1MOUNT.CEPH(8) Ceph MOUNT.CEPH(8)
2
6 mount.ceph - mount a Ceph file system
7
9 mount.ceph name*@*fsid.*fs_name*=/[subdir] dir [-o options ]
10
13 mount.ceph is a helper for mounting the Ceph file system on a Linux
14 host. It serves to resolve monitor hostname(s) into IP addresses and
15 read authentication keys from disk; the Linux kernel client component
16 does most of the real work. To mount a Ceph file system use:
17 mount.ceph name@07fe3187-00d9-42a3-814b-72a4d5e7d5be.fs_name=/ /mnt/mycephfs -o mon_addr=1.2.3.4
19 Mount helper can fill in the cluster FSID by reading the ceph configu‐
21 ration file. Its recommended to call the mount helper via mount(8) as
22 per:
23 mount -t ceph name@.fs_name=/ /mnt/mycephfs -o mon_addr=1.2.3.4
25 Note that the dot . still needs to be a part of the device string in
27 this case.
28 The first argument is the device part of the mount command. It includes
30 the RADOS user for authentication, the file system name and a path
31 within CephFS that will be mounted at the mount point.
32 Monitor addresses can be passed using mon_addr mount option. Multiple
34 monitor addresses can be passed by separating addresses with a slash
35 (/). Only one monitor is needed to mount successfully; the client will
36 learn about all monitors from any responsive monitor. However, it is a
37 good idea to specify more than one in case the one happens to be down
38 at the time of mount. Monitor addresses takes the form ip_ad‐
39 dress[:port]. If the port is not specified, the Ceph default of 6789 is
40 assumed.
41 If monitor addresses are not specified, then mount.ceph will attempt to
43 determine monitor addresses using local configuration files and/or DNS
44 SRV records. In similar way, if authentication is enabled on Ceph clus‐
45 ter (which is done using CephX) and options secret and secretfile are
46 not specified in the command, the mount helper will spawn a child
47 process that will use the standard Ceph library routines to find a
48 keyring and fetch the secret from it (including the monitor address and
49 FSID if those not specified).
50 A sub-directory of the file system can be mounted by specifying the
52 (absolute) path to the sub-directory right after "=" in the device part
53 of the mount command.
54 Mount helper application conventions dictate that the first two options
56 are device to be mounted and the mountpoint for that device. Options
57 must be passed only after these fixed arguments.
58
60 Basic
61 conf Path to a ceph.conf file. This is used to initialize the Ceph
62 context for autodiscovery of monitor addresses and auth secrets.
63 The default is to use the standard search path for ceph.conf
64 files.
65 mount_timeout
67 int (seconds), Default: 60
68 ms_mode=<legacy|crc|secure|prefer-crc|prefer-secure>
70 Set the connection mode that the client uses for transport. The
71 available modes are:
72 • legacy: use messenger v1 protocol to talk to the cluster
74 • crc: use messenger v2, without on-the-wire encryption
76 • secure: use messenger v2, with on-the-wire encryption
78 • prefer-crc: crc mode, if denied agree to secure mode
80 • prefer-secure: secure mode, if denied agree to crc mode
82 mon_addr
84 Monitor address of the cluster in the form of ip_address[:port]
85 fsid Cluster FSID. This can be found using ceph fsid command.
87 secret secret key for use with CephX. This option is insecure because
89 it exposes the secret on the command line. To avoid this, use
90 the secretfile option.
91 secretfile
93 path to file containing the secret key to use with CephX
94 recover_session=<no|clean>
96 Set auto reconnect mode in the case where the client is block‐
97 listed. The available modes are no and clean. The default is no.
98 • no: never attempt to reconnect when client detects that it has
100 been blocklisted. Blocklisted clients will not attempt to re‐
101 connect and their operations will fail too.
102 • clean: client reconnects to the Ceph cluster automatically
104 when it detects that it has been blocklisted. During recon‐
105 nect, client drops dirty data/metadata, invalidates page
106 caches and writable file handles. After reconnect, file locks
107 become stale because the MDS loses track of them. If an inode
108 contains any stale file locks, read/write on the inode is not
109 allowed until applications release all stale file locks.
110 Advanced
112 cap_release_safety
113 int, Default: calculated
114 caps_wanted_delay_max
116 int, cap release delay, Default: 60
117 caps_wanted_delay_min
119 int, cap release delay, Default: 5
120 dirstat
122 funky cat dirname for stats, Default: off
123 nodirstat
125 no funky cat dirname for stats
126 ip my ip
128 noasyncreaddir
130 no dcache readdir
131 nocrc no data crc on writes
133 noshare
135 create a new client instance, instead of sharing an existing in‐
136 stance of a client mounting the same cluster
137 osdkeepalive
139 int, Default: 5
140 osd_idle_ttl
142 int (seconds), Default: 60
143 rasize int (bytes), max readahead. Default: 8388608 (8192*1024)
145 rbytes Report the recursive size of the directory contents for st_size
147 on directories. Default: off
148 norbytes
150 Do not report the recursive size of the directory contents for
151 st_size on directories.
152 readdir_max_bytes
154 int, Default: 524288 (512*1024)
155 readdir_max_entries
157 int, Default: 1024
158 rsize int (bytes), max read size. Default: 16777216 (16*1024*1024)
160 snapdirname
162 string, set the name of the hidden snapdir. Default: .snap
163 write_congestion_kb
165 int (kb), max writeback in flight. scale with available memory.
166 Default: calculated from available memory
167 wsize int (bytes), max write size. Default: 16777216 (16*1024*1024)
169 (writeback uses smaller of wsize and stripe unit)
170 wsync Execute all namespace operations synchronously. This ensures
172 that the namespace operation will only complete after receiving
173 a reply from the MDS. This is the default.
174 nowsync
176 Allow the client to do namespace operations asynchronously. When
177 this option is enabled, a namespace operation may complete be‐
178 fore the MDS replies, if it has sufficient capabilities to do
179 so.
180
182 Mount the full file system:
183 mount -t ceph fs_user@.mycephfs2=/ /mnt/mycephfs
185 Mount only part of the namespace/file system:
187 mount.ceph fs_user@.mycephfs2=/some/directory/in/cephfs /mnt/mycephfs
189 Pass the monitor host's IP address, optionally:
191 mount.ceph fs_user@.mycephfs2=/ /mnt/mycephfs -o mon_addr=192.168.0.1
193 Pass the port along with IP address if it's running on a non-standard
195 port:
196 mount.ceph fs_user@.mycephfs2=/ /mnt/mycephfs -o mon_addr=192.168.0.1:7000
198 If there are multiple monitors, pass each address separated by a /:
200 mount.ceph fs_user@.mycephfs2=/ /mnt/mycephfs -o mon_addr=192.168.0.1/192.168.0.2/192.168.0.3
202 Pass secret key for CephX user optionally:
204 mount.ceph fs_user@.mycephfs2=/ /mnt/mycephfs -o secret=AQATSKdNGBnwLhAAnNDKnH65FmVKpXZJVasUeQ==
206 Pass file containing secret key to avoid leaving secret key in shell's
208 command history:
209 mount.ceph fs_user@.mycephfs2=/ /mnt/mycephfs -o secretfile=/etc/ceph/fs_username.secret
211 If authentication is disabled on Ceph cluster, omit the credential re‐
213 lated option:
214 mount.ceph fs_user@.mycephfs2=/ /mnt/mycephfs
216
218 mount.ceph is part of Ceph, a massively scalable, open-source, distrib‐
219 uted storage system. Please refer to the Ceph documentation at
220 https://docs.ceph.com for more information.
221
223 The recover_session= option was added to mainline Linux kernels in
224 v5.4. wsync and nowsync were added in v5.7.
225
227 ceph-fuse(8), ceph(8)
228
230 2010-2022, Inktank Storage, Inc. and contributors. Licensed under Cre‐
231 ative Commons Attribution Share Alike 3.0 (CC-BY-SA-3.0)
232dev Oct 18, 2022 MOUNT.CEPH(8)