diff options
Diffstat (limited to '')
-rw-r--r-- | doc/cephfs/administration.rst | 375 |
1 files changed, 375 insertions, 0 deletions
diff --git a/doc/cephfs/administration.rst b/doc/cephfs/administration.rst new file mode 100644 index 000000000..966b45ee5 --- /dev/null +++ b/doc/cephfs/administration.rst @@ -0,0 +1,375 @@ +.. _cephfs-administration: + +CephFS Administrative commands +============================== + +File Systems +------------ + +.. note:: The names of the file systems, metadata pools, and data pools can + only have characters in the set [a-zA-Z0-9\_-.]. + +These commands operate on the CephFS file systems in your Ceph cluster. +Note that by default only one file system is permitted: to enable +creation of multiple file systems use ``ceph fs flag set enable_multiple true``. + +:: + + fs new <file system name> <metadata pool name> <data pool name> + +This command creates a new file system. The file system name and metadata pool +name are self-explanatory. The specified data pool is the default data pool and +cannot be changed once set. Each file system has its own set of MDS daemons +assigned to ranks so ensure that you have sufficient standby daemons available +to accommodate the new file system. + +:: + + fs ls + +List all file systems by name. + +:: + + fs dump [epoch] + +This dumps the FSMap at the given epoch (default: current) which includes all +file system settings, MDS daemons and the ranks they hold, and the list of +standby MDS daemons. + + +:: + + fs rm <file system name> [--yes-i-really-mean-it] + +Destroy a CephFS file system. This wipes information about the state of the +file system from the FSMap. The metadata pool and data pools are untouched and +must be destroyed separately. + +:: + + fs get <file system name> + +Get information about the named file system, including settings and ranks. This +is a subset of the same information from the ``fs dump`` command. + +:: + + fs set <file system name> <var> <val> + +Change a setting on a file system. These settings are specific to the named +file system and do not affect other file systems. + +:: + + fs add_data_pool <file system name> <pool name/id> + +Add a data pool to the file system. This pool can be used for file layouts +as an alternate location to store file data. + +:: + + fs rm_data_pool <file system name> <pool name/id> + +This command removes the specified pool from the list of data pools for the +file system. If any files have layouts for the removed data pool, the file +data will become unavailable. The default data pool (when creating the file +system) cannot be removed. + + +Settings +-------- + +:: + + fs set <fs name> max_file_size <size in bytes> + +CephFS has a configurable maximum file size, and it's 1TB by default. +You may wish to set this limit higher if you expect to store large files +in CephFS. It is a 64-bit field. + +Setting ``max_file_size`` to 0 does not disable the limit. It would +simply limit clients to only creating empty files. + + +Maximum file sizes and performance +---------------------------------- + +CephFS enforces the maximum file size limit at the point of appending to +files or setting their size. It does not affect how anything is stored. + +When users create a file of an enormous size (without necessarily +writing any data to it), some operations (such as deletes) cause the MDS +to have to do a large number of operations to check if any of the RADOS +objects within the range that could exist (according to the file size) +really existed. + +The ``max_file_size`` setting prevents users from creating files that +appear to be eg. exabytes in size, causing load on the MDS as it tries +to enumerate the objects during operations like stats or deletes. + + +Taking the cluster down +----------------------- + +Taking a CephFS cluster down is done by setting the down flag: + +:: + + fs set <fs_name> down true + +To bring the cluster back online: + +:: + + fs set <fs_name> down false + +This will also restore the previous value of max_mds. MDS daemons are brought +down in a way such that journals are flushed to the metadata pool and all +client I/O is stopped. + + +Taking the cluster down rapidly for deletion or disaster recovery +----------------------------------------------------------------- + +To allow rapidly deleting a file system (for testing) or to quickly bring the +file system and MDS daemons down, use the ``fs fail`` command: + +:: + + fs fail <fs_name> + +This command sets a file system flag to prevent standbys from +activating on the file system (the ``joinable`` flag). + +This process can also be done manually by doing the following: + +:: + + fs set <fs_name> joinable false + +Then the operator can fail all of the ranks which causes the MDS daemons to +respawn as standbys. The file system will be left in a degraded state. + +:: + + # For all ranks, 0-N: + mds fail <fs_name>:<n> + +Once all ranks are inactive, the file system may also be deleted or left in +this state for other purposes (perhaps disaster recovery). + +To bring the cluster back up, simply set the joinable flag: + +:: + + fs set <fs_name> joinable true + + +Daemons +------- + +Most commands manipulating MDSs take a ``<role>`` argument which can take one +of three forms: + +:: + + <fs_name>:<rank> + <fs_id>:<rank> + <rank> + +Commands to manipulate MDS daemons: + +:: + + mds fail <gid/name/role> + +Mark an MDS daemon as failed. This is equivalent to what the cluster +would do if an MDS daemon had failed to send a message to the mon +for ``mds_beacon_grace`` second. If the daemon was active and a suitable +standby is available, using ``mds fail`` will force a failover to the standby. + +If the MDS daemon was in reality still running, then using ``mds fail`` +will cause the daemon to restart. If it was active and a standby was +available, then the "failed" daemon will return as a standby. + + +:: + + tell mds.<daemon name> command ... + +Send a command to the MDS daemon(s). Use ``mds.*`` to send a command to all +daemons. Use ``ceph tell mds.* help`` to learn available commands. + +:: + + mds metadata <gid/name/role> + +Get metadata about the given MDS known to the Monitors. + +:: + + mds repaired <role> + +Mark the file system rank as repaired. Unlike the name suggests, this command +does not change a MDS; it manipulates the file system rank which has been +marked damaged. + + +Required Client Features +------------------------ + +It is sometimes desirable to set features that clients must support to talk to +CephFS. Clients without those features may disrupt other clients or behave in +surprising ways. Or, you may want to require newer features to prevent older +and possibly buggy clients from connecting. + +Commands to manipulate required client features of a file system: + +:: + + fs required_client_features <fs name> add reply_encoding + fs required_client_features <fs name> rm reply_encoding + +To list all CephFS features + +:: + + fs feature ls + +Clients that are missing newly added features will be evicted automatically. + +Here are the current CephFS features and first release they came out: + ++------------------+--------------+-----------------+ +| Feature | Ceph release | Upstream Kernel | ++==================+==============+=================+ +| jewel | jewel | 4.5 | ++------------------+--------------+-----------------+ +| kraken | kraken | 4.13 | ++------------------+--------------+-----------------+ +| luminous | luminous | 4.13 | ++------------------+--------------+-----------------+ +| mimic | mimic | 4.19 | ++------------------+--------------+-----------------+ +| reply_encoding | nautilus | 5.1 | ++------------------+--------------+-----------------+ +| reclaim_client | nautilus | N/A | ++------------------+--------------+-----------------+ +| lazy_caps_wanted | nautilus | 5.1 | ++------------------+--------------+-----------------+ +| multi_reconnect | nautilus | 5.1 | ++------------------+--------------+-----------------+ +| deleg_ino | octopus | 5.6 | ++------------------+--------------+-----------------+ +| metric_collect | pacific | N/A | ++------------------+--------------+-----------------+ +| alternate_name | pacific | PLANNED | ++------------------+--------------+-----------------+ + +CephFS Feature Descriptions + + +:: + + reply_encoding + +MDS encodes request reply in extensible format if client supports this feature. + + +:: + + reclaim_client + +MDS allows new client to reclaim another (dead) client's states. This feature +is used by NFS-Ganesha. + + +:: + + lazy_caps_wanted + +When a stale client resumes, if the client supports this feature, mds only needs +to re-issue caps that are explicitly wanted. + + +:: + + multi_reconnect + +When mds failover, client sends reconnect messages to mds, to reestablish cache +states. If MDS supports this feature, client can split large reconnect message +into multiple ones. + + +:: + + deleg_ino + +MDS delegate inode numbers to client if client supports this feature. Having +delegated inode numbers is a prerequisite for client to do async file creation. + + +:: + + metric_collect + +Clients can send performance metric to MDS if MDS support this feature. + +:: + + alternate_name + +Clients can set and understand "alternate names" for directory entries. This is +to be used for encrypted file name support. + + +Global settings +--------------- + + +:: + + fs flag set <flag name> <flag val> [<confirmation string>] + +Sets a global CephFS flag (i.e. not specific to a particular file system). +Currently, the only flag setting is 'enable_multiple' which allows having +multiple CephFS file systems. + +Some flags require you to confirm your intentions with "--yes-i-really-mean-it" +or a similar string they will prompt you with. Consider these actions carefully +before proceeding; they are placed on especially dangerous activities. + +.. _advanced-cephfs-admin-settings: + +Advanced +-------- + +These commands are not required in normal operation, and exist +for use in exceptional circumstances. Incorrect use of these +commands may cause serious problems, such as an inaccessible +file system. + +:: + + mds rmfailed + +This removes a rank from the failed set. + +:: + + fs reset <file system name> + +This command resets the file system state to defaults, except for the name and +pools. Non-zero ranks are saved in the stopped set. + + +:: + + fs new <file system name> <metadata pool name> <data pool name> --fscid <fscid> --force + +This command creates a file system with a specific **fscid** (file system cluster ID). +You may want to do this when an application expects the file system's ID to be +stable after it has been recovered, e.g., after monitor databases are lost and +rebuilt. Consequently, file system IDs don't always keep increasing with newer +file systems. |