How to Manually Create an Nfsshare On a Volume After Cluster Start

[XaverStiensmeier]

1. Create a new volume under OpenStack Project/Volumes/Volumes.
2. Through Edit Volume, you can now select Manage Attachments and choose your master instance there. Attached To in the list will then show where the volume was mounted. For me, for example, /dev/vdc.
3. On the master: you can display the available disks using sudo lshw -class disk -short. The mounted disk should now appear there.
4. I format my newly mounted volume (if already formatted, this will overwrite all existing data!) with mkfs.ext4 /dev/vdc.
5. I mount the volume on /vol/spool2. For this, I first create the directory sudo mkdir -p /vol/spool2 and then mount it sudo mount /dev/vdc /vol/spool2. Since my volume uses ext4, I will find a lost+found directory at /vol/spool2 showing me that mounting has been successful.
6. Now I want to set up NFS and do this through BiBiGrid, so I don't have to manually change each started worker myself. For this, I ensure that all workers are shut down. I then modify the playbook/vars/common_configuration.yml.

nfs_mounts:
- dst: //vol/spool
  src: //vol/spool

becomes

nfs_mounts:
- dst: //vol/spool
  src: //vol/spool
- dst: //vol/spool2
  src: //vol/spool2

By executing bibiplay -l master -t nfs, Ansible will re-setup nfs on the master. The new nfsshare is now setup.

showmount -e localhost confirms that both nfsshares - the old and the new - are now available:

Export list for localhost:
/vol/spool2 192.168.20.0/24
/vol/spool  192.168.20.0/24

We are done. Workers that power up now, also have the second nfsshare.