projects / ceph.git / blame
| Commit | Line | Data |
|---|---|---|
| 7c673cae FG |
1 | :orphan: |
| 2 | ||
| 3 | ========================================= | |
| 4 | ceph-mds -- ceph metadata server daemon | |
| 5 | ========================================= | |
| 6 | ||
| 7 | .. program:: ceph-mds | |
| 8 | ||
| 9 | Synopsis | |
| 10 | ======== | |
| 11 | ||
| 12 | | **ceph-mds** -i *name* [ --hot-standby [*rank*] ] | |
| 13 | ||
| 14 | ||
| 15 | Description | |
| 16 | =========== | |
| 17 | ||
| 18 | **ceph-mds** is the metadata server daemon for the Ceph distributed file | |
| 19 | system. One or more instances of ceph-mds collectively manage the file | |
| 20 | system namespace, coordinating access to the shared OSD cluster. | |
| 21 | ||
| 22 | Each ceph-mds daemon instance should have a unique name. The name is used | |
| 23 | to identify daemon instances in the ceph.conf. | |
| 24 | ||
| 25 | Once the daemon has started, the monitor cluster will normally assign | |
| 26 | it a logical rank, or put it in a standby pool to take over for | |
| 27 | another daemon that crashes. Some of the specified options can cause | |
| 28 | other behaviors. | |
| 29 | ||
| 30 | If you specify hot-standby, you must either specify the rank on the command | |
| 31 | line, or specify one of the mds_standby_for_[rank|name] parameters in the | |
| 32 | config. The command line specification overrides the config, and specifying | |
| 33 | the rank overrides specifying the name. | |
| 34 | ||
| 35 | ||
| 36 | Options | |
| 37 | ======= | |
| 38 | ||
| 39 | .. option:: -f, --foreground | |
| 40 | ||
| 41 | Foreground: do not daemonize after startup (run in foreground). Do | |
| 42 | not generate a pid file. Useful when run via :doc:`ceph-run | |
| 43 | <ceph-run>`\(8). | |
| 44 | ||
| 45 | .. option:: -d | |
| 46 | ||
| 47 | Debug mode: like ``-f``, but also send all log output to stderr. | |
| 48 | ||
| 49 | .. option:: --setuser userorgid | |
| 50 | ||
| 51 | Set uid after starting. If a username is specified, the user | |
| 52 | record is looked up to get a uid and a gid, and the gid is also set | |
| 53 | as well, unless --setgroup is also specified. | |
| 54 | ||
| 55 | .. option:: --setgroup grouporgid | |
| 56 | ||
| 57 | Set gid after starting. If a group name is specified the group | |
| 58 | record is looked up to get a gid. | |
| 59 | ||
| 60 | .. option:: -c ceph.conf, --conf=ceph.conf | |
| 61 | ||
| 62 | Use *ceph.conf* configuration file instead of the default | |
| 63 | ``/etc/ceph/ceph.conf`` to determine monitor addresses during | |
| 64 | startup. | |
| 65 | ||
| 66 | .. option:: -m monaddress[:port] | |
| 67 | ||
| 68 | Connect to specified monitor (instead of looking through | |
| 69 | ``ceph.conf``). | |
| 70 | ||
| 71 | .. option:: --hot-standby <rank> | |
| 72 | ||
| 73 | Start as a hot standby for MDS <rank>. | |
| 74 | ||
| 75 | Availability | |
| 76 | ============ | |
| 77 | ||
| 78 | **ceph-mds** is part of Ceph, a massively scalable, open-source, distributed storage system. Please refer to the Ceph documentation at | |
| 79 | http://ceph.com/docs for more information. | |
| 80 | ||
| 81 | ||
| 82 | See also | |
| 83 | ======== | |
| 84 | ||
| 85 | :doc:`ceph <ceph>`\(8), | |
| 86 | :doc:`ceph-mon <ceph-mon>`\(8), | |
| 87 | :doc:`ceph-osd <ceph-osd>`\(8) |