock-start shared_vg1
      VG shared_vg1 starting dlm lockspace
      Starting locking.  Waiting until locks are ready...
    [root@z2 ~]# vgchange --lock-start shared_vg2
      VG shared_vg2 starting dlm lockspace
      Starting locking.  Waiting until locks are ready...
  10. On one node in the cluster, create the shared logical volumes and format the volumes with a GFS2 file system. One journal is required for each node that mounts the file system. Ensure that you create enough journals for each of the nodes in your cluster.

    [root@z1 ~]# lvcreate --activate sy -L5G -n shared_lv1 shared_vg1
      Logical volume "shared_lv1" created.
    [root@z1 ~]# lvcreate --activate sy -L5G -n shared_lv2 shared_vg1
      Logical volume "shared_lv2" created.
    [root@z1 ~]# lvcreate --activate sy -L5G -n shared_lv1 shared_vg2
      Logical volume "shared_lv1" created.
    
    [root@z1 ~]# mkfs.gfs2 -j2 -p lock_dlm -t my_cluster:gfs2-demo1 /dev/shared_vg1/shared_lv1
    [root@z1 ~]# mkfs.gfs2 -j2 -p lock_dlm -t my_cluster:gfs2-demo2 /dev/shared_vg1/shared_lv2
    [root@z1 ~]# mkfs.gfs2 -j2 -p lock_dlm -t my_cluster:gfs2-demo3 /dev/shared_vg2/shared_lv1
  11. Create an LVM-activate resource for each logical volume to automatically activate that logical volume on all nodes.

    1. Create an LVM-activate resource named sharedlv1 for the logical volume shared_lv1 in volume group shared_vg1. This command also creates the resource group shared_vg1 that includes the resource. In this example, the resource group has the same name as the shared volume group that includes the logical volume.

      [root@z1 ~]# pcs resource create sharedlv1 --group shared_vg1 ocf:heartbeat:LVM-activate lvname=shared_lv1 vgname=shared_vg1 activation_mode=shared vg_access_mode=lvmlockd
    2. Create an LVM-activate resource named sharedlv2 for the logical volume shared_lv2 in volume group shared_vg1. This resource will also be part of the resource group shared_vg1.

      [root@z1 ~]# pcs resource create sharedlv2 --group shared_vg1 ocf:heartbeat:LVM-activate lvname=shared_lv2 vgname=shared_vg1 activation_mode=shared vg_access_mode=lvmlockd
    3. Create an LVM-activate resource named sharedlv3 for the logical volume shared_lv1 in volume group shared_vg2. This command also creates the resource group shared_vg2 that includes the resource.

      [root@z1 ~]# pcs resource create sharedlv3 --group shared_vg2 ocf:heartbeat:LVM-activate lvname=shared_lv1 vgname=shared_vg2 activation_mode=shared vg_access_mode=lvmlockd
  12. Clone the two new resource groups.

    [root@z1 ~]# pcs resource clone shared_vg1 interleave=true
    [root@z1 ~]# pcs resource clone shared_vg2 interleave=true
  13. Configure ordering constraints to ensure that the locking resource group that includes the dlm and lvmlockd resources starts first.

    [root@z1 ~]# pcs constraint order start locking-clone then shared_vg1-clone
    Adding locking-clone shared_vg1-clone (kind: Mandatory) (Options: first-action=start then-action=start)
    [root@z1 ~]# pcs constraint order start locking-clone then shared_vg2-clone
    Adding locking-clone shared_vg2-clone (kind: Mandatory) (Options: first-action=start then-action=start)
  14. Configure colocation constraints to ensure that the vg1 and vg2 resource groups start on the same node as the locking resource group.

    [root@z1 ~]# pcs constraint colocation add shared_vg1-clone with locking-clone
    [root@z1 ~]# pcs constraint colocation add shared_vg2-clone with locking-clone
  15. On both nodes in the cluster, verify that the logical volumes are active. There may be a delay of a few seconds.

    [root@z1 ~]# lvs
      LV         VG          Attr       LSize
      shared_lv1 shared_vg1  -wi-a----- 5.00g
      shared_lv2 shared_vg1  -wi-a----- 5.00g
      shared_lv1 shared_vg2  -wi-a----- 5.00g
    
    [root@z2 ~]# lvs
      LV         VG          Attr       LSize
      shared_lv1 shared_vg1  -wi-a----- 5.00g
      shared_lv2 shared_vg1  -wi-a----- 5.00g
      shared_lv1 shared_vg2  -wi-a----- 5.00g
  16. Create a file system resource to automatically mount each GFS2 file system on all nodes.

    You should not add the file system to the /etc/fstab file because it will be managed as a Pacemaker cluster resource. Mount options can be specified as part of the resource configuration with options=options. Run the pcs resource describe Filesystem command for full configuration options.

    The following commands create the file system resources. These commands add each resource to the resource group that includes the logical volume resource for that file system.

    [root@z1 ~]# pcs resource create sharedfs1 --group shared_vg1 ocf:heartbeat:Filesystem device="/dev/shared_vg1/shared_lv1" directory="/mnt/gfs1" fstype="gfs2" options=noatime op monitor interval=10s on-fail=fence
    [root@z1 ~]# pcs resource create sharedfs2 --group shared_vg1 ocf:heartbeat:Filesystem device="/dev/shared_vg1/shared_lv2" directory="/mnt/gfs2" fstype="gfs2" options=noatime op monitor interval=10s on-fail=fence
    [root@z1 ~]# pcs resource create sharedfs3 --group shared_vg2 ocf:heartbeat:Filesystem device="/dev/shared_vg2/shared_lv1" directory="/mnt/gfs3" fstype="gfs2" options=noatime op monitor interval=10s on-fail=fence
  17. Verify that the GFS2 file systems are mounted on both nodes of the cluster.

    [root@z1 ~]# mount | grep gfs2
    /dev/mapper/shared_vg1-shared_lv1 on /mnt/gfs1 type gfs2 (rw,noatime,seclabel)
    /dev/mapper/shared_vg1-shared_lv2 on /mnt/gfs2 type gfs2 (rw,noatime,seclabel)
    /dev/mapper/shared_vg2-shared_lv1 on /mnt/gfs3 type gfs2 (rw,noatime,seclabel)
    
    [root@z2 ~]# mount | grep gfs2
    /dev/mapper/shared_vg1-shared_lv1 on /mnt/gfs1 type gfs2 (rw,noatime,seclabel)
    /dev/mapper/shared_vg1-shared_lv2 on /mnt/gfs2 type gfs2 (rw,noatime,seclabel)
    /dev/mapper/shared_vg2-shared_lv1 on /mnt/gfs3 type gfs2 (rw,noatime,seclabel)
  18. Check the status of the cluster.

    [root@z1 ~]# pcs status --full
    Cluster name: my_cluster
    [...1
    
    Full list of resources:
    
     smoke-apc      (stonith:fence_apc):    Started z1.example.com
     Clone Set: locking-clone [locking]
         Resource Group: locking:0
             dlm    (ocf::pacemaker:controld):      Started z2.example.com
             lvmlockd       (ocf::heartbeat:lvmlockd):      Started z2.example.com
         Resource Group: locking:1
             dlm    (ocf::pacemaker:controld):      Started z1.example.com
             lvmlockd       (ocf::heartbeat:lvmlockd):      Started z1.example.com
         Started: [ z1.example.com z2.example.com ]
     Clone Set: shared_vg1-clone [shared_vg1]
         Resource Group: shared_vg1:0
             sharedlv1      (ocf::heartbeat:LVM-activate):  Started z2.example.com
             sharedlv2      (ocf::heartbeat:LVM-activate):  Started z2.example.com
             sharedfs1      (ocf::heartbeat:Filesystem):    Started z2.example.com
             sharedfs2      (ocf::heartbeat:Filesystem):    Started z2.example.com
         Resource Group: shared_vg1:1
             sharedlv1      (ocf::heartbeat:LVM-activate):  Started z1.example.com
             sharedlv2      (ocf::heartbeat:LVM-activate):  Started z1.example.com
             sharedfs1      (ocf::heartbeat:Filesystem):    Started z1.example.com
             sharedfs2      (ocf::heartbeat:Filesystem):    Started example.co
         Started: [ z1.example.com z2.example.com ]
     Clone Set: shared_vg2-clone [shared_vg2]
         Resource Group: shared_vg2:0
             sharedlv3      (ocf::heartbeat:LVM-activate):  Started z2.example.com
             sharedfs3      (ocf::heartbeat:Filesystem):    Started z2.example.com
         Resource Group: shared_vg2:1
             sharedlv3      (ocf::heartbeat:LVM-activate):  Started z1.example.com
             sharedfs3      (ocf::heartbeat:Filesystem):    Started z1.example.com
         Started: [ z1.example.com z2.example.com ]
    
    ...

7.2. Migrating a GFS2 file system from RHEL7 to RHEL8

In Red Hat Enterprise Linux 8, LVM uses the LVM lock daemon lvmlockd instead of clvmd for managing shared storage devices in an active/active cluster. This requires that you configure the logical volumes that your active/active cluster will require as shared logical volumes. Additionally, this requires that you use the LVM-activate resource to manage an LVM volume and that you use the lvmlockd resource agent to manage the lvmlockd daemon. See Configuring a GFS2 file system in a cluster for a full procedure for configuring a Pacemaker cluster that includes GFS2 file systems using shared logical volumes.

To use your existing Red Hat Enterprise Linux 7 logical volumes when configuring a RHEL8 cluster that includes GFS2 file systems, perform the following procedure from the RHEL8 cluster. In this example, the clustered RHEL 7 logical volume is part of the volume group upgrade_gfs_vg.

Note

The RHEL8 cluster must have the same name as the RHEL7 cluster that includes the GFS2 file system in order for the existing file system to be valid.

  1. Ensure that the logical volumes containing the GFS2 file systems are currently inactive. This procedure is safe only if all nodes have stopped using the volume group.
  2. From one node in the cluster, forcibly change the volume group to be local.

    [root@rhel8-01 ~]# vgchange --lock-type none --lock-opt force upgrade_gfs_vg
    Forcibly change VG lock type to none? [y/n]: y
      Volume group "upgrade_gfs_vg" successfully changed
  3. From one node in the cluster, change the local volume group to a shared volume group

    [root@rhel8-01 ~]# vgchange --lock-type dlm upgrade_gfs_vg
       Volume group "upgrade_gfs_vg" successfully changed
  4. On each node in the cluster, start locking for the volume group.

    [root@rhel8-01 ~]# vgchange --lock-start upgrade_gfs_vg
      VG upgrade_gfs_vg starting dlm lockspace
      Starting locking.  Waiting until locks are ready...
    [root@rhel8-02 ~]# vgchange --lock-start upgrade_gfs_vg
      VG upgrade_gfs_vg starting dlm lockspace
      Starting locking.  Waiting until locks are ready...

After performing this procedure, you can create an LVM-activate resource for each logical volume.