1RADOS(8)                             Ceph                             RADOS(8)
2
3
4

NAME

6       rados - rados object storage utility
7

SYNOPSIS

9       rados [ -m monaddr ] [ mkpool | rmpool foo ] [ -p | --pool
10       pool ] [ -s | --snap snap ] [ -i infile ] [ -o outfile ]
11       command ...
12
13

DESCRIPTION

15       rados  is  a utility for interacting with a Ceph object storage cluster
16       (RADOS), part of the Ceph distributed storage system.
17

OPTIONS

19       -p pool, --pool pool
20              Interact with the given pool. Required by most commands.
21
22       -s snap, --snap snap
23              Read from the given pool snapshot. Valid for  all  pool-specific
24              read operations.
25
26       -i infile
27              will  specify an input file to be passed along as a payload with
28              the command to the monitor cluster. This is only used  for  spe‐
29              cific monitor commands.
30
31       -o outfile
32              will  write any payload returned by the monitor cluster with its
33              reply to outfile.  Only  specific  monitor  commands  (e.g.  osd
34              getmap) return a payload.
35
36       -c ceph.conf, --conf=ceph.conf
37              Use   ceph.conf   configuration  file  instead  of  the  default
38              /etc/ceph/ceph.conf  to  determine  monitor   addresses   during
39              startup.
40
41       -m monaddress[:port]
42              Connect   to  specified  monitor  (instead  of  looking  through
43              ceph.conf).
44
45       -b block_size
46              Set the block size for put/get/append ops and for  write  bench‐
47              marking.
48
49       --striper
50              Uses  the  striping  API  of  rados rather than the default one.
51              Available for stat, get, put, append, truncate, rm, ls  and  all
52              xattr related operation
53

GLOBAL COMMANDS

55       lspools
56              List object pools
57
58       df     Show  utilization  statistics,  including disk usage (bytes) and
59              object counts, over the entire system and broken down by pool.
60
61       mkpool foo
62              Create a pool with name foo.
63
64       rmpool foo [ foo --yes-i-really-really-mean-it ]
65              Delete the pool foo (and all its data).
66
67       list-inconsistent-pg pool
68              List inconsistent PGs in given pool.
69
70       list-inconsistent-obj pgid
71              List inconsistent objects in given PG.
72
73       list-inconsistent-snapset pgid
74              List inconsistent snapsets in given PG.
75

POOL SPECIFIC COMMANDS

77       get name outfile
78              Read object name from the cluster and write it to outfile.
79
80       put name infile [--offset offset]
81              Write object name with start offset (default:0) to  the  cluster
82              with contents from infile.
83
84       append name infile
85              Append object name to the cluster with contents from infile.
86
87       rm name
88              Remove object name.
89
90       listwatchers name
91              List the watchers of object name.
92
93       ls outfile
94              List objects in given pool and write to outfile.
95
96       lssnap List snapshots for given pool.
97
98       clonedata srcname dstname --object-locator key
99              Clone  object  byte  data from srcname to dstname.  Both objects
100              must be stored with the locator key key (usually either  srcname
101              or  dstname).  Object attributes and omap keys are not copied or
102              cloned.
103
104       mksnap foo
105              Create pool snapshot named foo.
106
107       rmsnap foo
108              Remove pool snapshot named foo.
109
110       bench seconds mode [ -b objsize ] [ -t threads ]
111              Benchmark for seconds. The mode can be write, seq, or rand.  seq
112              and  rand  are  read  benchmarks,  either  sequential or random.
113              Before running one of the reading benchmarks, run a write bench‐
114              mark  with the --no-cleanup option. The default object size is 4
115              MB, and  the  default  number  of  simulated  threads  (parallel
116              writes)  is  16.  The  --run-name  <label>  option is useful for
117              benchmarking a workload test from multiple clients. The  <label>
118              is  an arbitrary object name. It is "benchmark_last_metadata" by
119              default, and is used as the underlying object  name  for  "read"
120              and "write" ops.  Note: -b objsize option is valid only in write
121              mode.  Note: write and seq must be run on the same  host  other‐
122              wise the objects created by write will have names that will fail
123              seq.
124
125       cleanup [ --run-name run_name ] [ --prefix prefix ]
126              Clean up a previous  benchmark  operation.   Note:  the  default
127              run-name is "benchmark_last_metadata"
128
129       listxattr name
130              List all extended attributes of an object.
131
132       getxattr name attr
133              Dump the extended attribute value of attr of an object.
134
135       setxattr name attr value
136              Set the value of attr in the extended attributes of an object.
137
138       rmxattr name attr
139              Remove attr from the extended attributes of an object.
140
141       listomapkeys name
142              List all the keys stored in the object map of object name.
143
144       listomapvals name
145              List  all  key/value  pairs  stored  in the object map of object
146              name.  The values are dumped in hexadecimal.
147
148       getomapval [ --omap-key-file file ] name key [ out-file ]
149              Dump the hexadecimal value of key in the object  map  of  object
150              name.   If  the  optional out-file argument is not provided, the
151              value will be written to standard output.
152
153       setomapval [ --omap-key-file file ] name key [ value ]
154              Set the value of key in the object map of object  name.  If  the
155              optional  value argument is not provided, the value will be read
156              from standard input.
157
158       rmomapkey [ --omap-key-file file ] name key
159              Remove key from the object map of object name.
160
161       getomapheader name
162              Dump the hexadecimal value of the object map  header  of  object
163              name.
164
165       setomapheader name value
166              Set the value of the object map header of object name.
167

EXAMPLES

169       To view cluster utilization:
170
171          rados df
172
173       To get a list object in pool foo sent to stdout:
174
175          rados -p foo ls -
176
177       To write an object:
178
179          rados -p foo put myobject blah.txt
180
181       To create a snapshot:
182
183          rados -p foo mksnap mysnap
184
185       To delete the object:
186
187          rados -p foo rm myobject
188
189       To read a previously snapshotted version of an object:
190
191          rados -p foo -s mysnap get myobject blah.txt.old
192
193       To list inconsistent objects in PG 0.6:
194
195          rados list-inconsistent-obj 0.6 --format=json-pretty
196

AVAILABILITY

198       rados  is  part of Ceph, a massively scalable, open-source, distributed
199       storage  system.  Please   refer   to   the   Ceph   documentation   at
200       http://ceph.com/docs for more information.
201

SEE ALSO

203       ceph(8)
204
206       2010-2014,  Inktank Storage, Inc. and contributors. Licensed under Cre‐
207       ative Commons Attribution Share Alike 3.0 (CC-BY-SA-3.0)
208
209
210
211
212dev                              Apr 14, 2019                         RADOS(8)
Impressum