3 ========================================
4 ceph-osd -- ceph object storage daemon
5 ========================================
12 | **ceph-osd** -i *osdnum* [ --osd-data *datapath* ] [ --osd-journal
13 *journal* ] [ --mkfs ] [ --mkjournal ] [--flush-journal] [--check-allows-journal] [--check-wants-journal] [--check-needs-journal] [ --mkkey ]
19 **ceph-osd** is the object storage daemon for the Ceph distributed file
20 system. It is responsible for storing objects on a local file system
21 and providing access to them over the network.
23 The datapath argument should be a directory on a xfs file system
24 where the object data resides. The journal is optional, and is only
25 useful performance-wise when it resides on a different disk than
26 datapath with low latency (ideally, an NVRAM device).
32 .. option:: -f, --foreground
34 Foreground: do not daemonize after startup (run in foreground). Do
35 not generate a pid file. Useful when run via :doc:`ceph-run <ceph-run>`\(8).
39 Debug mode: like ``-f``, but also send all log output to stderr.
41 .. option:: --setuser userorgid
43 Set uid after starting. If a username is specified, the user
44 record is looked up to get a uid and a gid, and the gid is also set
45 as well, unless --setgroup is also specified.
47 .. option:: --setgroup grouporgid
49 Set gid after starting. If a group name is specified the group
50 record is looked up to get a gid.
52 .. option:: --osd-data osddata
54 Use object store at *osddata*.
56 .. option:: --osd-journal journal
58 Journal updates to *journal*.
60 .. option:: --check-wants-journal
62 Check whether a journal is desired.
64 .. option:: --check-allows-journal
66 Check whether a journal is allowed.
68 .. option:: --check-needs-journal
70 Check whether a journal is required.
74 Create an empty object repository. This also initializes the journal
79 Generate a new secret key. This is normally used in combination
80 with ``--mkfs`` as it is more convenient than generating a key by
81 hand with :doc:`ceph-authtool <ceph-authtool>`\(8).
83 .. option:: --mkjournal
85 Create a new journal file to match an existing object repository.
86 This is useful if the journal device or file is wiped out due to a
87 disk or file system failure.
89 .. option:: --flush-journal
91 Flush the journal to permanent store. This runs in the foreground
92 so you know when it's completed. This can be useful if you want to
93 resize the journal or need to otherwise destroy it: this guarantees
96 .. option:: --get-cluster-fsid
98 Print the cluster fsid (uuid) and exit.
100 .. option:: --get-osd-fsid
102 Print the OSD's fsid and exit. The OSD's uuid is generated at
103 --mkfs time and is thus unique to a particular instantiation of
106 .. option:: --get-journal-fsid
108 Print the journal's uuid. The journal fsid is set to match the OSD
111 .. option:: -c ceph.conf, --conf=ceph.conf
113 Use *ceph.conf* configuration file instead of the default
114 ``/etc/ceph/ceph.conf`` for runtime configuration options.
116 .. option:: -m monaddress[:port]
118 Connect to specified monitor (instead of looking through
125 **ceph-osd** is part of Ceph, a massively scalable, open-source, distributed storage system. Please refer to
126 the Ceph documentation at http://ceph.com/docs for more information.
131 :doc:`ceph <ceph>`\(8),
132 :doc:`ceph-mds <ceph-mds>`\(8),
133 :doc:`ceph-mon <ceph-mon>`\(8),
134 :doc:`ceph-authtool <ceph-authtool>`\(8)