1RADOS(8) Ceph RADOS(8)
2
6 rados - rados object storage utility
7
9 rados [ options ] [ command ]
10
13 rados is a utility for interacting with a Ceph object storage cluster
14 (RADOS), part of the Ceph distributed storage system.
15
17 -p pool, --pool pool
18 Interact with the given pool. Required by most commands.
19 --pgid As an alternative to --pool, --pgid also allow users to specify
21 the PG id to which the command will be directed. With this
22 option, certain commands like ls allow users to limit the scope
23 of the command to the given PG.
24 -N namespace, --namespace namespace
26 Specify the rados namespace to use for the object.
27 -s snap, --snap snap
29 Read from the given pool snapshot. Valid for all pool-specific
30 read operations.
31 -i infile
33 will specify an input file to be passed along as a payload with
34 the command to the monitor cluster. This is only used for spe‐
35 cific monitor commands.
36 -o outfile
38 will write any payload returned by the monitor cluster with its
39 reply to outfile. Only specific monitor commands (e.g. osd
40 getmap) return a payload.
41 -c ceph.conf, --conf=ceph.conf
43 Use ceph.conf configuration file instead of the default
44 /etc/ceph/ceph.conf to determine monitor addresses during
45 startup.
46 -m monaddress[:port]
48 Connect to specified monitor (instead of looking through
49 ceph.conf).
50 -b block_size
52 Set the block size for put/get/append ops and for write bench‐
53 marking.
54 --striper
56 Uses the striping API of rados rather than the default one.
57 Available for stat, stat2, get, put, append, truncate, rm, ls
58 and all xattr related operation
59 -O object_size
61 Set the object size for put/get ops and for write benchmarking
62 --lock-cookie locker-cookie
64 Will set the lock cookie for acquiring advisory lock (lock get
65 command). If the cookie is not empty, this option must be
66 passed to lock break command to find the correct lock when
67 releasing lock.
68
70 lspools
71 List object pools
72 df Show utilization statistics, including disk usage (bytes) and
74 object counts, over the entire system and broken down by pool.
75 list-inconsistent-pg pool
77 List inconsistent PGs in given pool.
78 list-inconsistent-obj pgid
80 List inconsistent objects in given PG.
81 list-inconsistent-snapset pgid
83 List inconsistent snapsets in given PG.
84
86 get name outfile
87 Read object name from the cluster and write it to outfile.
88 put name infile [--offset offset]
90 Write object name with start offset (default:0) to the cluster
91 with contents from infile. Warning: The put command creates a
92 single RADOS object, sized just as large as your input file.
93 Unless your objects are of reasonable and consistent sizes, that
94 is probably not what you want -- consider using RGW/S3, CephFS,
95 or RBD instead.
96 append name infile
98 Append object name to the cluster with contents from infile.
99 rm name
101 Remove object name.
102 listwatchers name
104 List the watchers of object name.
105 ls outfile
107 List objects in the given pool and write to outfile. Instead of
108 --pool if --pgid will be specified, ls will only list the
109 objects in the given PG.
110 lssnap List snapshots for given pool.
112 clonedata srcname dstname --object-locator key
114 Clone object byte data from srcname to dstname. Both objects
115 must be stored with the locator key key (usually either srcname
116 or dstname). Object attributes and omap keys are not copied or
117 cloned.
118 mksnap foo
120 Create pool snapshot named foo.
121 rmsnap foo
123 Remove pool snapshot named foo.
124 bench seconds mode [ -b objsize ] [ -t threads ]
126 Benchmark for seconds. The mode can be write, seq, or rand. seq
127 and rand are read benchmarks, either sequential or random.
128 Before running one of the reading benchmarks, run a write bench‐
129 mark with the --no-cleanup option. The default object size is 4
130 MB, and the default number of simulated threads (parallel
131 writes) is 16. The --run-name <label> option is useful for
132 benchmarking a workload test from multiple clients. The <label>
133 is an arbitrary object name. It is "benchmark_last_metadata" by
134 default, and is used as the underlying object name for "read"
135 and "write" ops. Note: -b objsize option is valid only in write
136 mode. Note: write and seq must be run on the same host other‐
137 wise the objects created by write will have names that will fail
138 seq.
139 cleanup [ --run-name run_name ] [ --prefix prefix ]
141 Clean up a previous benchmark operation. Note: the default
142 run-name is "benchmark_last_metadata"
143 listxattr name
145 List all extended attributes of an object.
146 getxattr name attr
148 Dump the extended attribute value of attr of an object.
149 setxattr name attr value
151 Set the value of attr in the extended attributes of an object.
152 rmxattr name attr
154 Remove attr from the extended attributes of an object.
155 stat name
157 Get stat (ie. mtime, size) of given object
158 stat2 name
160 Get stat (similar to stat, but with high precision time) of
161 given object
162 listomapkeys name
164 List all the keys stored in the object map of object name.
165 listomapvals name
167 List all key/value pairs stored in the object map of object
168 name. The values are dumped in hexadecimal.
169 getomapval [ --omap-key-file file ] name key [ out-file ]
171 Dump the hexadecimal value of key in the object map of object
172 name. If the optional out-file argument is not provided, the
173 value will be written to standard output.
174 setomapval [ --omap-key-file file ] name key [ value ]
176 Set the value of key in the object map of object name. If the
177 optional value argument is not provided, the value will be read
178 from standard input.
179 rmomapkey [ --omap-key-file file ] name key
181 Remove key from the object map of object name.
182 getomapheader name
184 Dump the hexadecimal value of the object map header of object
185 name.
186 setomapheader name value
188 Set the value of the object map header of object name.
189 export filename
191 Serialize pool contents to a file or standard output.n"
192 import [--dry-run] [--no-overwrite] < filename | - >
194 Load pool contents from a file or standard input
195
197 To view cluster utilization:
198 rados df
200 To get a list object in pool foo sent to stdout:
202 rados -p foo ls -
204 To get a list of objects in PG 0.6:
206 rados --pgid 0.6 ls
208 To write an object:
210 rados -p foo put myobject blah.txt
212 To create a snapshot:
214 rados -p foo mksnap mysnap
216 To delete the object:
218 rados -p foo rm myobject
220 To read a previously snapshotted version of an object:
222 rados -p foo -s mysnap get myobject blah.txt.old
224 To list inconsistent objects in PG 0.6:
226 rados list-inconsistent-obj 0.6 --format=json-pretty
228
230 rados is part of Ceph, a massively scalable, open-source, distributed
231 storage system. Please refer to the Ceph documentation at
232 http://ceph.com/docs for more information.
233
235 ceph(8)
236
238 2010-2021, Inktank Storage, Inc. and contributors. Licensed under Cre‐
239 ative Commons Attribution Share Alike 3.0 (CC-BY-SA-3.0)
240dev Mar 18, 2021 RADOS(8)