Taking down a CephFS cluster

You can take down Ceph File System (CephFS) cluster by setting the down flag to true. Setting the flag to true gracefully shuts down the Metadata Server (MDS) daemons by flushing journals to the metadata pool and stopping all client I/O.

Before you begin

Be sure that you have root-level access to a Ceph Monitor node.

About this task

You can take the CephFS cluster down quickly to test the deletion of a file system and bring the Metadata Server (MDS) daemons down. For example, when practicing a disaster recovery scenario. Doing this sets the jointable flag to prevent the MDS standby daemons from activating the file system.

Procedure

Use one of the following procedures to take down a CephFS cluster. To bring the CephFS cluster back up, see Bringing the CephFS cluster back up.
  • Mark the CephFS cluster down.
    ceph fs set FS_NAME down true
    For example:
    [ceph: root@host01 /]# ceph fs set cephfs down true
  • Quickly take down a CephFS cluster.
    ceph fs fail FS_NAME
    For example:
    [ceph: root@host01 /]# ceph fs fail cephfs

Bringing the CephFS cluster back up

Procedure

Use one of the following procedures to bring a CephFS cluster back up after taking it down.
  • Bring the CephFS cluster back up.
    ceph fs set FS_NAME down false
    For example:
    [ceph: root@host01 /]# ceph fs set cephfs down false
  • Get the CephFS cluster back up, by setting cephfs to joinable.
    ceph fs set FS_NAME joinable true
    For example:
    [ceph: root@host01 /]# ceph fs set cephfs joinable true

cephfs marked joinable; MDS may join as newly active.