1 .. SPDX-License-Identifier: GPL-2.0-only 1 .. SPDX-License-Identifier: GPL-2.0-only
2 2
3 dm-vdo 3 dm-vdo
4 ====== 4 ======
5 5
6 The dm-vdo (virtual data optimizer) device map 6 The dm-vdo (virtual data optimizer) device mapper target provides
7 block-level deduplication, compression, and th 7 block-level deduplication, compression, and thin provisioning. As a device
8 mapper target, it can add these features to th 8 mapper target, it can add these features to the storage stack, compatible
9 with any file system. The vdo target does not 9 with any file system. The vdo target does not protect against data
10 corruption, relying instead on integrity prote 10 corruption, relying instead on integrity protection of the storage below
11 it. It is strongly recommended that lvm be use 11 it. It is strongly recommended that lvm be used to manage vdo volumes. See
12 lvmvdo(7). 12 lvmvdo(7).
13 13
14 Userspace component 14 Userspace component
15 =================== 15 ===================
16 16
17 Formatting a vdo volume requires the use of th 17 Formatting a vdo volume requires the use of the 'vdoformat' tool, available
18 at: 18 at:
19 19
20 https://github.com/dm-vdo/vdo/ 20 https://github.com/dm-vdo/vdo/
21 21
22 In most cases, a vdo target will recover from 22 In most cases, a vdo target will recover from a crash automatically the
23 next time it is started. In cases where it enc 23 next time it is started. In cases where it encountered an unrecoverable
24 error (either during normal operation or crash 24 error (either during normal operation or crash recovery) the target will
25 enter or come up in read-only mode. Because re 25 enter or come up in read-only mode. Because read-only mode is indicative of
26 data-loss, a positive action must be taken to 26 data-loss, a positive action must be taken to bring vdo out of read-only
27 mode. The 'vdoforcerebuild' tool, available fr 27 mode. The 'vdoforcerebuild' tool, available from the same repo, is used to
28 prepare a read-only vdo to exit read-only mode 28 prepare a read-only vdo to exit read-only mode. After running this tool,
29 the vdo target will rebuild its metadata the n 29 the vdo target will rebuild its metadata the next time it is
30 started. Although some data may be lost, the r 30 started. Although some data may be lost, the rebuilt vdo's metadata will be
31 internally consistent and the target will be w 31 internally consistent and the target will be writable again.
32 32
33 The repo also contains additional userspace to 33 The repo also contains additional userspace tools which can be used to
34 inspect a vdo target's on-disk metadata. Fortu 34 inspect a vdo target's on-disk metadata. Fortunately, these tools are
35 rarely needed except by dm-vdo developers. 35 rarely needed except by dm-vdo developers.
36 36
37 Metadata requirements 37 Metadata requirements
38 ===================== 38 =====================
39 39
40 Each vdo volume reserves 3GB of space for meta 40 Each vdo volume reserves 3GB of space for metadata, or more depending on
41 its configuration. It is helpful to check that 41 its configuration. It is helpful to check that the space saved by
42 deduplication and compression is not cancelled 42 deduplication and compression is not cancelled out by the metadata
43 requirements. An estimation of the space saved 43 requirements. An estimation of the space saved for a specific dataset can
44 be computed with the vdo estimator tool, which 44 be computed with the vdo estimator tool, which is available at:
45 45
46 https://github.com/dm-vdo/vdoestimator/ 46 https://github.com/dm-vdo/vdoestimator/
47 47
48 Target interface 48 Target interface
49 ================ 49 ================
50 50
51 Table line 51 Table line
52 ---------- 52 ----------
53 53
54 :: 54 ::
55 55
56 <offset> <logical device size> vdo V4 56 <offset> <logical device size> vdo V4 <storage device>
57 <storage device size> <minimum I/O siz 57 <storage device size> <minimum I/O size> <block map cache size>
58 <block map era length> [optional argum 58 <block map era length> [optional arguments]
59 59
60 60
61 Required parameters: 61 Required parameters:
62 62
63 offset: 63 offset:
64 The offset, in sectors, at whi 64 The offset, in sectors, at which the vdo volume's logical
65 space begins. 65 space begins.
66 66
67 logical device size: 67 logical device size:
68 The size of the device which t 68 The size of the device which the vdo volume will service,
69 in sectors. Must match the cur 69 in sectors. Must match the current logical size of the vdo
70 volume. 70 volume.
71 71
72 storage device: 72 storage device:
73 The device holding the vdo vol 73 The device holding the vdo volume's data and metadata.
74 74
75 storage device size: 75 storage device size:
76 The size of the device holding 76 The size of the device holding the vdo volume, as a number
77 of 4096-byte blocks. Must matc 77 of 4096-byte blocks. Must match the current size of the vdo
78 volume. 78 volume.
79 79
80 minimum I/O size: 80 minimum I/O size:
81 The minimum I/O size for this 81 The minimum I/O size for this vdo volume to accept, in
82 bytes. Valid values are 512 or 82 bytes. Valid values are 512 or 4096. The recommended value
83 is 4096. 83 is 4096.
84 84
85 block map cache size: 85 block map cache size:
86 The size of the block map cach 86 The size of the block map cache, as a number of 4096-byte
87 blocks. The minimum and recomm 87 blocks. The minimum and recommended value is 32768 blocks.
88 If the logical thread count is 88 If the logical thread count is non-zero, the cache size
89 must be at least 4096 blocks p 89 must be at least 4096 blocks per logical thread.
90 90
91 block map era length: 91 block map era length:
92 The speed with which the block 92 The speed with which the block map cache writes out
93 modified block map pages. A sm 93 modified block map pages. A smaller era length is likely to
94 reduce the amount of time spen 94 reduce the amount of time spent rebuilding, at the cost of
95 increased block map writes dur 95 increased block map writes during normal operation. The
96 maximum and recommended value 96 maximum and recommended value is 16380; the minimum value
97 is 1. 97 is 1.
98 98
99 Optional parameters: 99 Optional parameters:
100 -------------------- 100 --------------------
101 Some or all of these parameters may be specifi 101 Some or all of these parameters may be specified as <key> <value> pairs.
102 102
103 Thread related parameters: 103 Thread related parameters:
104 104
105 Different categories of work are assigned to s 105 Different categories of work are assigned to separate thread groups, and
106 the number of threads in each group can be con 106 the number of threads in each group can be configured separately.
107 107
108 If <hash>, <logical>, and <physical> are all s 108 If <hash>, <logical>, and <physical> are all set to 0, the work handled by
109 all three thread types will be handled by a si 109 all three thread types will be handled by a single thread. If any of these
110 values are non-zero, all of them must be non-z 110 values are non-zero, all of them must be non-zero.
111 111
112 ack: 112 ack:
113 The number of threads used to 113 The number of threads used to complete bios. Since
114 completing a bio calls an arbi 114 completing a bio calls an arbitrary completion function
115 outside the vdo volume, thread 115 outside the vdo volume, threads of this type allow the vdo
116 volume to continue processing 116 volume to continue processing requests even when bio
117 completion is slow. The defaul 117 completion is slow. The default is 1.
118 118
119 bio: 119 bio:
120 The number of threads used to 120 The number of threads used to issue bios to the underlying
121 storage. Threads of this type 121 storage. Threads of this type allow the vdo volume to
122 continue processing requests e 122 continue processing requests even when bio submission is
123 slow. The default is 4. 123 slow. The default is 4.
124 124
125 bioRotationInterval: 125 bioRotationInterval:
126 The number of bios to enqueue 126 The number of bios to enqueue on each bio thread before
127 switching to the next thread. 127 switching to the next thread. The value must be greater
128 than 0 and not more than 1024; 128 than 0 and not more than 1024; the default is 64.
129 129
130 cpu: 130 cpu:
131 The number of threads used to 131 The number of threads used to do CPU-intensive work, such
132 as hashing and compression. Th 132 as hashing and compression. The default is 1.
133 133
134 hash: 134 hash:
135 The number of threads used to 135 The number of threads used to manage data comparisons for
136 deduplication based on the has 136 deduplication based on the hash value of data blocks. The
137 default is 0. 137 default is 0.
138 138
139 logical: 139 logical:
140 The number of threads used to 140 The number of threads used to manage caching and locking
141 based on the logical address o 141 based on the logical address of incoming bios. The default
142 is 0; the maximum is 60. 142 is 0; the maximum is 60.
143 143
144 physical: 144 physical:
145 The number of threads used to 145 The number of threads used to manage administration of the
146 underlying storage device. At 146 underlying storage device. At format time, a slab size for
147 the vdo is chosen; the vdo sto 147 the vdo is chosen; the vdo storage device must be large
148 enough to have at least 1 slab 148 enough to have at least 1 slab per physical thread. The
149 default is 0; the maximum is 1 149 default is 0; the maximum is 16.
150 150
151 Miscellaneous parameters: 151 Miscellaneous parameters:
152 152
153 maxDiscard: 153 maxDiscard:
154 The maximum size of discard bi 154 The maximum size of discard bio accepted, in 4096-byte
155 blocks. I/O requests to a vdo 155 blocks. I/O requests to a vdo volume are normally split
156 into 4096-byte blocks, and pro 156 into 4096-byte blocks, and processed up to 2048 at a time.
157 However, discard requests to a 157 However, discard requests to a vdo volume can be
158 automatically split to a large 158 automatically split to a larger size, up to <maxDiscard>
159 4096-byte blocks in a single b 159 4096-byte blocks in a single bio, and are limited to 1500
160 at a time. Increasing this val 160 at a time. Increasing this value may provide better overall
161 performance, at the cost of in 161 performance, at the cost of increased latency for the
162 individual discard requests. T 162 individual discard requests. The default and minimum is 1;
163 the maximum is UINT_MAX / 4096 163 the maximum is UINT_MAX / 4096.
164 164
165 deduplication: 165 deduplication:
166 Whether deduplication is enabl 166 Whether deduplication is enabled. The default is 'on'; the
167 acceptable values are 'on' and 167 acceptable values are 'on' and 'off'.
168 168
169 compression: 169 compression:
170 Whether compression is enabled 170 Whether compression is enabled. The default is 'off'; the
171 acceptable values are 'on' and 171 acceptable values are 'on' and 'off'.
172 172
173 Device modification 173 Device modification
174 ------------------- 174 -------------------
175 175
176 A modified table may be loaded into a running, 176 A modified table may be loaded into a running, non-suspended vdo volume.
177 The modifications will take effect when the de 177 The modifications will take effect when the device is next resumed. The
178 modifiable parameters are <logical device size 178 modifiable parameters are <logical device size>, <physical device size>,
179 <maxDiscard>, <compression>, and <deduplicatio 179 <maxDiscard>, <compression>, and <deduplication>.
180 180
181 If the logical device size or physical device 181 If the logical device size or physical device size are changed, upon
182 successful resume vdo will store the new value 182 successful resume vdo will store the new values and require them on future
183 startups. These two parameters may not be decr 183 startups. These two parameters may not be decreased. The logical device
184 size may not exceed 4 PB. The physical device 184 size may not exceed 4 PB. The physical device size must increase by at
185 least 32832 4096-byte blocks if at all, and mu 185 least 32832 4096-byte blocks if at all, and must not exceed the size of the
186 underlying storage device. Additionally, when 186 underlying storage device. Additionally, when formatting the vdo device, a
187 slab size is chosen: the physical device size 187 slab size is chosen: the physical device size may never increase above the
188 size which provides 8192 slabs, and each incre 188 size which provides 8192 slabs, and each increase must be large enough to
189 add at least one new slab. 189 add at least one new slab.
190 190
191 Examples: 191 Examples:
192 192
193 Start a previously-formatted vdo volume with 1 193 Start a previously-formatted vdo volume with 1 GB logical space and 1 GB
194 physical space, storing to /dev/dm-1 which has 194 physical space, storing to /dev/dm-1 which has more than 1 GB of space.
195 195
196 :: 196 ::
197 197
198 dmsetup create vdo0 --table \ 198 dmsetup create vdo0 --table \
199 "0 2097152 vdo V4 /dev/dm-1 262144 409 199 "0 2097152 vdo V4 /dev/dm-1 262144 4096 32768 16380"
200 200
201 Grow the logical size to 4 GB. 201 Grow the logical size to 4 GB.
202 202
203 :: 203 ::
204 204
205 dmsetup reload vdo0 --table \ 205 dmsetup reload vdo0 --table \
206 "0 8388608 vdo V4 /dev/dm-1 262144 409 206 "0 8388608 vdo V4 /dev/dm-1 262144 4096 32768 16380"
207 dmsetup resume vdo0 207 dmsetup resume vdo0
208 208
209 Grow the physical size to 2 GB. 209 Grow the physical size to 2 GB.
210 210
211 :: 211 ::
212 212
213 dmsetup reload vdo0 --table \ 213 dmsetup reload vdo0 --table \
214 "0 8388608 vdo V4 /dev/dm-1 524288 409 214 "0 8388608 vdo V4 /dev/dm-1 524288 4096 32768 16380"
215 dmsetup resume vdo0 215 dmsetup resume vdo0
216 216
217 Grow the physical size by 1 GB more and increa 217 Grow the physical size by 1 GB more and increase max discard sectors.
218 218
219 :: 219 ::
220 220
221 dmsetup reload vdo0 --table \ 221 dmsetup reload vdo0 --table \
222 "0 10485760 vdo V4 /dev/dm-1 786432 40 222 "0 10485760 vdo V4 /dev/dm-1 786432 4096 32768 16380 maxDiscard 8"
223 dmsetup resume vdo0 223 dmsetup resume vdo0
224 224
225 Stop the vdo volume. 225 Stop the vdo volume.
226 226
227 :: 227 ::
228 228
229 dmsetup remove vdo0 229 dmsetup remove vdo0
230 230
231 Start the vdo volume again. Note that the logi 231 Start the vdo volume again. Note that the logical and physical device sizes
232 must still match, but other parameters can cha 232 must still match, but other parameters can change.
233 233
234 :: 234 ::
235 235
236 dmsetup create vdo1 --table \ 236 dmsetup create vdo1 --table \
237 "0 10485760 vdo V4 /dev/dm-1 786432 51 237 "0 10485760 vdo V4 /dev/dm-1 786432 512 65550 5000 hash 1 logical 3 physical 2"
238 238
239 Messages 239 Messages
240 -------- 240 --------
241 All vdo devices accept messages in the form: 241 All vdo devices accept messages in the form:
242 242
243 :: 243 ::
244 <<
245 dmsetup message <target-name> 0 <messa 244 dmsetup message <target-name> 0 <message-name> <message-parameters>
246 245
247 The messages are: 246 The messages are:
248 247
249 stats: 248 stats:
250 Outputs the current view of th 249 Outputs the current view of the vdo statistics. Mostly used
251 by the vdostats userspace prog 250 by the vdostats userspace program to interpret the output
252 buffer. 251 buffer.
253 252
254 config: !! 253 dump:
255 Outputs useful vdo configurati <<
256 by users who want to recreate <<
257 want to know the creation conf <<
258 <<
259 dump: <<
260 Dumps many internal structures 254 Dumps many internal structures to the system log. This is
261 not always safe to run, so it 255 not always safe to run, so it should only be used to debug
262 a hung vdo. Optional parameter 256 a hung vdo. Optional parameters to specify structures to
263 dump are: 257 dump are:
264 258
265 viopool: The pool of I 259 viopool: The pool of I/O requests incoming bios
266 pools: A synonym of 'v 260 pools: A synonym of 'viopool'
267 vdo: Most of the struc 261 vdo: Most of the structures managing on-disk data
268 queues: Basic informat 262 queues: Basic information about each vdo thread
269 threads: A synonym of 263 threads: A synonym of 'queues'
270 default: Equivalent to 264 default: Equivalent to 'queues vdo'
271 all: All of the above. 265 all: All of the above.
272 266
273 dump-on-shutdown: 267 dump-on-shutdown:
274 Perform a default dump next ti 268 Perform a default dump next time vdo shuts down.
275 269
276 270
277 Status 271 Status
278 ------ 272 ------
279 273
280 :: 274 ::
281 275
282 <device> <operating mode> <in recovery> <i 276 <device> <operating mode> <in recovery> <index state>
283 <compression state> <physical blocks used> 277 <compression state> <physical blocks used> <total physical blocks>
284 278
285 device: 279 device:
286 The name of the vdo volume. 280 The name of the vdo volume.
287 281
288 operating mode: 282 operating mode:
289 The current operating mode of 283 The current operating mode of the vdo volume; values may be
290 'normal', 'recovering' (the vo 284 'normal', 'recovering' (the volume has detected an issue
291 with its metadata and is attem 285 with its metadata and is attempting to repair itself), and
292 'read-only' (an error has occu 286 'read-only' (an error has occurred that forces the vdo
293 volume to only support read op 287 volume to only support read operations and not writes).
294 288
295 in recovery: 289 in recovery:
296 Whether the vdo volume is curr 290 Whether the vdo volume is currently in recovery mode;
297 values may be 'recovering' or 291 values may be 'recovering' or '-' which indicates not
298 recovering. 292 recovering.
299 293
300 index state: 294 index state:
301 The current state of the dedup 295 The current state of the deduplication index in the vdo
302 volume; values may be 'closed' 296 volume; values may be 'closed', 'closing', 'error',
303 'offline', 'online', 'opening' 297 'offline', 'online', 'opening', and 'unknown'.
304 298
305 compression state: 299 compression state:
306 The current state of compressi 300 The current state of compression in the vdo volume; values
307 may be 'offline' and 'online'. 301 may be 'offline' and 'online'.
308 302
309 used physical blocks: 303 used physical blocks:
310 The number of physical blocks 304 The number of physical blocks in use by the vdo volume.
311 305
312 total physical blocks: 306 total physical blocks:
313 The total number of physical b 307 The total number of physical blocks the vdo volume may use;
314 the difference between this va 308 the difference between this value and the
315 <used physical blocks> is the 309 <used physical blocks> is the number of blocks the vdo
316 volume has left before being f 310 volume has left before being full.
317 311
318 Memory Requirements 312 Memory Requirements
319 =================== 313 ===================
320 314
321 A vdo target requires a fixed 38 MB of RAM alo 315 A vdo target requires a fixed 38 MB of RAM along with the following amounts
322 that scale with the target: 316 that scale with the target:
323 317
324 - 1.15 MB of RAM for each 1 MB of configured b 318 - 1.15 MB of RAM for each 1 MB of configured block map cache size. The
325 block map cache requires a minimum of 150 MB 319 block map cache requires a minimum of 150 MB.
326 - 1.6 MB of RAM for each 1 TB of logical space 320 - 1.6 MB of RAM for each 1 TB of logical space.
327 - 268 MB of RAM for each 1 TB of physical stor 321 - 268 MB of RAM for each 1 TB of physical storage managed by the volume.
328 322
329 The deduplication index requires additional me 323 The deduplication index requires additional memory which scales with the
330 size of the deduplication window. For dense in 324 size of the deduplication window. For dense indexes, the index requires 1
331 GB of RAM per 1 TB of window. For sparse index 325 GB of RAM per 1 TB of window. For sparse indexes, the index requires 1 GB
332 of RAM per 10 TB of window. The index configur 326 of RAM per 10 TB of window. The index configuration is set when the target
333 is formatted and may not be modified. 327 is formatted and may not be modified.
334 328
335 Module Parameters 329 Module Parameters
336 ================= 330 =================
337 331
338 The vdo driver has a numeric parameter 'log_le 332 The vdo driver has a numeric parameter 'log_level' which controls the
339 verbosity of logging from the driver. The defa 333 verbosity of logging from the driver. The default setting is 6
340 (LOGLEVEL_INFO and more severe messages). 334 (LOGLEVEL_INFO and more severe messages).
341 335
342 Run-time Usage 336 Run-time Usage
343 ============== 337 ==============
344 338
345 When using dm-vdo, it is important to be aware 339 When using dm-vdo, it is important to be aware of the ways in which its
346 behavior differs from other storage targets. 340 behavior differs from other storage targets.
347 341
348 - There is no guarantee that over-writes of ex 342 - There is no guarantee that over-writes of existing blocks will succeed.
349 Because the underlying storage may be multip 343 Because the underlying storage may be multiply referenced, over-writing
350 an existing block generally requires a vdo t 344 an existing block generally requires a vdo to have a free block
351 available. 345 available.
352 346
353 - When blocks are no longer in use, sending a 347 - When blocks are no longer in use, sending a discard request for those
354 blocks lets the vdo release references for t 348 blocks lets the vdo release references for those blocks. If the vdo is
355 thinly provisioned, discarding unused blocks 349 thinly provisioned, discarding unused blocks is essential to prevent the
356 target from running out of space. However, d 350 target from running out of space. However, due to the sharing of
357 duplicate blocks, no discard request for any 351 duplicate blocks, no discard request for any given logical block is
358 guaranteed to reclaim space. 352 guaranteed to reclaim space.
359 353
360 - Assuming the underlying storage properly imp 354 - Assuming the underlying storage properly implements flush requests, vdo
361 is resilient against crashes, however, unflu 355 is resilient against crashes, however, unflushed writes may or may not
362 persist after a crash. 356 persist after a crash.
363 357
364 - Each write to a vdo target entails a signifi 358 - Each write to a vdo target entails a significant amount of processing.
365 However, much of the work is paralellizable. 359 However, much of the work is paralellizable. Therefore, vdo targets
366 achieve better throughput at higher I/O dept 360 achieve better throughput at higher I/O depths, and can support up 2048
367 requests in parallel. 361 requests in parallel.
368 362
369 Tuning 363 Tuning
370 ====== 364 ======
371 365
372 The vdo device has many options, and it can be 366 The vdo device has many options, and it can be difficult to make optimal
373 choices without perfect knowledge of the workl 367 choices without perfect knowledge of the workload. Additionally, most
374 configuration options must be set when a vdo t 368 configuration options must be set when a vdo target is started, and cannot
375 be changed without shutting it down completely 369 be changed without shutting it down completely; the configuration cannot be
376 changed while the target is active. Ideally, t 370 changed while the target is active. Ideally, tuning with simulated
377 workloads should be performed before deploying 371 workloads should be performed before deploying vdo in production
378 environments. 372 environments.
379 373
380 The most important value to adjust is the bloc 374 The most important value to adjust is the block map cache size. In order to
381 service a request for any logical address, a v 375 service a request for any logical address, a vdo must load the portion of
382 the block map which holds the relevant mapping 376 the block map which holds the relevant mapping. These mappings are cached.
383 Performance will suffer when the working set d 377 Performance will suffer when the working set does not fit in the cache. By
384 default, a vdo allocates 128 MB of metadata ca 378 default, a vdo allocates 128 MB of metadata cache in RAM to support
385 efficient access to 100 GB of logical space at 379 efficient access to 100 GB of logical space at a time. It should be scaled
386 up proportionally for larger working sets. 380 up proportionally for larger working sets.
387 381
388 The logical and physical thread counts should 382 The logical and physical thread counts should also be adjusted. A logical
389 thread controls a disjoint section of the bloc 383 thread controls a disjoint section of the block map, so additional logical
390 threads increase parallelism and can increase 384 threads increase parallelism and can increase throughput. Physical threads
391 control a disjoint section of the data blocks, 385 control a disjoint section of the data blocks, so additional physical
392 threads can also increase throughput. However, 386 threads can also increase throughput. However, excess threads can waste
393 resources and increase contention. 387 resources and increase contention.
394 388
395 Bio submission threads control the parallelism 389 Bio submission threads control the parallelism involved in sending I/O to
396 the underlying storage; fewer threads mean the 390 the underlying storage; fewer threads mean there is more opportunity to
397 reorder I/O requests for performance benefit, 391 reorder I/O requests for performance benefit, but also that each I/O
398 request has to wait longer before being submit 392 request has to wait longer before being submitted.
399 393
400 Bio acknowledgment threads are used for finish 394 Bio acknowledgment threads are used for finishing I/O requests. This is
401 done on dedicated threads since the amount of 395 done on dedicated threads since the amount of work required to execute a
402 bio's callback can not be controlled by the vd 396 bio's callback can not be controlled by the vdo itself. Usually one thread
403 is sufficient but additional threads may be be 397 is sufficient but additional threads may be beneficial, particularly when
404 bios have CPU-heavy callbacks. 398 bios have CPU-heavy callbacks.
405 399
406 CPU threads are used for hashing and for compr 400 CPU threads are used for hashing and for compression; in workloads with
407 compression enabled, more threads may result i 401 compression enabled, more threads may result in higher throughput.
408 402
409 Hash threads are used to sort active requests 403 Hash threads are used to sort active requests by hash and determine whether
410 they should deduplicate; the most CPU intensiv 404 they should deduplicate; the most CPU intensive actions done by these
411 threads are comparison of 4096-byte data block 405 threads are comparison of 4096-byte data blocks. In most cases, a single
412 hash thread is sufficient. 406 hash thread is sufficient.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.