Loading...
1.. SPDX-License-Identifier: GPL-2.0
2
3============================
4Ceph Distributed File System
5============================
6
7Ceph is a distributed network file system designed to provide good
8performance, reliability, and scalability.
9
10Basic features include:
11
12 * POSIX semantics
13 * Seamless scaling from 1 to many thousands of nodes
14 * High availability and reliability. No single point of failure.
15 * N-way replication of data across storage nodes
16 * Fast recovery from node failures
17 * Automatic rebalancing of data on node addition/removal
18 * Easy deployment: most FS components are userspace daemons
19
20Also,
21
22 * Flexible snapshots (on any directory)
23 * Recursive accounting (nested files, directories, bytes)
24
25In contrast to cluster filesystems like GFS, OCFS2, and GPFS that rely
26on symmetric access by all clients to shared block devices, Ceph
27separates data and metadata management into independent server
28clusters, similar to Lustre. Unlike Lustre, however, metadata and
29storage nodes run entirely as user space daemons. File data is striped
30across storage nodes in large chunks to distribute workload and
31facilitate high throughputs. When storage nodes fail, data is
32re-replicated in a distributed fashion by the storage nodes themselves
33(with some minimal coordination from a cluster monitor), making the
34system extremely efficient and scalable.
35
36Metadata servers effectively form a large, consistent, distributed
37in-memory cache above the file namespace that is extremely scalable,
38dynamically redistributes metadata in response to workload changes,
39and can tolerate arbitrary (well, non-Byzantine) node failures. The
40metadata server takes a somewhat unconventional approach to metadata
41storage to significantly improve performance for common workloads. In
42particular, inodes with only a single link are embedded in
43directories, allowing entire directories of dentries and inodes to be
44loaded into its cache with a single I/O operation. The contents of
45extremely large directories can be fragmented and managed by
46independent metadata servers, allowing scalable concurrent access.
47
48The system offers automatic data rebalancing/migration when scaling
49from a small cluster of just a few nodes to many hundreds, without
50requiring an administrator carve the data set into static volumes or
51go through the tedious process of migrating data between servers.
52When the file system approaches full, new nodes can be easily added
53and things will "just work."
54
55Ceph includes flexible snapshot mechanism that allows a user to create
56a snapshot on any subdirectory (and its nested contents) in the
57system. Snapshot creation and deletion are as simple as 'mkdir
58.snap/foo' and 'rmdir .snap/foo'.
59
60Snapshot names have two limitations:
61
62* They can not start with an underscore ('_'), as these names are reserved
63 for internal usage by the MDS.
64* They can not exceed 240 characters in size. This is because the MDS makes
65 use of long snapshot names internally, which follow the format:
66 `_<SNAPSHOT-NAME>_<INODE-NUMBER>`. Since filenames in general can't have
67 more than 255 characters, and `<node-id>` takes 13 characters, the long
68 snapshot names can take as much as 255 - 1 - 1 - 13 = 240.
69
70Ceph also provides some recursive accounting on directories for nested files
71and bytes. You can run the commands::
72
73 getfattr -n ceph.dir.rfiles /some/dir
74 getfattr -n ceph.dir.rbytes /some/dir
75
76to get the total number of nested files and their combined size, respectively.
77This makes the identification of large disk space consumers relatively quick,
78as no 'du' or similar recursive scan of the file system is required.
79
80Finally, Ceph also allows quotas to be set on any directory in the system.
81The quota can restrict the number of bytes or the number of files stored
82beneath that point in the directory hierarchy. Quotas can be set using
83extended attributes 'ceph.quota.max_files' and 'ceph.quota.max_bytes', eg::
84
85 setfattr -n ceph.quota.max_bytes -v 100000000 /some/dir
86 getfattr -n ceph.quota.max_bytes /some/dir
87
88A limitation of the current quotas implementation is that it relies on the
89cooperation of the client mounting the file system to stop writers when a
90limit is reached. A modified or adversarial client cannot be prevented
91from writing as much data as it needs.
92
93Mount Syntax
94============
95
96The basic mount syntax is::
97
98 # mount -t ceph user@fsid.fs_name=/[subdir] mnt -o mon_addr=monip1[:port][/monip2[:port]]
99
100You only need to specify a single monitor, as the client will get the
101full list when it connects. (However, if the monitor you specify
102happens to be down, the mount won't succeed.) The port can be left
103off if the monitor is using the default. So if the monitor is at
1041.2.3.4::
105
106 # mount -t ceph cephuser@07fe3187-00d9-42a3-814b-72a4d5e7d5be.cephfs=/ /mnt/ceph -o mon_addr=1.2.3.4
107
108is sufficient. If /sbin/mount.ceph is installed, a hostname can be
109used instead of an IP address and the cluster FSID can be left out
110(as the mount helper will fill it in by reading the ceph configuration
111file)::
112
113 # mount -t ceph cephuser@cephfs=/ /mnt/ceph -o mon_addr=mon-addr
114
115Multiple monitor addresses can be passed by separating each address with a slash (`/`)::
116
117 # mount -t ceph cephuser@cephfs=/ /mnt/ceph -o mon_addr=192.168.1.100/192.168.1.101
118
119When using the mount helper, monitor address can be read from ceph
120configuration file if available. Note that, the cluster FSID (passed as part
121of the device string) is validated by checking it with the FSID reported by
122the monitor.
123
124Mount Options
125=============
126
127 mon_addr=ip_address[:port][/ip_address[:port]]
128 Monitor address to the cluster. This is used to bootstrap the
129 connection to the cluster. Once connection is established, the
130 monitor addresses in the monitor map are followed.
131
132 fsid=cluster-id
133 FSID of the cluster (from `ceph fsid` command).
134
135 ip=A.B.C.D[:N]
136 Specify the IP and/or port the client should bind to locally.
137 There is normally not much reason to do this. If the IP is not
138 specified, the client's IP address is determined by looking at the
139 address its connection to the monitor originates from.
140
141 wsize=X
142 Specify the maximum write size in bytes. Default: 64 MB.
143
144 rsize=X
145 Specify the maximum read size in bytes. Default: 64 MB.
146
147 rasize=X
148 Specify the maximum readahead size in bytes. Default: 8 MB.
149
150 mount_timeout=X
151 Specify the timeout value for mount (in seconds), in the case
152 of a non-responsive Ceph file system. The default is 60
153 seconds.
154
155 caps_max=X
156 Specify the maximum number of caps to hold. Unused caps are released
157 when number of caps exceeds the limit. The default is 0 (no limit)
158
159 rbytes
160 When stat() is called on a directory, set st_size to 'rbytes',
161 the summation of file sizes over all files nested beneath that
162 directory. This is the default.
163
164 norbytes
165 When stat() is called on a directory, set st_size to the
166 number of entries in that directory.
167
168 nocrc
169 Disable CRC32C calculation for data writes. If set, the storage node
170 must rely on TCP's error correction to detect data corruption
171 in the data payload.
172
173 dcache
174 Use the dcache contents to perform negative lookups and
175 readdir when the client has the entire directory contents in
176 its cache. (This does not change correctness; the client uses
177 cached metadata only when a lease or capability ensures it is
178 valid.)
179
180 nodcache
181 Do not use the dcache as above. This avoids a significant amount of
182 complex code, sacrificing performance without affecting correctness,
183 and is useful for tracking down bugs.
184
185 noasyncreaddir
186 Do not use the dcache as above for readdir.
187
188 noquotadf
189 Report overall filesystem usage in statfs instead of using the root
190 directory quota.
191
192 nocopyfrom
193 Don't use the RADOS 'copy-from' operation to perform remote object
194 copies. Currently, it's only used in copy_file_range, which will revert
195 to the default VFS implementation if this option is used.
196
197 recover_session=<no|clean>
198 Set auto reconnect mode in the case where the client is blocklisted. The
199 available modes are "no" and "clean". The default is "no".
200
201 * no: never attempt to reconnect when client detects that it has been
202 blocklisted. Operations will generally fail after being blocklisted.
203
204 * clean: client reconnects to the ceph cluster automatically when it
205 detects that it has been blocklisted. During reconnect, client drops
206 dirty data/metadata, invalidates page caches and writable file handles.
207 After reconnect, file locks become stale because the MDS loses track
208 of them. If an inode contains any stale file locks, read/write on the
209 inode is not allowed until applications release all stale file locks.
210
211More Information
212================
213
214For more information on Ceph, see the home page at
215 https://ceph.com/
216
217The Linux kernel client source tree is available at
218 - https://github.com/ceph/ceph-client.git
219
220and the source for the full system is at
221 https://github.com/ceph/ceph.git
1.. SPDX-License-Identifier: GPL-2.0
2
3============================
4Ceph Distributed File System
5============================
6
7Ceph is a distributed network file system designed to provide good
8performance, reliability, and scalability.
9
10Basic features include:
11
12 * POSIX semantics
13 * Seamless scaling from 1 to many thousands of nodes
14 * High availability and reliability. No single point of failure.
15 * N-way replication of data across storage nodes
16 * Fast recovery from node failures
17 * Automatic rebalancing of data on node addition/removal
18 * Easy deployment: most FS components are userspace daemons
19
20Also,
21
22 * Flexible snapshots (on any directory)
23 * Recursive accounting (nested files, directories, bytes)
24
25In contrast to cluster filesystems like GFS, OCFS2, and GPFS that rely
26on symmetric access by all clients to shared block devices, Ceph
27separates data and metadata management into independent server
28clusters, similar to Lustre. Unlike Lustre, however, metadata and
29storage nodes run entirely as user space daemons. File data is striped
30across storage nodes in large chunks to distribute workload and
31facilitate high throughputs. When storage nodes fail, data is
32re-replicated in a distributed fashion by the storage nodes themselves
33(with some minimal coordination from a cluster monitor), making the
34system extremely efficient and scalable.
35
36Metadata servers effectively form a large, consistent, distributed
37in-memory cache above the file namespace that is extremely scalable,
38dynamically redistributes metadata in response to workload changes,
39and can tolerate arbitrary (well, non-Byzantine) node failures. The
40metadata server takes a somewhat unconventional approach to metadata
41storage to significantly improve performance for common workloads. In
42particular, inodes with only a single link are embedded in
43directories, allowing entire directories of dentries and inodes to be
44loaded into its cache with a single I/O operation. The contents of
45extremely large directories can be fragmented and managed by
46independent metadata servers, allowing scalable concurrent access.
47
48The system offers automatic data rebalancing/migration when scaling
49from a small cluster of just a few nodes to many hundreds, without
50requiring an administrator carve the data set into static volumes or
51go through the tedious process of migrating data between servers.
52When the file system approaches full, new nodes can be easily added
53and things will "just work."
54
55Ceph includes flexible snapshot mechanism that allows a user to create
56a snapshot on any subdirectory (and its nested contents) in the
57system. Snapshot creation and deletion are as simple as 'mkdir
58.snap/foo' and 'rmdir .snap/foo'.
59
60Ceph also provides some recursive accounting on directories for nested
61files and bytes. That is, a 'getfattr -d foo' on any directory in the
62system will reveal the total number of nested regular files and
63subdirectories, and a summation of all nested file sizes. This makes
64the identification of large disk space consumers relatively quick, as
65no 'du' or similar recursive scan of the file system is required.
66
67Finally, Ceph also allows quotas to be set on any directory in the system.
68The quota can restrict the number of bytes or the number of files stored
69beneath that point in the directory hierarchy. Quotas can be set using
70extended attributes 'ceph.quota.max_files' and 'ceph.quota.max_bytes', eg::
71
72 setfattr -n ceph.quota.max_bytes -v 100000000 /some/dir
73 getfattr -n ceph.quota.max_bytes /some/dir
74
75A limitation of the current quotas implementation is that it relies on the
76cooperation of the client mounting the file system to stop writers when a
77limit is reached. A modified or adversarial client cannot be prevented
78from writing as much data as it needs.
79
80Mount Syntax
81============
82
83The basic mount syntax is::
84
85 # mount -t ceph monip[:port][,monip2[:port]...]:/[subdir] mnt
86
87You only need to specify a single monitor, as the client will get the
88full list when it connects. (However, if the monitor you specify
89happens to be down, the mount won't succeed.) The port can be left
90off if the monitor is using the default. So if the monitor is at
911.2.3.4::
92
93 # mount -t ceph 1.2.3.4:/ /mnt/ceph
94
95is sufficient. If /sbin/mount.ceph is installed, a hostname can be
96used instead of an IP address.
97
98
99
100Mount Options
101=============
102
103 ip=A.B.C.D[:N]
104 Specify the IP and/or port the client should bind to locally.
105 There is normally not much reason to do this. If the IP is not
106 specified, the client's IP address is determined by looking at the
107 address its connection to the monitor originates from.
108
109 wsize=X
110 Specify the maximum write size in bytes. Default: 64 MB.
111
112 rsize=X
113 Specify the maximum read size in bytes. Default: 64 MB.
114
115 rasize=X
116 Specify the maximum readahead size in bytes. Default: 8 MB.
117
118 mount_timeout=X
119 Specify the timeout value for mount (in seconds), in the case
120 of a non-responsive Ceph file system. The default is 60
121 seconds.
122
123 caps_max=X
124 Specify the maximum number of caps to hold. Unused caps are released
125 when number of caps exceeds the limit. The default is 0 (no limit)
126
127 rbytes
128 When stat() is called on a directory, set st_size to 'rbytes',
129 the summation of file sizes over all files nested beneath that
130 directory. This is the default.
131
132 norbytes
133 When stat() is called on a directory, set st_size to the
134 number of entries in that directory.
135
136 nocrc
137 Disable CRC32C calculation for data writes. If set, the storage node
138 must rely on TCP's error correction to detect data corruption
139 in the data payload.
140
141 dcache
142 Use the dcache contents to perform negative lookups and
143 readdir when the client has the entire directory contents in
144 its cache. (This does not change correctness; the client uses
145 cached metadata only when a lease or capability ensures it is
146 valid.)
147
148 nodcache
149 Do not use the dcache as above. This avoids a significant amount of
150 complex code, sacrificing performance without affecting correctness,
151 and is useful for tracking down bugs.
152
153 noasyncreaddir
154 Do not use the dcache as above for readdir.
155
156 noquotadf
157 Report overall filesystem usage in statfs instead of using the root
158 directory quota.
159
160 nocopyfrom
161 Don't use the RADOS 'copy-from' operation to perform remote object
162 copies. Currently, it's only used in copy_file_range, which will revert
163 to the default VFS implementation if this option is used.
164
165 recover_session=<no|clean>
166 Set auto reconnect mode in the case where the client is blacklisted. The
167 available modes are "no" and "clean". The default is "no".
168
169 * no: never attempt to reconnect when client detects that it has been
170 blacklisted. Operations will generally fail after being blacklisted.
171
172 * clean: client reconnects to the ceph cluster automatically when it
173 detects that it has been blacklisted. During reconnect, client drops
174 dirty data/metadata, invalidates page caches and writable file handles.
175 After reconnect, file locks become stale because the MDS loses track
176 of them. If an inode contains any stale file locks, read/write on the
177 inode is not allowed until applications release all stale file locks.
178
179More Information
180================
181
182For more information on Ceph, see the home page at
183 https://ceph.com/
184
185The Linux kernel client source tree is available at
186 - https://github.com/ceph/ceph-client.git
187 - git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client.git
188
189and the source for the full system is at
190 https://github.com/ceph/ceph.git