Commit | Line | Data |
---|---|---|
f80dba8d MT |
1 | How fio works |
2 | ------------- | |
3 | ||
4 | The first step in getting fio to simulate a desired I/O workload, is writing a | |
5 | job file describing that specific setup. A job file may contain any number of | |
6 | threads and/or files -- the typical contents of the job file is a *global* | |
7 | section defining shared parameters, and one or more job sections describing the | |
8 | jobs involved. When run, fio parses this file and sets everything up as | |
9 | described. If we break down a job from top to bottom, it contains the following | |
10 | basic parameters: | |
11 | ||
12 | `I/O type`_ | |
13 | ||
14 | Defines the I/O pattern issued to the file(s). We may only be reading | |
15 | sequentially from this file(s), or we may be writing randomly. Or even | |
16 | mixing reads and writes, sequentially or randomly. | |
17 | Should we be doing buffered I/O, or direct/raw I/O? | |
18 | ||
19 | `Block size`_ | |
20 | ||
21 | In how large chunks are we issuing I/O? This may be a single value, | |
22 | or it may describe a range of block sizes. | |
23 | ||
24 | `I/O size`_ | |
25 | ||
26 | How much data are we going to be reading/writing. | |
27 | ||
28 | `I/O engine`_ | |
29 | ||
30 | How do we issue I/O? We could be memory mapping the file, we could be | |
31 | using regular read/write, we could be using splice, async I/O, or even | |
32 | SG (SCSI generic sg). | |
33 | ||
34 | `I/O depth`_ | |
35 | ||
36 | If the I/O engine is async, how large a queuing depth do we want to | |
37 | maintain? | |
38 | ||
39 | ||
40 | `Target file/device`_ | |
41 | ||
42 | How many files are we spreading the workload over. | |
43 | ||
44 | `Threads, processes and job synchronization`_ | |
45 | ||
46 | How many threads or processes should we spread this workload over. | |
47 | ||
48 | The above are the basic parameters defined for a workload, in addition there's a | |
49 | multitude of parameters that modify other aspects of how this job behaves. | |
50 | ||
51 | ||
52 | Command line options | |
53 | -------------------- | |
54 | ||
55 | .. option:: --debug=type | |
56 | ||
b034c0dd TK |
57 | Enable verbose tracing of various fio actions. May be ``all`` for all types |
58 | or individual types separated by a comma (e.g. ``--debug=file,mem`` will | |
59 | enable file and memory debugging). Currently, additional logging is | |
60 | available for: | |
f80dba8d | 61 | |
b034c0dd | 62 | *process* |
f80dba8d | 63 | Dump info related to processes. |
b034c0dd | 64 | *file* |
f80dba8d | 65 | Dump info related to file actions. |
b034c0dd | 66 | *io* |
f80dba8d | 67 | Dump info related to I/O queuing. |
b034c0dd | 68 | *mem* |
f80dba8d | 69 | Dump info related to memory allocations. |
b034c0dd | 70 | *blktrace* |
f80dba8d | 71 | Dump info related to blktrace setup. |
b034c0dd | 72 | *verify* |
f80dba8d | 73 | Dump info related to I/O verification. |
b034c0dd | 74 | *all* |
f80dba8d | 75 | Enable all debug options. |
b034c0dd | 76 | *random* |
f80dba8d | 77 | Dump info related to random offset generation. |
b034c0dd | 78 | *parse* |
f80dba8d | 79 | Dump info related to option matching and parsing. |
b034c0dd | 80 | *diskutil* |
f80dba8d | 81 | Dump info related to disk utilization updates. |
b034c0dd | 82 | *job:x* |
f80dba8d | 83 | Dump info only related to job number x. |
b034c0dd | 84 | *mutex* |
f80dba8d | 85 | Dump info only related to mutex up/down ops. |
b034c0dd | 86 | *profile* |
f80dba8d | 87 | Dump info related to profile extensions. |
b034c0dd | 88 | *time* |
f80dba8d | 89 | Dump info related to internal time keeping. |
b034c0dd | 90 | *net* |
f80dba8d | 91 | Dump info related to networking connections. |
b034c0dd | 92 | *rate* |
f80dba8d | 93 | Dump info related to I/O rate switching. |
b034c0dd | 94 | *compress* |
f80dba8d | 95 | Dump info related to log compress/decompress. |
b034c0dd | 96 | *?* or *help* |
f80dba8d MT |
97 | Show available debug options. |
98 | ||
99 | .. option:: --parse-only | |
100 | ||
25cd4b95 | 101 | Parse options only, don't start any I/O. |
f80dba8d MT |
102 | |
103 | .. option:: --output=filename | |
104 | ||
105 | Write output to file `filename`. | |
106 | ||
b8f7e412 TK |
107 | .. option:: --output-format=type |
108 | ||
109 | Set the reporting format to `normal`, `terse`, `json`, or `json+`. Multiple | |
110 | formats can be selected, separated by a comma. `terse` is a CSV based | |
111 | format. `json+` is like `json`, except it adds a full dump of the latency | |
112 | buckets. | |
113 | ||
114 | .. option:: --runtime | |
115 | Limit run time to runtime seconds. | |
116 | ||
f80dba8d MT |
117 | .. option:: --bandwidth-log |
118 | ||
119 | Generate aggregate bandwidth logs. | |
120 | ||
121 | .. option:: --minimal | |
122 | ||
123 | Print statistics in a terse, semicolon-delimited format. | |
124 | ||
125 | .. option:: --append-terse | |
126 | ||
b034c0dd TK |
127 | Print statistics in selected mode AND terse, semicolon-delimited format. |
128 | **Deprecated**, use :option:`--output-format` instead to select multiple | |
129 | formats. | |
f80dba8d | 130 | |
f80dba8d MT |
131 | .. option:: --terse-version=type |
132 | ||
a2c95580 | 133 | Set terse version output format (default 3, or 2 or 4 or 5). |
f80dba8d MT |
134 | |
135 | .. option:: --version | |
136 | ||
b8f7e412 | 137 | Print version information and exit. |
f80dba8d MT |
138 | |
139 | .. option:: --help | |
140 | ||
113f0e7c | 141 | Print a summary of the command line options and exit. |
f80dba8d MT |
142 | |
143 | .. option:: --cpuclock-test | |
144 | ||
145 | Perform test and validation of internal CPU clock. | |
146 | ||
113f0e7c | 147 | .. option:: --crctest=[test] |
f80dba8d | 148 | |
b034c0dd TK |
149 | Test the speed of the built-in checksumming functions. If no argument is |
150 | given, all of them are tested. Alternatively, a comma separated list can | |
151 | be passed, in which case the given ones are tested. | |
f80dba8d MT |
152 | |
153 | .. option:: --cmdhelp=command | |
154 | ||
155 | Print help information for `command`. May be ``all`` for all commands. | |
156 | ||
157 | .. option:: --enghelp=[ioengine[,command]] | |
158 | ||
b034c0dd TK |
159 | List all commands defined by :option:`ioengine`, or print help for `command` |
160 | defined by :option:`ioengine`. If no :option:`ioengine` is given, list all | |
161 | available ioengines. | |
f80dba8d MT |
162 | |
163 | .. option:: --showcmd=jobfile | |
164 | ||
b8f7e412 | 165 | Convert `jobfile` to a set of command-line options. |
f80dba8d MT |
166 | |
167 | .. option:: --readonly | |
168 | ||
b034c0dd TK |
169 | Turn on safety read-only checks, preventing writes. The ``--readonly`` |
170 | option is an extra safety guard to prevent users from accidentally starting | |
171 | a write workload when that is not desired. Fio will only write if | |
172 | `rw=write/randwrite/rw/randrw` is given. This extra safety net can be used | |
173 | as an extra precaution as ``--readonly`` will also enable a write check in | |
174 | the I/O engine core to prevent writes due to unknown user space bug(s). | |
f80dba8d MT |
175 | |
176 | .. option:: --eta=when | |
177 | ||
b8f7e412 TK |
178 | Specifies when real-time ETA estimate should be printed. `when` may be |
179 | `always`, `never` or `auto`. | |
f80dba8d MT |
180 | |
181 | .. option:: --eta-newline=time | |
182 | ||
947e0fe0 SW |
183 | Force a new line for every `time` period passed. When the unit is omitted, |
184 | the value is interpreted in seconds. | |
f80dba8d MT |
185 | |
186 | .. option:: --status-interval=time | |
187 | ||
947e0fe0 SW |
188 | Force full status dump every `time` period passed. When the unit is |
189 | omitted, the value is interpreted in seconds. | |
f80dba8d MT |
190 | |
191 | .. option:: --section=name | |
192 | ||
b034c0dd TK |
193 | Only run specified section `name` in job file. Multiple sections can be specified. |
194 | The ``--section`` option allows one to combine related jobs into one file. | |
195 | E.g. one job file could define light, moderate, and heavy sections. Tell | |
196 | fio to run only the "heavy" section by giving ``--section=heavy`` | |
197 | command line option. One can also specify the "write" operations in one | |
198 | section and "verify" operation in another section. The ``--section`` option | |
199 | only applies to job sections. The reserved *global* section is always | |
200 | parsed and used. | |
f80dba8d MT |
201 | |
202 | .. option:: --alloc-size=kb | |
203 | ||
b034c0dd TK |
204 | Set the internal smalloc pool size to `kb` in KiB. The |
205 | ``--alloc-size`` switch allows one to use a larger pool size for smalloc. | |
206 | If running large jobs with randommap enabled, fio can run out of memory. | |
207 | Smalloc is an internal allocator for shared structures from a fixed size | |
208 | memory pool and can grow to 16 pools. The pool size defaults to 16MiB. | |
f80dba8d | 209 | |
b034c0dd TK |
210 | NOTE: While running :file:`.fio_smalloc.*` backing store files are visible |
211 | in :file:`/tmp`. | |
f80dba8d MT |
212 | |
213 | .. option:: --warnings-fatal | |
214 | ||
b034c0dd TK |
215 | All fio parser warnings are fatal, causing fio to exit with an |
216 | error. | |
f80dba8d MT |
217 | |
218 | .. option:: --max-jobs=nr | |
219 | ||
b8f7e412 | 220 | Set the maximum number of threads/processes to support. |
f80dba8d MT |
221 | |
222 | .. option:: --server=args | |
223 | ||
b034c0dd TK |
224 | Start a backend server, with `args` specifying what to listen to. |
225 | See `Client/Server`_ section. | |
f80dba8d MT |
226 | |
227 | .. option:: --daemonize=pidfile | |
228 | ||
b034c0dd | 229 | Background a fio server, writing the pid to the given `pidfile` file. |
f80dba8d MT |
230 | |
231 | .. option:: --client=hostname | |
232 | ||
b034c0dd TK |
233 | Instead of running the jobs locally, send and run them on the given host or |
234 | set of hosts. See `Client/Server`_ section. | |
f80dba8d MT |
235 | |
236 | .. option:: --remote-config=file | |
237 | ||
238 | Tell fio server to load this local file. | |
239 | ||
240 | .. option:: --idle-prof=option | |
241 | ||
b8f7e412 | 242 | Report CPU idleness. `option` is one of the following: |
113f0e7c SW |
243 | |
244 | **calibrate** | |
245 | Run unit work calibration only and exit. | |
246 | ||
247 | **system** | |
248 | Show aggregate system idleness and unit work. | |
249 | ||
250 | **percpu** | |
251 | As **system** but also show per CPU idleness. | |
f80dba8d MT |
252 | |
253 | .. option:: --inflate-log=log | |
254 | ||
255 | Inflate and output compressed log. | |
256 | ||
257 | .. option:: --trigger-file=file | |
258 | ||
259 | Execute trigger cmd when file exists. | |
260 | ||
261 | .. option:: --trigger-timeout=t | |
262 | ||
263 | Execute trigger at this time. | |
264 | ||
265 | .. option:: --trigger=cmd | |
266 | ||
267 | Set this command as local trigger. | |
268 | ||
269 | .. option:: --trigger-remote=cmd | |
270 | ||
271 | Set this command as remote trigger. | |
272 | ||
273 | .. option:: --aux-path=path | |
274 | ||
275 | Use this path for fio state generated files. | |
276 | ||
277 | Any parameters following the options will be assumed to be job files, unless | |
278 | they match a job file parameter. Multiple job files can be listed and each job | |
279 | file will be regarded as a separate group. Fio will :option:`stonewall` | |
280 | execution between each group. | |
281 | ||
282 | ||
283 | Job file format | |
284 | --------------- | |
285 | ||
286 | As previously described, fio accepts one or more job files describing what it is | |
287 | supposed to do. The job file format is the classic ini file, where the names | |
c60ebc45 | 288 | enclosed in [] brackets define the job name. You are free to use any ASCII name |
f80dba8d MT |
289 | you want, except *global* which has special meaning. Following the job name is |
290 | a sequence of zero or more parameters, one per line, that define the behavior of | |
291 | the job. If the first character in a line is a ';' or a '#', the entire line is | |
292 | discarded as a comment. | |
293 | ||
294 | A *global* section sets defaults for the jobs described in that file. A job may | |
295 | override a *global* section parameter, and a job file may even have several | |
296 | *global* sections if so desired. A job is only affected by a *global* section | |
297 | residing above it. | |
298 | ||
299 | The :option:`--cmdhelp` option also lists all options. If used with an `option` | |
300 | argument, :option:`--cmdhelp` will detail the given `option`. | |
301 | ||
302 | See the `examples/` directory for inspiration on how to write job files. Note | |
303 | the copyright and license requirements currently apply to `examples/` files. | |
304 | ||
305 | So let's look at a really simple job file that defines two processes, each | |
306 | randomly reading from a 128MiB file: | |
307 | ||
308 | .. code-block:: ini | |
309 | ||
310 | ; -- start job file -- | |
311 | [global] | |
312 | rw=randread | |
313 | size=128m | |
314 | ||
315 | [job1] | |
316 | ||
317 | [job2] | |
318 | ||
319 | ; -- end job file -- | |
320 | ||
321 | As you can see, the job file sections themselves are empty as all the described | |
322 | parameters are shared. As no :option:`filename` option is given, fio makes up a | |
323 | `filename` for each of the jobs as it sees fit. On the command line, this job | |
324 | would look as follows:: | |
325 | ||
326 | $ fio --name=global --rw=randread --size=128m --name=job1 --name=job2 | |
327 | ||
328 | ||
329 | Let's look at an example that has a number of processes writing randomly to | |
330 | files: | |
331 | ||
332 | .. code-block:: ini | |
333 | ||
334 | ; -- start job file -- | |
335 | [random-writers] | |
336 | ioengine=libaio | |
337 | iodepth=4 | |
338 | rw=randwrite | |
339 | bs=32k | |
340 | direct=0 | |
341 | size=64m | |
342 | numjobs=4 | |
343 | ; -- end job file -- | |
344 | ||
345 | Here we have no *global* section, as we only have one job defined anyway. We | |
346 | want to use async I/O here, with a depth of 4 for each file. We also increased | |
347 | the buffer size used to 32KiB and define numjobs to 4 to fork 4 identical | |
348 | jobs. The result is 4 processes each randomly writing to their own 64MiB | |
349 | file. Instead of using the above job file, you could have given the parameters | |
350 | on the command line. For this case, you would specify:: | |
351 | ||
352 | $ fio --name=random-writers --ioengine=libaio --iodepth=4 --rw=randwrite --bs=32k --direct=0 --size=64m --numjobs=4 | |
353 | ||
354 | When fio is utilized as a basis of any reasonably large test suite, it might be | |
355 | desirable to share a set of standardized settings across multiple job files. | |
356 | Instead of copy/pasting such settings, any section may pull in an external | |
357 | :file:`filename.fio` file with *include filename* directive, as in the following | |
358 | example:: | |
359 | ||
360 | ; -- start job file including.fio -- | |
361 | [global] | |
362 | filename=/tmp/test | |
363 | filesize=1m | |
364 | include glob-include.fio | |
365 | ||
366 | [test] | |
367 | rw=randread | |
368 | bs=4k | |
369 | time_based=1 | |
370 | runtime=10 | |
371 | include test-include.fio | |
372 | ; -- end job file including.fio -- | |
373 | ||
374 | .. code-block:: ini | |
375 | ||
376 | ; -- start job file glob-include.fio -- | |
377 | thread=1 | |
378 | group_reporting=1 | |
379 | ; -- end job file glob-include.fio -- | |
380 | ||
381 | .. code-block:: ini | |
382 | ||
383 | ; -- start job file test-include.fio -- | |
384 | ioengine=libaio | |
385 | iodepth=4 | |
386 | ; -- end job file test-include.fio -- | |
387 | ||
388 | Settings pulled into a section apply to that section only (except *global* | |
389 | section). Include directives may be nested in that any included file may contain | |
390 | further include directive(s). Include files may not contain [] sections. | |
391 | ||
392 | ||
393 | Environment variables | |
394 | ~~~~~~~~~~~~~~~~~~~~~ | |
395 | ||
396 | Fio also supports environment variable expansion in job files. Any sub-string of | |
397 | the form ``${VARNAME}`` as part of an option value (in other words, on the right | |
398 | of the '='), will be expanded to the value of the environment variable called | |
399 | `VARNAME`. If no such environment variable is defined, or `VARNAME` is the | |
400 | empty string, the empty string will be substituted. | |
401 | ||
402 | As an example, let's look at a sample fio invocation and job file:: | |
403 | ||
404 | $ SIZE=64m NUMJOBS=4 fio jobfile.fio | |
405 | ||
406 | .. code-block:: ini | |
407 | ||
408 | ; -- start job file -- | |
409 | [random-writers] | |
410 | rw=randwrite | |
411 | size=${SIZE} | |
412 | numjobs=${NUMJOBS} | |
413 | ; -- end job file -- | |
414 | ||
415 | This will expand to the following equivalent job file at runtime: | |
416 | ||
417 | .. code-block:: ini | |
418 | ||
419 | ; -- start job file -- | |
420 | [random-writers] | |
421 | rw=randwrite | |
422 | size=64m | |
423 | numjobs=4 | |
424 | ; -- end job file -- | |
425 | ||
426 | Fio ships with a few example job files, you can also look there for inspiration. | |
427 | ||
428 | Reserved keywords | |
429 | ~~~~~~~~~~~~~~~~~ | |
430 | ||
431 | Additionally, fio has a set of reserved keywords that will be replaced | |
432 | internally with the appropriate value. Those keywords are: | |
433 | ||
434 | **$pagesize** | |
435 | ||
436 | The architecture page size of the running system. | |
437 | ||
438 | **$mb_memory** | |
439 | ||
440 | Megabytes of total memory in the system. | |
441 | ||
442 | **$ncpus** | |
443 | ||
444 | Number of online available CPUs. | |
445 | ||
446 | These can be used on the command line or in the job file, and will be | |
447 | automatically substituted with the current system values when the job is | |
448 | run. Simple math is also supported on these keywords, so you can perform actions | |
449 | like:: | |
450 | ||
b034c0dd | 451 | size=8*$mb_memory |
f80dba8d MT |
452 | |
453 | and get that properly expanded to 8 times the size of memory in the machine. | |
454 | ||
455 | ||
456 | Job file parameters | |
457 | ------------------- | |
458 | ||
459 | This section describes in details each parameter associated with a job. Some | |
460 | parameters take an option of a given type, such as an integer or a | |
461 | string. Anywhere a numeric value is required, an arithmetic expression may be | |
462 | used, provided it is surrounded by parentheses. Supported operators are: | |
463 | ||
464 | - addition (+) | |
465 | - subtraction (-) | |
466 | - multiplication (*) | |
467 | - division (/) | |
468 | - modulus (%) | |
469 | - exponentiation (^) | |
470 | ||
471 | For time values in expressions, units are microseconds by default. This is | |
472 | different than for time values not in expressions (not enclosed in | |
473 | parentheses). The following types are used: | |
474 | ||
475 | ||
476 | Parameter types | |
477 | ~~~~~~~~~~~~~~~ | |
478 | ||
479 | **str** | |
b034c0dd | 480 | String: A sequence of alphanumeric characters. |
f80dba8d MT |
481 | |
482 | **time** | |
008d0feb SW |
483 | Integer with possible time suffix. Without a unit value is interpreted as |
484 | seconds unless otherwise specified. Accepts a suffix of 'd' for days, 'h' for | |
485 | hours, 'm' for minutes, 's' for seconds, 'ms' (or 'msec') for milliseconds and | |
486 | 'us' (or 'usec') for microseconds. For example, use 10m for 10 minutes. | |
f80dba8d MT |
487 | |
488 | .. _int: | |
489 | ||
490 | **int** | |
491 | Integer. A whole number value, which may contain an integer prefix | |
492 | and an integer suffix: | |
493 | ||
b034c0dd | 494 | [*integer prefix*] **number** [*integer suffix*] |
f80dba8d MT |
495 | |
496 | The optional *integer prefix* specifies the number's base. The default | |
497 | is decimal. *0x* specifies hexadecimal. | |
498 | ||
499 | The optional *integer suffix* specifies the number's units, and includes an | |
500 | optional unit prefix and an optional unit. For quantities of data, the | |
947e0fe0 SW |
501 | default unit is bytes. For quantities of time, the default unit is seconds |
502 | unless otherwise specified. | |
f80dba8d | 503 | |
9207a0cb | 504 | With :option:`kb_base`\=1000, fio follows international standards for unit |
f80dba8d MT |
505 | prefixes. To specify power-of-10 decimal values defined in the |
506 | International System of Units (SI): | |
507 | ||
508 | * *Ki* -- means kilo (K) or 1000 | |
509 | * *Mi* -- means mega (M) or 1000**2 | |
510 | * *Gi* -- means giga (G) or 1000**3 | |
511 | * *Ti* -- means tera (T) or 1000**4 | |
512 | * *Pi* -- means peta (P) or 1000**5 | |
513 | ||
514 | To specify power-of-2 binary values defined in IEC 80000-13: | |
515 | ||
516 | * *k* -- means kibi (Ki) or 1024 | |
517 | * *M* -- means mebi (Mi) or 1024**2 | |
518 | * *G* -- means gibi (Gi) or 1024**3 | |
519 | * *T* -- means tebi (Ti) or 1024**4 | |
520 | * *P* -- means pebi (Pi) or 1024**5 | |
521 | ||
9207a0cb | 522 | With :option:`kb_base`\=1024 (the default), the unit prefixes are opposite |
f80dba8d MT |
523 | from those specified in the SI and IEC 80000-13 standards to provide |
524 | compatibility with old scripts. For example, 4k means 4096. | |
525 | ||
526 | For quantities of data, an optional unit of 'B' may be included | |
b8f7e412 | 527 | (e.g., 'kB' is the same as 'k'). |
f80dba8d MT |
528 | |
529 | The *integer suffix* is not case sensitive (e.g., m/mi mean mebi/mega, | |
530 | not milli). 'b' and 'B' both mean byte, not bit. | |
531 | ||
9207a0cb | 532 | Examples with :option:`kb_base`\=1000: |
f80dba8d MT |
533 | |
534 | * *4 KiB*: 4096, 4096b, 4096B, 4ki, 4kib, 4kiB, 4Ki, 4KiB | |
535 | * *1 MiB*: 1048576, 1mi, 1024ki | |
536 | * *1 MB*: 1000000, 1m, 1000k | |
537 | * *1 TiB*: 1099511627776, 1ti, 1024gi, 1048576mi | |
538 | * *1 TB*: 1000000000, 1t, 1000m, 1000000k | |
539 | ||
9207a0cb | 540 | Examples with :option:`kb_base`\=1024 (default): |
f80dba8d MT |
541 | |
542 | * *4 KiB*: 4096, 4096b, 4096B, 4k, 4kb, 4kB, 4K, 4KB | |
543 | * *1 MiB*: 1048576, 1m, 1024k | |
544 | * *1 MB*: 1000000, 1mi, 1000ki | |
545 | * *1 TiB*: 1099511627776, 1t, 1024g, 1048576m | |
546 | * *1 TB*: 1000000000, 1ti, 1000mi, 1000000ki | |
547 | ||
548 | To specify times (units are not case sensitive): | |
549 | ||
550 | * *D* -- means days | |
551 | * *H* -- means hours | |
4502cb42 | 552 | * *M* -- means minutes |
f80dba8d MT |
553 | * *s* -- or sec means seconds (default) |
554 | * *ms* -- or *msec* means milliseconds | |
555 | * *us* -- or *usec* means microseconds | |
556 | ||
557 | If the option accepts an upper and lower range, use a colon ':' or | |
558 | minus '-' to separate such values. See :ref:`irange <irange>`. | |
4502cb42 SW |
559 | If the lower value specified happens to be larger than the upper value |
560 | the two values are swapped. | |
f80dba8d MT |
561 | |
562 | .. _bool: | |
563 | ||
564 | **bool** | |
565 | Boolean. Usually parsed as an integer, however only defined for | |
566 | true and false (1 and 0). | |
567 | ||
568 | .. _irange: | |
569 | ||
570 | **irange** | |
571 | Integer range with suffix. Allows value range to be given, such as | |
c60ebc45 | 572 | 1024-4096. A colon may also be used as the separator, e.g. 1k:4k. If the |
f80dba8d MT |
573 | option allows two sets of ranges, they can be specified with a ',' or '/' |
574 | delimiter: 1k-4k/8k-32k. Also see :ref:`int <int>`. | |
575 | ||
576 | **float_list** | |
577 | A list of floating point numbers, separated by a ':' character. | |
578 | ||
579 | ||
580 | Units | |
581 | ~~~~~ | |
582 | ||
583 | .. option:: kb_base=int | |
584 | ||
585 | Select the interpretation of unit prefixes in input parameters. | |
586 | ||
587 | **1000** | |
588 | Inputs comply with IEC 80000-13 and the International | |
589 | System of Units (SI). Use: | |
590 | ||
591 | - power-of-2 values with IEC prefixes (e.g., KiB) | |
592 | - power-of-10 values with SI prefixes (e.g., kB) | |
593 | ||
594 | **1024** | |
595 | Compatibility mode (default). To avoid breaking old scripts: | |
596 | ||
597 | - power-of-2 values with SI prefixes | |
598 | - power-of-10 values with IEC prefixes | |
599 | ||
600 | See :option:`bs` for more details on input parameters. | |
601 | ||
602 | Outputs always use correct prefixes. Most outputs include both | |
603 | side-by-side, like:: | |
604 | ||
605 | bw=2383.3kB/s (2327.4KiB/s) | |
606 | ||
607 | If only one value is reported, then kb_base selects the one to use: | |
608 | ||
609 | **1000** -- SI prefixes | |
610 | ||
611 | **1024** -- IEC prefixes | |
612 | ||
613 | .. option:: unit_base=int | |
614 | ||
615 | Base unit for reporting. Allowed values are: | |
616 | ||
617 | **0** | |
618 | Use auto-detection (default). | |
619 | **8** | |
620 | Byte based. | |
621 | **1** | |
622 | Bit based. | |
623 | ||
624 | ||
625 | With the above in mind, here follows the complete list of fio job parameters. | |
626 | ||
627 | ||
628 | Job description | |
629 | ~~~~~~~~~~~~~~~ | |
630 | ||
631 | .. option:: name=str | |
632 | ||
633 | ASCII name of the job. This may be used to override the name printed by fio | |
634 | for this job. Otherwise the job name is used. On the command line this | |
635 | parameter has the special purpose of also signaling the start of a new job. | |
636 | ||
637 | .. option:: description=str | |
638 | ||
639 | Text description of the job. Doesn't do anything except dump this text | |
640 | description when this job is run. It's not parsed. | |
641 | ||
642 | .. option:: loops=int | |
643 | ||
644 | Run the specified number of iterations of this job. Used to repeat the same | |
645 | workload a given number of times. Defaults to 1. | |
646 | ||
647 | .. option:: numjobs=int | |
648 | ||
79591fa9 TK |
649 | Create the specified number of clones of this job. Each clone of job |
650 | is spawned as an independent thread or process. May be used to setup a | |
f80dba8d MT |
651 | larger number of threads/processes doing the same thing. Each thread is |
652 | reported separately; to see statistics for all clones as a whole, use | |
653 | :option:`group_reporting` in conjunction with :option:`new_group`. | |
a47b697c | 654 | See :option:`--max-jobs`. Default: 1. |
f80dba8d MT |
655 | |
656 | ||
657 | Time related parameters | |
658 | ~~~~~~~~~~~~~~~~~~~~~~~ | |
659 | ||
660 | .. option:: runtime=time | |
661 | ||
f75ede1d | 662 | Tell fio to terminate processing after the specified period of time. It |
f80dba8d | 663 | can be quite hard to determine for how long a specified job will run, so |
f75ede1d | 664 | this parameter is handy to cap the total runtime to a given time. When |
947e0fe0 | 665 | the unit is omitted, the value is intepreted in seconds. |
f80dba8d MT |
666 | |
667 | .. option:: time_based | |
668 | ||
669 | If set, fio will run for the duration of the :option:`runtime` specified | |
670 | even if the file(s) are completely read or written. It will simply loop over | |
671 | the same workload as many times as the :option:`runtime` allows. | |
672 | ||
a881438b | 673 | .. option:: startdelay=irange(time) |
f80dba8d | 674 | |
947e0fe0 SW |
675 | Delay the start of job for the specified amount of time. Can be a single |
676 | value or a range. When given as a range, each thread will choose a value | |
677 | randomly from within the range. Value is in seconds if a unit is omitted. | |
f80dba8d MT |
678 | |
679 | .. option:: ramp_time=time | |
680 | ||
681 | If set, fio will run the specified workload for this amount of time before | |
682 | logging any performance numbers. Useful for letting performance settle | |
683 | before logging results, thus minimizing the runtime required for stable | |
684 | results. Note that the ``ramp_time`` is considered lead in time for a job, | |
685 | thus it will increase the total runtime if a special timeout or | |
f75ede1d SW |
686 | :option:`runtime` is specified. When the unit is omitted, the value is |
687 | given in seconds. | |
f80dba8d MT |
688 | |
689 | .. option:: clocksource=str | |
690 | ||
691 | Use the given clocksource as the base of timing. The supported options are: | |
692 | ||
693 | **gettimeofday** | |
694 | :manpage:`gettimeofday(2)` | |
695 | ||
696 | **clock_gettime** | |
697 | :manpage:`clock_gettime(2)` | |
698 | ||
699 | **cpu** | |
700 | Internal CPU clock source | |
701 | ||
702 | cpu is the preferred clocksource if it is reliable, as it is very fast (and | |
703 | fio is heavy on time calls). Fio will automatically use this clocksource if | |
704 | it's supported and considered reliable on the system it is running on, | |
705 | unless another clocksource is specifically set. For x86/x86-64 CPUs, this | |
706 | means supporting TSC Invariant. | |
707 | ||
708 | .. option:: gtod_reduce=bool | |
709 | ||
710 | Enable all of the :manpage:`gettimeofday(2)` reducing options | |
f75ede1d | 711 | (:option:`disable_clat`, :option:`disable_slat`, :option:`disable_bw_measurement`) plus |
f80dba8d MT |
712 | reduce precision of the timeout somewhat to really shrink the |
713 | :manpage:`gettimeofday(2)` call count. With this option enabled, we only do | |
714 | about 0.4% of the :manpage:`gettimeofday(2)` calls we would have done if all | |
715 | time keeping was enabled. | |
716 | ||
717 | .. option:: gtod_cpu=int | |
718 | ||
719 | Sometimes it's cheaper to dedicate a single thread of execution to just | |
720 | getting the current time. Fio (and databases, for instance) are very | |
721 | intensive on :manpage:`gettimeofday(2)` calls. With this option, you can set | |
722 | one CPU aside for doing nothing but logging current time to a shared memory | |
723 | location. Then the other threads/processes that run I/O workloads need only | |
724 | copy that segment, instead of entering the kernel with a | |
725 | :manpage:`gettimeofday(2)` call. The CPU set aside for doing these time | |
726 | calls will be excluded from other uses. Fio will manually clear it from the | |
727 | CPU mask of other jobs. | |
728 | ||
729 | ||
730 | Target file/device | |
731 | ~~~~~~~~~~~~~~~~~~ | |
732 | ||
733 | .. option:: directory=str | |
734 | ||
735 | Prefix filenames with this directory. Used to place files in a different | |
736 | location than :file:`./`. You can specify a number of directories by | |
737 | separating the names with a ':' character. These directories will be | |
02dd2689 | 738 | assigned equally distributed to job clones created by :option:`numjobs` as |
f80dba8d MT |
739 | long as they are using generated filenames. If specific `filename(s)` are |
740 | set fio will use the first listed directory, and thereby matching the | |
741 | `filename` semantic which generates a file each clone if not specified, but | |
742 | let all clones use the same if set. | |
743 | ||
02dd2689 SW |
744 | See the :option:`filename` option for information on how to escape "``:``" and |
745 | "``\``" characters within the directory path itself. | |
f80dba8d MT |
746 | |
747 | .. option:: filename=str | |
748 | ||
749 | Fio normally makes up a `filename` based on the job name, thread number, and | |
02dd2689 SW |
750 | file number (see :option:`filename_format`). If you want to share files |
751 | between threads in a job or several | |
79591fa9 TK |
752 | jobs with fixed file paths, specify a `filename` for each of them to override |
753 | the default. If the ioengine is file based, you can specify a number of files | |
754 | by separating the names with a ':' colon. So if you wanted a job to open | |
755 | :file:`/dev/sda` and :file:`/dev/sdb` as the two working files, you would use | |
756 | ``filename=/dev/sda:/dev/sdb``. This also means that whenever this option is | |
757 | specified, :option:`nrfiles` is ignored. The size of regular files specified | |
02dd2689 | 758 | by this option will be :option:`size` divided by number of files unless an |
79591fa9 TK |
759 | explicit size is specified by :option:`filesize`. |
760 | ||
02dd2689 SW |
761 | Each colon and backslash in the wanted path must be escaped with a ``\`` |
762 | character. For instance, if the path is :file:`/dev/dsk/foo@3,0:c` then you | |
763 | would use ``filename=/dev/dsk/foo@3,0\:c`` and if the path is | |
764 | :file:`F:\\filename` then you would use ``filename=F\:\\filename``. | |
765 | ||
f80dba8d MT |
766 | On Windows, disk devices are accessed as :file:`\\\\.\\PhysicalDrive0` for |
767 | the first device, :file:`\\\\.\\PhysicalDrive1` for the second etc. | |
768 | Note: Windows and FreeBSD prevent write access to areas | |
02dd2689 SW |
769 | of the disk containing in-use data (e.g. filesystems). |
770 | ||
771 | The filename "`-`" is a reserved name, meaning *stdin* or *stdout*. Which | |
772 | of the two depends on the read/write direction set. | |
f80dba8d MT |
773 | |
774 | .. option:: filename_format=str | |
775 | ||
776 | If sharing multiple files between jobs, it is usually necessary to have fio | |
777 | generate the exact names that you want. By default, fio will name a file | |
778 | based on the default file format specification of | |
779 | :file:`jobname.jobnumber.filenumber`. With this option, that can be | |
780 | customized. Fio will recognize and replace the following keywords in this | |
781 | string: | |
782 | ||
783 | **$jobname** | |
784 | The name of the worker thread or process. | |
785 | **$jobnum** | |
786 | The incremental number of the worker thread or process. | |
787 | **$filenum** | |
788 | The incremental number of the file for that worker thread or | |
789 | process. | |
790 | ||
791 | To have dependent jobs share a set of files, this option can be set to have | |
792 | fio generate filenames that are shared between the two. For instance, if | |
793 | :file:`testfiles.$filenum` is specified, file number 4 for any job will be | |
794 | named :file:`testfiles.4`. The default of :file:`$jobname.$jobnum.$filenum` | |
795 | will be used if no other format specifier is given. | |
796 | ||
797 | .. option:: unique_filename=bool | |
798 | ||
799 | To avoid collisions between networked clients, fio defaults to prefixing any | |
800 | generated filenames (with a directory specified) with the source of the | |
801 | client connecting. To disable this behavior, set this option to 0. | |
802 | ||
803 | .. option:: opendir=str | |
804 | ||
805 | Recursively open any files below directory `str`. | |
806 | ||
807 | .. option:: lockfile=str | |
808 | ||
809 | Fio defaults to not locking any files before it does I/O to them. If a file | |
810 | or file descriptor is shared, fio can serialize I/O to that file to make the | |
811 | end result consistent. This is usual for emulating real workloads that share | |
812 | files. The lock modes are: | |
813 | ||
814 | **none** | |
815 | No locking. The default. | |
816 | **exclusive** | |
817 | Only one thread or process may do I/O at a time, excluding all | |
818 | others. | |
819 | **readwrite** | |
820 | Read-write locking on the file. Many readers may | |
821 | access the file at the same time, but writes get exclusive access. | |
822 | ||
823 | .. option:: nrfiles=int | |
824 | ||
79591fa9 TK |
825 | Number of files to use for this job. Defaults to 1. The size of files |
826 | will be :option:`size` divided by this unless explicit size is specified by | |
827 | :option:`filesize`. Files are created for each thread separately, and each | |
828 | file will have a file number within its name by default, as explained in | |
829 | :option:`filename` section. | |
830 | ||
f80dba8d MT |
831 | |
832 | .. option:: openfiles=int | |
833 | ||
834 | Number of files to keep open at the same time. Defaults to the same as | |
835 | :option:`nrfiles`, can be set smaller to limit the number simultaneous | |
836 | opens. | |
837 | ||
838 | .. option:: file_service_type=str | |
839 | ||
840 | Defines how fio decides which file from a job to service next. The following | |
841 | types are defined: | |
842 | ||
843 | **random** | |
844 | Choose a file at random. | |
845 | ||
846 | **roundrobin** | |
847 | Round robin over opened files. This is the default. | |
848 | ||
849 | **sequential** | |
850 | Finish one file before moving on to the next. Multiple files can | |
851 | still be open depending on 'openfiles'. | |
852 | ||
853 | **zipf** | |
c60ebc45 | 854 | Use a *Zipf* distribution to decide what file to access. |
f80dba8d MT |
855 | |
856 | **pareto** | |
c60ebc45 | 857 | Use a *Pareto* distribution to decide what file to access. |
f80dba8d | 858 | |
dd3503d3 | 859 | **normal** |
c60ebc45 | 860 | Use a *Gaussian* (normal) distribution to decide what file to |
f80dba8d MT |
861 | access. |
862 | ||
dd3503d3 SW |
863 | **gauss** |
864 | Alias for normal. | |
865 | ||
f80dba8d MT |
866 | For *random*, *roundrobin*, and *sequential*, a postfix can be appended to |
867 | tell fio how many I/Os to issue before switching to a new file. For example, | |
868 | specifying ``file_service_type=random:8`` would cause fio to issue | |
869 | 8 I/Os before selecting a new file at random. For the non-uniform | |
870 | distributions, a floating point postfix can be given to influence how the | |
871 | distribution is skewed. See :option:`random_distribution` for a description | |
872 | of how that would work. | |
873 | ||
874 | .. option:: ioscheduler=str | |
875 | ||
876 | Attempt to switch the device hosting the file to the specified I/O scheduler | |
877 | before running. | |
878 | ||
879 | .. option:: create_serialize=bool | |
880 | ||
881 | If true, serialize the file creation for the jobs. This may be handy to | |
882 | avoid interleaving of data files, which may greatly depend on the filesystem | |
a47b697c | 883 | used and even the number of processors in the system. Default: true. |
f80dba8d MT |
884 | |
885 | .. option:: create_fsync=bool | |
886 | ||
22413915 | 887 | :manpage:`fsync(2)` the data file after creation. This is the default. |
f80dba8d MT |
888 | |
889 | .. option:: create_on_open=bool | |
890 | ||
730bd7d9 SW |
891 | If true, don't pre-create files but allow the job's open() to create a file |
892 | when it's time to do I/O. Default: false -- pre-create all necessary files | |
893 | when the job starts. | |
f80dba8d MT |
894 | |
895 | .. option:: create_only=bool | |
896 | ||
897 | If true, fio will only run the setup phase of the job. If files need to be | |
4502cb42 | 898 | laid out or updated on disk, only that will be done -- the actual job contents |
a47b697c | 899 | are not executed. Default: false. |
f80dba8d MT |
900 | |
901 | .. option:: allow_file_create=bool | |
902 | ||
730bd7d9 SW |
903 | If true, fio is permitted to create files as part of its workload. If this |
904 | option is false, then fio will error out if | |
f80dba8d MT |
905 | the files it needs to use don't already exist. Default: true. |
906 | ||
907 | .. option:: allow_mounted_write=bool | |
908 | ||
c60ebc45 | 909 | If this isn't set, fio will abort jobs that are destructive (e.g. that write) |
f80dba8d MT |
910 | to what appears to be a mounted device or partition. This should help catch |
911 | creating inadvertently destructive tests, not realizing that the test will | |
b1db0375 TK |
912 | destroy data on the mounted file system. Note that some platforms don't allow |
913 | writing against a mounted device regardless of this option. Default: false. | |
f80dba8d MT |
914 | |
915 | .. option:: pre_read=bool | |
916 | ||
917 | If this is given, files will be pre-read into memory before starting the | |
918 | given I/O operation. This will also clear the :option:`invalidate` flag, | |
919 | since it is pointless to pre-read and then drop the cache. This will only | |
920 | work for I/O engines that are seek-able, since they allow you to read the | |
a47b697c SW |
921 | same data multiple times. Thus it will not work on non-seekable I/O engines |
922 | (e.g. network, splice). Default: false. | |
f80dba8d MT |
923 | |
924 | .. option:: unlink=bool | |
925 | ||
926 | Unlink the job files when done. Not the default, as repeated runs of that | |
a47b697c SW |
927 | job would then waste time recreating the file set again and again. Default: |
928 | false. | |
f80dba8d MT |
929 | |
930 | .. option:: unlink_each_loop=bool | |
931 | ||
a47b697c | 932 | Unlink job files after each iteration or loop. Default: false. |
f80dba8d MT |
933 | |
934 | .. option:: zonesize=int | |
935 | ||
936 | Divide a file into zones of the specified size. See :option:`zoneskip`. | |
937 | ||
938 | .. option:: zonerange=int | |
939 | ||
940 | Give size of an I/O zone. See :option:`zoneskip`. | |
941 | ||
942 | .. option:: zoneskip=int | |
943 | ||
944 | Skip the specified number of bytes when :option:`zonesize` data has been | |
945 | read. The two zone options can be used to only do I/O on zones of a file. | |
946 | ||
947 | ||
948 | I/O type | |
949 | ~~~~~~~~ | |
950 | ||
951 | .. option:: direct=bool | |
952 | ||
953 | If value is true, use non-buffered I/O. This is usually O_DIRECT. Note that | |
954 | ZFS on Solaris doesn't support direct I/O. On Windows the synchronous | |
955 | ioengines don't support direct I/O. Default: false. | |
956 | ||
957 | .. option:: atomic=bool | |
958 | ||
959 | If value is true, attempt to use atomic direct I/O. Atomic writes are | |
960 | guaranteed to be stable once acknowledged by the operating system. Only | |
961 | Linux supports O_ATOMIC right now. | |
962 | ||
963 | .. option:: buffered=bool | |
964 | ||
965 | If value is true, use buffered I/O. This is the opposite of the | |
966 | :option:`direct` option. Defaults to true. | |
967 | ||
968 | .. option:: readwrite=str, rw=str | |
969 | ||
970 | Type of I/O pattern. Accepted values are: | |
971 | ||
972 | **read** | |
973 | Sequential reads. | |
974 | **write** | |
975 | Sequential writes. | |
976 | **trim** | |
977 | Sequential trims (Linux block devices only). | |
f80dba8d MT |
978 | **randread** |
979 | Random reads. | |
2831be97 SW |
980 | **randwrite** |
981 | Random writes. | |
f80dba8d MT |
982 | **randtrim** |
983 | Random trims (Linux block devices only). | |
984 | **rw,readwrite** | |
985 | Sequential mixed reads and writes. | |
986 | **randrw** | |
987 | Random mixed reads and writes. | |
988 | **trimwrite** | |
989 | Sequential trim+write sequences. Blocks will be trimmed first, | |
990 | then the same blocks will be written to. | |
991 | ||
992 | Fio defaults to read if the option is not specified. For the mixed I/O | |
993 | types, the default is to split them 50/50. For certain types of I/O the | |
730bd7d9 SW |
994 | result may still be skewed a bit, since the speed may be different. |
995 | ||
996 | It is possible to specify the number of I/Os to do before getting a new | |
997 | offset by appending ``:<nr>`` to the end of the string given. For a | |
f80dba8d MT |
998 | random read, it would look like ``rw=randread:8`` for passing in an offset |
999 | modifier with a value of 8. If the suffix is used with a sequential I/O | |
730bd7d9 SW |
1000 | pattern, then the *<nr>* value specified will be **added** to the generated |
1001 | offset for each I/O turning sequential I/O into sequential I/O with holes. | |
1002 | For instance, using ``rw=write:4k`` will skip 4k for every write. Also see | |
1003 | the :option:`rw_sequencer` option. | |
f80dba8d MT |
1004 | |
1005 | .. option:: rw_sequencer=str | |
1006 | ||
1007 | If an offset modifier is given by appending a number to the ``rw=<str>`` | |
1008 | line, then this option controls how that number modifies the I/O offset | |
1009 | being generated. Accepted values are: | |
1010 | ||
1011 | **sequential** | |
1012 | Generate sequential offset. | |
1013 | **identical** | |
1014 | Generate the same offset. | |
1015 | ||
1016 | ``sequential`` is only useful for random I/O, where fio would normally | |
c60ebc45 | 1017 | generate a new random offset for every I/O. If you append e.g. 8 to randread, |
f80dba8d MT |
1018 | you would get a new random offset for every 8 I/O's. The result would be a |
1019 | seek for only every 8 I/O's, instead of for every I/O. Use ``rw=randread:8`` | |
1020 | to specify that. As sequential I/O is already sequential, setting | |
1021 | ``sequential`` for that would not result in any differences. ``identical`` | |
1022 | behaves in a similar fashion, except it sends the same offset 8 number of | |
1023 | times before generating a new offset. | |
1024 | ||
1025 | .. option:: unified_rw_reporting=bool | |
1026 | ||
1027 | Fio normally reports statistics on a per data direction basis, meaning that | |
1028 | reads, writes, and trims are accounted and reported separately. If this | |
1029 | option is set fio sums the results and report them as "mixed" instead. | |
1030 | ||
1031 | .. option:: randrepeat=bool | |
1032 | ||
1033 | Seed the random number generator used for random I/O patterns in a | |
1034 | predictable way so the pattern is repeatable across runs. Default: true. | |
1035 | ||
1036 | .. option:: allrandrepeat=bool | |
1037 | ||
1038 | Seed all random number generators in a predictable way so results are | |
1039 | repeatable across runs. Default: false. | |
1040 | ||
1041 | .. option:: randseed=int | |
1042 | ||
1043 | Seed the random number generators based on this seed value, to be able to | |
1044 | control what sequence of output is being generated. If not set, the random | |
1045 | sequence depends on the :option:`randrepeat` setting. | |
1046 | ||
1047 | .. option:: fallocate=str | |
1048 | ||
1049 | Whether pre-allocation is performed when laying down files. | |
1050 | Accepted values are: | |
1051 | ||
1052 | **none** | |
1053 | Do not pre-allocate space. | |
1054 | ||
2c3e17be SW |
1055 | **native** |
1056 | Use a platform's native pre-allocation call but fall back to | |
1057 | **none** behavior if it fails/is not implemented. | |
1058 | ||
f80dba8d MT |
1059 | **posix** |
1060 | Pre-allocate via :manpage:`posix_fallocate(3)`. | |
1061 | ||
1062 | **keep** | |
1063 | Pre-allocate via :manpage:`fallocate(2)` with | |
1064 | FALLOC_FL_KEEP_SIZE set. | |
1065 | ||
1066 | **0** | |
1067 | Backward-compatible alias for **none**. | |
1068 | ||
1069 | **1** | |
1070 | Backward-compatible alias for **posix**. | |
1071 | ||
1072 | May not be available on all supported platforms. **keep** is only available | |
2c3e17be SW |
1073 | on Linux. If using ZFS on Solaris this cannot be set to **posix** |
1074 | because ZFS doesn't support pre-allocation. Default: **native** if any | |
1075 | pre-allocation methods are available, **none** if not. | |
f80dba8d MT |
1076 | |
1077 | .. option:: fadvise_hint=str | |
1078 | ||
1079 | Use :manpage:`posix_fadvise(2)` to advise the kernel on what I/O patterns | |
1080 | are likely to be issued. Accepted values are: | |
1081 | ||
1082 | **0** | |
1083 | Backwards-compatible hint for "no hint". | |
1084 | ||
1085 | **1** | |
1086 | Backwards compatible hint for "advise with fio workload type". This | |
1087 | uses **FADV_RANDOM** for a random workload, and **FADV_SEQUENTIAL** | |
1088 | for a sequential workload. | |
1089 | ||
1090 | **sequential** | |
1091 | Advise using **FADV_SEQUENTIAL**. | |
1092 | ||
1093 | **random** | |
1094 | Advise using **FADV_RANDOM**. | |
1095 | ||
8f4b9f24 | 1096 | .. option:: write_hint=str |
f80dba8d | 1097 | |
8f4b9f24 JA |
1098 | Use :manpage:`fcntl(2)` to advise the kernel what life time to expect |
1099 | from a write. Only supported on Linux, as of version 4.13. Accepted | |
1100 | values are: | |
1101 | ||
1102 | **none** | |
1103 | No particular life time associated with this file. | |
1104 | ||
1105 | **short** | |
1106 | Data written to this file has a short life time. | |
1107 | ||
1108 | **medium** | |
1109 | Data written to this file has a medium life time. | |
1110 | ||
1111 | **long** | |
1112 | Data written to this file has a long life time. | |
1113 | ||
1114 | **extreme** | |
1115 | Data written to this file has a very long life time. | |
1116 | ||
1117 | The values are all relative to each other, and no absolute meaning | |
1118 | should be associated with them. | |
f80dba8d MT |
1119 | |
1120 | .. option:: offset=int | |
1121 | ||
82dbb8cb TK |
1122 | Start I/O at the provided offset in the file, given as either a fixed size in |
1123 | bytes or a percentage. If a percentage is given, the next ``blockalign``-ed | |
1124 | offset will be used. Data before the given offset will not be touched. This | |
89978a6b BW |
1125 | effectively caps the file size at `real_size - offset`. Can be combined with |
1126 | :option:`size` to constrain the start and end range of the I/O workload. | |
44bb1142 TK |
1127 | A percentage can be specified by a number between 1 and 100 followed by '%', |
1128 | for example, ``offset=20%`` to specify 20%. | |
f80dba8d MT |
1129 | |
1130 | .. option:: offset_increment=int | |
1131 | ||
1132 | If this is provided, then the real offset becomes `offset + offset_increment | |
1133 | * thread_number`, where the thread number is a counter that starts at 0 and | |
1134 | is incremented for each sub-job (i.e. when :option:`numjobs` option is | |
1135 | specified). This option is useful if there are several jobs which are | |
1136 | intended to operate on a file in parallel disjoint segments, with even | |
1137 | spacing between the starting points. | |
1138 | ||
1139 | .. option:: number_ios=int | |
1140 | ||
c60ebc45 | 1141 | Fio will normally perform I/Os until it has exhausted the size of the region |
f80dba8d MT |
1142 | set by :option:`size`, or if it exhaust the allocated time (or hits an error |
1143 | condition). With this setting, the range/size can be set independently of | |
c60ebc45 | 1144 | the number of I/Os to perform. When fio reaches this number, it will exit |
f80dba8d MT |
1145 | normally and report status. Note that this does not extend the amount of I/O |
1146 | that will be done, it will only stop fio if this condition is met before | |
1147 | other end-of-job criteria. | |
1148 | ||
1149 | .. option:: fsync=int | |
1150 | ||
730bd7d9 SW |
1151 | If writing to a file, issue an :manpage:`fsync(2)` (or its equivalent) of |
1152 | the dirty data for every number of blocks given. For example, if you give 32 | |
1153 | as a parameter, fio will sync the file after every 32 writes issued. If fio is | |
1154 | using non-buffered I/O, we may not sync the file. The exception is the sg | |
1155 | I/O engine, which synchronizes the disk cache anyway. Defaults to 0, which | |
1156 | means fio does not periodically issue and wait for a sync to complete. Also | |
1157 | see :option:`end_fsync` and :option:`fsync_on_close`. | |
f80dba8d MT |
1158 | |
1159 | .. option:: fdatasync=int | |
1160 | ||
1161 | Like :option:`fsync` but uses :manpage:`fdatasync(2)` to only sync data and | |
000a5f1c | 1162 | not metadata blocks. In Windows, FreeBSD, and DragonFlyBSD there is no |
730bd7d9 SW |
1163 | :manpage:`fdatasync(2)` so this falls back to using :manpage:`fsync(2)`. |
1164 | Defaults to 0, which means fio does not periodically issue and wait for a | |
1165 | data-only sync to complete. | |
f80dba8d MT |
1166 | |
1167 | .. option:: write_barrier=int | |
1168 | ||
2831be97 | 1169 | Make every `N-th` write a barrier write. |
f80dba8d MT |
1170 | |
1171 | .. option:: sync_file_range=str:val | |
1172 | ||
1173 | Use :manpage:`sync_file_range(2)` for every `val` number of write | |
1174 | operations. Fio will track range of writes that have happened since the last | |
1175 | :manpage:`sync_file_range(2)` call. `str` can currently be one or more of: | |
1176 | ||
1177 | **wait_before** | |
1178 | SYNC_FILE_RANGE_WAIT_BEFORE | |
1179 | **write** | |
1180 | SYNC_FILE_RANGE_WRITE | |
1181 | **wait_after** | |
1182 | SYNC_FILE_RANGE_WAIT_AFTER | |
1183 | ||
1184 | So if you do ``sync_file_range=wait_before,write:8``, fio would use | |
1185 | ``SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE`` for every 8 | |
1186 | writes. Also see the :manpage:`sync_file_range(2)` man page. This option is | |
1187 | Linux specific. | |
1188 | ||
1189 | .. option:: overwrite=bool | |
1190 | ||
1191 | If true, writes to a file will always overwrite existing data. If the file | |
1192 | doesn't already exist, it will be created before the write phase begins. If | |
1193 | the file exists and is large enough for the specified write phase, nothing | |
a47b697c | 1194 | will be done. Default: false. |
f80dba8d MT |
1195 | |
1196 | .. option:: end_fsync=bool | |
1197 | ||
a47b697c SW |
1198 | If true, :manpage:`fsync(2)` file contents when a write stage has completed. |
1199 | Default: false. | |
f80dba8d MT |
1200 | |
1201 | .. option:: fsync_on_close=bool | |
1202 | ||
1203 | If true, fio will :manpage:`fsync(2)` a dirty file on close. This differs | |
a47b697c SW |
1204 | from :option:`end_fsync` in that it will happen on every file close, not |
1205 | just at the end of the job. Default: false. | |
f80dba8d MT |
1206 | |
1207 | .. option:: rwmixread=int | |
1208 | ||
1209 | Percentage of a mixed workload that should be reads. Default: 50. | |
1210 | ||
1211 | .. option:: rwmixwrite=int | |
1212 | ||
1213 | Percentage of a mixed workload that should be writes. If both | |
1214 | :option:`rwmixread` and :option:`rwmixwrite` is given and the values do not | |
1215 | add up to 100%, the latter of the two will be used to override the | |
1216 | first. This may interfere with a given rate setting, if fio is asked to | |
1217 | limit reads or writes to a certain rate. If that is the case, then the | |
1218 | distribution may be skewed. Default: 50. | |
1219 | ||
1220 | .. option:: random_distribution=str:float[,str:float][,str:float] | |
1221 | ||
1222 | By default, fio will use a completely uniform random distribution when asked | |
1223 | to perform random I/O. Sometimes it is useful to skew the distribution in | |
1224 | specific ways, ensuring that some parts of the data is more hot than others. | |
1225 | fio includes the following distribution models: | |
1226 | ||
1227 | **random** | |
1228 | Uniform random distribution | |
1229 | ||
1230 | **zipf** | |
1231 | Zipf distribution | |
1232 | ||
1233 | **pareto** | |
1234 | Pareto distribution | |
1235 | ||
b2f4b559 | 1236 | **normal** |
c60ebc45 | 1237 | Normal (Gaussian) distribution |
f80dba8d MT |
1238 | |
1239 | **zoned** | |
1240 | Zoned random distribution | |
1241 | ||
1242 | When using a **zipf** or **pareto** distribution, an input value is also | |
1243 | needed to define the access pattern. For **zipf**, this is the `zipf | |
c60ebc45 | 1244 | theta`. For **pareto**, it's the `Pareto power`. Fio includes a test |
f80dba8d MT |
1245 | program, :command:`genzipf`, that can be used visualize what the given input |
1246 | values will yield in terms of hit rates. If you wanted to use **zipf** with | |
1247 | a `theta` of 1.2, you would use ``random_distribution=zipf:1.2`` as the | |
1248 | option. If a non-uniform model is used, fio will disable use of the random | |
b2f4b559 SW |
1249 | map. For the **normal** distribution, a normal (Gaussian) deviation is |
1250 | supplied as a value between 0 and 100. | |
f80dba8d MT |
1251 | |
1252 | For a **zoned** distribution, fio supports specifying percentages of I/O | |
1253 | access that should fall within what range of the file or device. For | |
1254 | example, given a criteria of: | |
1255 | ||
1256 | * 60% of accesses should be to the first 10% | |
1257 | * 30% of accesses should be to the next 20% | |
f334e05e | 1258 | * 8% of accesses should be to the next 30% |
f80dba8d MT |
1259 | * 2% of accesses should be to the next 40% |
1260 | ||
1261 | we can define that through zoning of the random accesses. For the above | |
1262 | example, the user would do:: | |
1263 | ||
1264 | random_distribution=zoned:60/10:30/20:8/30:2/40 | |
1265 | ||
1266 | similarly to how :option:`bssplit` works for setting ranges and percentages | |
1267 | of block sizes. Like :option:`bssplit`, it's possible to specify separate | |
1268 | zones for reads, writes, and trims. If just one set is given, it'll apply to | |
1269 | all of them. | |
1270 | ||
1271 | .. option:: percentage_random=int[,int][,int] | |
1272 | ||
1273 | For a random workload, set how big a percentage should be random. This | |
1274 | defaults to 100%, in which case the workload is fully random. It can be set | |
1275 | from anywhere from 0 to 100. Setting it to 0 would make the workload fully | |
1276 | sequential. Any setting in between will result in a random mix of sequential | |
1277 | and random I/O, at the given percentages. Comma-separated values may be | |
1278 | specified for reads, writes, and trims as described in :option:`blocksize`. | |
1279 | ||
1280 | .. option:: norandommap | |
1281 | ||
1282 | Normally fio will cover every block of the file when doing random I/O. If | |
1283 | this option is given, fio will just get a new random offset without looking | |
1284 | at past I/O history. This means that some blocks may not be read or written, | |
1285 | and that some blocks may be read/written more than once. If this option is | |
1286 | used with :option:`verify` and multiple blocksizes (via :option:`bsrange`), | |
1287 | only intact blocks are verified, i.e., partially-overwritten blocks are | |
1288 | ignored. | |
1289 | ||
1290 | .. option:: softrandommap=bool | |
1291 | ||
1292 | See :option:`norandommap`. If fio runs with the random block map enabled and | |
1293 | it fails to allocate the map, if this option is set it will continue without | |
1294 | a random block map. As coverage will not be as complete as with random maps, | |
1295 | this option is disabled by default. | |
1296 | ||
1297 | .. option:: random_generator=str | |
1298 | ||
1299 | Fio supports the following engines for generating | |
1300 | I/O offsets for random I/O: | |
1301 | ||
1302 | **tausworthe** | |
1303 | Strong 2^88 cycle random number generator | |
1304 | **lfsr** | |
1305 | Linear feedback shift register generator | |
1306 | **tausworthe64** | |
1307 | Strong 64-bit 2^258 cycle random number generator | |
1308 | ||
1309 | **tausworthe** is a strong random number generator, but it requires tracking | |
1310 | on the side if we want to ensure that blocks are only read or written | |
1311 | once. **LFSR** guarantees that we never generate the same offset twice, and | |
1312 | it's also less computationally expensive. It's not a true random generator, | |
1313 | however, though for I/O purposes it's typically good enough. **LFSR** only | |
1314 | works with single block sizes, not with workloads that use multiple block | |
1315 | sizes. If used with such a workload, fio may read or write some blocks | |
1316 | multiple times. The default value is **tausworthe**, unless the required | |
1317 | space exceeds 2^32 blocks. If it does, then **tausworthe64** is | |
1318 | selected automatically. | |
1319 | ||
1320 | ||
1321 | Block size | |
1322 | ~~~~~~~~~~ | |
1323 | ||
1324 | .. option:: blocksize=int[,int][,int], bs=int[,int][,int] | |
1325 | ||
1326 | The block size in bytes used for I/O units. Default: 4096. A single value | |
1327 | applies to reads, writes, and trims. Comma-separated values may be | |
1328 | specified for reads, writes, and trims. A value not terminated in a comma | |
1329 | applies to subsequent types. | |
1330 | ||
1331 | Examples: | |
1332 | ||
1333 | **bs=256k** | |
1334 | means 256k for reads, writes and trims. | |
1335 | ||
1336 | **bs=8k,32k** | |
1337 | means 8k for reads, 32k for writes and trims. | |
1338 | ||
1339 | **bs=8k,32k,** | |
1340 | means 8k for reads, 32k for writes, and default for trims. | |
1341 | ||
1342 | **bs=,8k** | |
1343 | means default for reads, 8k for writes and trims. | |
1344 | ||
1345 | **bs=,8k,** | |
b443ae44 | 1346 | means default for reads, 8k for writes, and default for trims. |
f80dba8d MT |
1347 | |
1348 | .. option:: blocksize_range=irange[,irange][,irange], bsrange=irange[,irange][,irange] | |
1349 | ||
1350 | A range of block sizes in bytes for I/O units. The issued I/O unit will | |
1351 | always be a multiple of the minimum size, unless | |
1352 | :option:`blocksize_unaligned` is set. | |
1353 | ||
1354 | Comma-separated ranges may be specified for reads, writes, and trims as | |
1355 | described in :option:`blocksize`. | |
1356 | ||
1357 | Example: ``bsrange=1k-4k,2k-8k``. | |
1358 | ||
1359 | .. option:: bssplit=str[,str][,str] | |
1360 | ||
1361 | Sometimes you want even finer grained control of the block sizes issued, not | |
1362 | just an even split between them. This option allows you to weight various | |
1363 | block sizes, so that you are able to define a specific amount of block sizes | |
1364 | issued. The format for this option is:: | |
1365 | ||
1366 | bssplit=blocksize/percentage:blocksize/percentage | |
1367 | ||
1368 | for as many block sizes as needed. So if you want to define a workload that | |
1369 | has 50% 64k blocks, 10% 4k blocks, and 40% 32k blocks, you would write:: | |
1370 | ||
1371 | bssplit=4k/10:64k/50:32k/40 | |
1372 | ||
1373 | Ordering does not matter. If the percentage is left blank, fio will fill in | |
1374 | the remaining values evenly. So a bssplit option like this one:: | |
1375 | ||
1376 | bssplit=4k/50:1k/:32k/ | |
1377 | ||
1378 | would have 50% 4k ios, and 25% 1k and 32k ios. The percentages always add up | |
1379 | to 100, if bssplit is given a range that adds up to more, it will error out. | |
1380 | ||
1381 | Comma-separated values may be specified for reads, writes, and trims as | |
1382 | described in :option:`blocksize`. | |
1383 | ||
1384 | If you want a workload that has 50% 2k reads and 50% 4k reads, while having | |
1385 | 90% 4k writes and 10% 8k writes, you would specify:: | |
1386 | ||
1387 | bssplit=2k/50:4k/50,4k/90,8k/10 | |
1388 | ||
1389 | .. option:: blocksize_unaligned, bs_unaligned | |
1390 | ||
1391 | If set, fio will issue I/O units with any size within | |
1392 | :option:`blocksize_range`, not just multiples of the minimum size. This | |
1393 | typically won't work with direct I/O, as that normally requires sector | |
1394 | alignment. | |
1395 | ||
1396 | .. option:: bs_is_seq_rand | |
1397 | ||
1398 | If this option is set, fio will use the normal read,write blocksize settings | |
1399 | as sequential,random blocksize settings instead. Any random read or write | |
1400 | will use the WRITE blocksize settings, and any sequential read or write will | |
1401 | use the READ blocksize settings. | |
1402 | ||
1403 | .. option:: blockalign=int[,int][,int], ba=int[,int][,int] | |
1404 | ||
1405 | Boundary to which fio will align random I/O units. Default: | |
1406 | :option:`blocksize`. Minimum alignment is typically 512b for using direct | |
1407 | I/O, though it usually depends on the hardware block size. This option is | |
1408 | mutually exclusive with using a random map for files, so it will turn off | |
1409 | that option. Comma-separated values may be specified for reads, writes, and | |
1410 | trims as described in :option:`blocksize`. | |
1411 | ||
1412 | ||
1413 | Buffers and memory | |
1414 | ~~~~~~~~~~~~~~~~~~ | |
1415 | ||
1416 | .. option:: zero_buffers | |
1417 | ||
1418 | Initialize buffers with all zeros. Default: fill buffers with random data. | |
1419 | ||
1420 | .. option:: refill_buffers | |
1421 | ||
1422 | If this option is given, fio will refill the I/O buffers on every | |
1423 | submit. The default is to only fill it at init time and reuse that | |
1424 | data. Only makes sense if zero_buffers isn't specified, naturally. If data | |
1425 | verification is enabled, `refill_buffers` is also automatically enabled. | |
1426 | ||
1427 | .. option:: scramble_buffers=bool | |
1428 | ||
1429 | If :option:`refill_buffers` is too costly and the target is using data | |
1430 | deduplication, then setting this option will slightly modify the I/O buffer | |
1431 | contents to defeat normal de-dupe attempts. This is not enough to defeat | |
1432 | more clever block compression attempts, but it will stop naive dedupe of | |
1433 | blocks. Default: true. | |
1434 | ||
1435 | .. option:: buffer_compress_percentage=int | |
1436 | ||
1437 | If this is set, then fio will attempt to provide I/O buffer content (on | |
730bd7d9 | 1438 | WRITEs) that compresses to the specified level. Fio does this by providing a |
22413915 | 1439 | mix of random data and a fixed pattern. The fixed pattern is either zeros, |
f80dba8d MT |
1440 | or the pattern specified by :option:`buffer_pattern`. If the pattern option |
1441 | is used, it might skew the compression ratio slightly. Note that this is per | |
1442 | block size unit, for file/disk wide compression level that matches this | |
1443 | setting, you'll also want to set :option:`refill_buffers`. | |
1444 | ||
1445 | .. option:: buffer_compress_chunk=int | |
1446 | ||
1447 | See :option:`buffer_compress_percentage`. This setting allows fio to manage | |
1448 | how big the ranges of random data and zeroed data is. Without this set, fio | |
1449 | will provide :option:`buffer_compress_percentage` of blocksize random data, | |
1450 | followed by the remaining zeroed. With this set to some chunk size smaller | |
1451 | than the block size, fio can alternate random and zeroed data throughout the | |
1452 | I/O buffer. | |
1453 | ||
1454 | .. option:: buffer_pattern=str | |
1455 | ||
a1554f65 SB |
1456 | If set, fio will fill the I/O buffers with this pattern or with the contents |
1457 | of a file. If not set, the contents of I/O buffers are defined by the other | |
1458 | options related to buffer contents. The setting can be any pattern of bytes, | |
1459 | and can be prefixed with 0x for hex values. It may also be a string, where | |
1460 | the string must then be wrapped with ``""``. Or it may also be a filename, | |
1461 | where the filename must be wrapped with ``''`` in which case the file is | |
1462 | opened and read. Note that not all the file contents will be read if that | |
1463 | would cause the buffers to overflow. So, for example:: | |
1464 | ||
1465 | buffer_pattern='filename' | |
1466 | ||
1467 | or:: | |
f80dba8d MT |
1468 | |
1469 | buffer_pattern="abcd" | |
1470 | ||
1471 | or:: | |
1472 | ||
1473 | buffer_pattern=-12 | |
1474 | ||
1475 | or:: | |
1476 | ||
1477 | buffer_pattern=0xdeadface | |
1478 | ||
1479 | Also you can combine everything together in any order:: | |
1480 | ||
a1554f65 | 1481 | buffer_pattern=0xdeadface"abcd"-12'filename' |
f80dba8d MT |
1482 | |
1483 | .. option:: dedupe_percentage=int | |
1484 | ||
1485 | If set, fio will generate this percentage of identical buffers when | |
1486 | writing. These buffers will be naturally dedupable. The contents of the | |
1487 | buffers depend on what other buffer compression settings have been set. It's | |
1488 | possible to have the individual buffers either fully compressible, or not at | |
1489 | all. This option only controls the distribution of unique buffers. | |
1490 | ||
1491 | .. option:: invalidate=bool | |
1492 | ||
730bd7d9 SW |
1493 | Invalidate the buffer/page cache parts of the files to be used prior to |
1494 | starting I/O if the platform and file type support it. Defaults to true. | |
21c1b29e TK |
1495 | This will be ignored if :option:`pre_read` is also specified for the |
1496 | same job. | |
f80dba8d MT |
1497 | |
1498 | .. option:: sync=bool | |
1499 | ||
1500 | Use synchronous I/O for buffered writes. For the majority of I/O engines, | |
1501 | this means using O_SYNC. Default: false. | |
1502 | ||
1503 | .. option:: iomem=str, mem=str | |
1504 | ||
1505 | Fio can use various types of memory as the I/O unit buffer. The allowed | |
1506 | values are: | |
1507 | ||
1508 | **malloc** | |
1509 | Use memory from :manpage:`malloc(3)` as the buffers. Default memory | |
1510 | type. | |
1511 | ||
1512 | **shm** | |
1513 | Use shared memory as the buffers. Allocated through | |
1514 | :manpage:`shmget(2)`. | |
1515 | ||
1516 | **shmhuge** | |
1517 | Same as shm, but use huge pages as backing. | |
1518 | ||
1519 | **mmap** | |
22413915 | 1520 | Use :manpage:`mmap(2)` to allocate buffers. May either be anonymous memory, or can |
f80dba8d MT |
1521 | be file backed if a filename is given after the option. The format |
1522 | is `mem=mmap:/path/to/file`. | |
1523 | ||
1524 | **mmaphuge** | |
1525 | Use a memory mapped huge file as the buffer backing. Append filename | |
1526 | after mmaphuge, ala `mem=mmaphuge:/hugetlbfs/file`. | |
1527 | ||
1528 | **mmapshared** | |
1529 | Same as mmap, but use a MMAP_SHARED mapping. | |
1530 | ||
03553853 YR |
1531 | **cudamalloc** |
1532 | Use GPU memory as the buffers for GPUDirect RDMA benchmark. | |
1533 | ||
f80dba8d MT |
1534 | The area allocated is a function of the maximum allowed bs size for the job, |
1535 | multiplied by the I/O depth given. Note that for **shmhuge** and | |
1536 | **mmaphuge** to work, the system must have free huge pages allocated. This | |
1537 | can normally be checked and set by reading/writing | |
1538 | :file:`/proc/sys/vm/nr_hugepages` on a Linux system. Fio assumes a huge page | |
1539 | is 4MiB in size. So to calculate the number of huge pages you need for a | |
1540 | given job file, add up the I/O depth of all jobs (normally one unless | |
1541 | :option:`iodepth` is used) and multiply by the maximum bs set. Then divide | |
1542 | that number by the huge page size. You can see the size of the huge pages in | |
1543 | :file:`/proc/meminfo`. If no huge pages are allocated by having a non-zero | |
1544 | number in `nr_hugepages`, using **mmaphuge** or **shmhuge** will fail. Also | |
1545 | see :option:`hugepage-size`. | |
1546 | ||
1547 | **mmaphuge** also needs to have hugetlbfs mounted and the file location | |
1548 | should point there. So if it's mounted in :file:`/huge`, you would use | |
1549 | `mem=mmaphuge:/huge/somefile`. | |
1550 | ||
1551 | .. option:: iomem_align=int | |
1552 | ||
1553 | This indicates the memory alignment of the I/O memory buffers. Note that | |
1554 | the given alignment is applied to the first I/O unit buffer, if using | |
1555 | :option:`iodepth` the alignment of the following buffers are given by the | |
1556 | :option:`bs` used. In other words, if using a :option:`bs` that is a | |
1557 | multiple of the page sized in the system, all buffers will be aligned to | |
1558 | this value. If using a :option:`bs` that is not page aligned, the alignment | |
1559 | of subsequent I/O memory buffers is the sum of the :option:`iomem_align` and | |
1560 | :option:`bs` used. | |
1561 | ||
1562 | .. option:: hugepage-size=int | |
1563 | ||
1564 | Defines the size of a huge page. Must at least be equal to the system | |
1565 | setting, see :file:`/proc/meminfo`. Defaults to 4MiB. Should probably | |
1566 | always be a multiple of megabytes, so using ``hugepage-size=Xm`` is the | |
1567 | preferred way to set this to avoid setting a non-pow-2 bad value. | |
1568 | ||
1569 | .. option:: lockmem=int | |
1570 | ||
1571 | Pin the specified amount of memory with :manpage:`mlock(2)`. Can be used to | |
1572 | simulate a smaller amount of memory. The amount specified is per worker. | |
1573 | ||
1574 | ||
1575 | I/O size | |
1576 | ~~~~~~~~ | |
1577 | ||
1578 | .. option:: size=int | |
1579 | ||
79591fa9 TK |
1580 | The total size of file I/O for each thread of this job. Fio will run until |
1581 | this many bytes has been transferred, unless runtime is limited by other options | |
1582 | (such as :option:`runtime`, for instance, or increased/decreased by :option:`io_size`). | |
1583 | Fio will divide this size between the available files determined by options | |
1584 | such as :option:`nrfiles`, :option:`filename`, unless :option:`filesize` is | |
1585 | specified by the job. If the result of division happens to be 0, the size is | |
c4aa2d08 | 1586 | set to the physical size of the given files or devices if they exist. |
79591fa9 | 1587 | If this option is not specified, fio will use the full size of the given |
f80dba8d MT |
1588 | files or devices. If the files do not exist, size must be given. It is also |
1589 | possible to give size as a percentage between 1 and 100. If ``size=20%`` is | |
1590 | given, fio will use 20% of the full size of the given files or devices. | |
9d25d068 SW |
1591 | Can be combined with :option:`offset` to constrain the start and end range |
1592 | that I/O will be done within. | |
f80dba8d MT |
1593 | |
1594 | .. option:: io_size=int, io_limit=int | |
1595 | ||
1596 | Normally fio operates within the region set by :option:`size`, which means | |
1597 | that the :option:`size` option sets both the region and size of I/O to be | |
1598 | performed. Sometimes that is not what you want. With this option, it is | |
1599 | possible to define just the amount of I/O that fio should do. For instance, | |
1600 | if :option:`size` is set to 20GiB and :option:`io_size` is set to 5GiB, fio | |
1601 | will perform I/O within the first 20GiB but exit when 5GiB have been | |
1602 | done. The opposite is also possible -- if :option:`size` is set to 20GiB, | |
1603 | and :option:`io_size` is set to 40GiB, then fio will do 40GiB of I/O within | |
1604 | the 0..20GiB region. | |
1605 | ||
7fdd97ca | 1606 | .. option:: filesize=irange(int) |
f80dba8d MT |
1607 | |
1608 | Individual file sizes. May be a range, in which case fio will select sizes | |
1609 | for files at random within the given range and limited to :option:`size` in | |
1610 | total (if that is given). If not given, each created file is the same size. | |
79591fa9 TK |
1611 | This option overrides :option:`size` in terms of file size, which means |
1612 | this value is used as a fixed size or possible range of each file. | |
f80dba8d MT |
1613 | |
1614 | .. option:: file_append=bool | |
1615 | ||
1616 | Perform I/O after the end of the file. Normally fio will operate within the | |
1617 | size of a file. If this option is set, then fio will append to the file | |
1618 | instead. This has identical behavior to setting :option:`offset` to the size | |
1619 | of a file. This option is ignored on non-regular files. | |
1620 | ||
1621 | .. option:: fill_device=bool, fill_fs=bool | |
1622 | ||
1623 | Sets size to something really large and waits for ENOSPC (no space left on | |
1624 | device) as the terminating condition. Only makes sense with sequential | |
1625 | write. For a read workload, the mount point will be filled first then I/O | |
1626 | started on the result. This option doesn't make sense if operating on a raw | |
1627 | device node, since the size of that is already known by the file system. | |
1628 | Additionally, writing beyond end-of-device will not return ENOSPC there. | |
1629 | ||
1630 | ||
1631 | I/O engine | |
1632 | ~~~~~~~~~~ | |
1633 | ||
1634 | .. option:: ioengine=str | |
1635 | ||
1636 | Defines how the job issues I/O to the file. The following types are defined: | |
1637 | ||
1638 | **sync** | |
1639 | Basic :manpage:`read(2)` or :manpage:`write(2)` | |
1640 | I/O. :manpage:`lseek(2)` is used to position the I/O location. | |
54227e6b | 1641 | See :option:`fsync` and :option:`fdatasync` for syncing write I/Os. |
f80dba8d MT |
1642 | |
1643 | **psync** | |
1644 | Basic :manpage:`pread(2)` or :manpage:`pwrite(2)` I/O. Default on | |
1645 | all supported operating systems except for Windows. | |
1646 | ||
1647 | **vsync** | |
1648 | Basic :manpage:`readv(2)` or :manpage:`writev(2)` I/O. Will emulate | |
c60ebc45 | 1649 | queuing by coalescing adjacent I/Os into a single submission. |
f80dba8d MT |
1650 | |
1651 | **pvsync** | |
1652 | Basic :manpage:`preadv(2)` or :manpage:`pwritev(2)` I/O. | |
1653 | ||
1654 | **pvsync2** | |
1655 | Basic :manpage:`preadv2(2)` or :manpage:`pwritev2(2)` I/O. | |
1656 | ||
1657 | **libaio** | |
1658 | Linux native asynchronous I/O. Note that Linux may only support | |
22413915 | 1659 | queued behavior with non-buffered I/O (set ``direct=1`` or |
f80dba8d MT |
1660 | ``buffered=0``). |
1661 | This engine defines engine specific options. | |
1662 | ||
1663 | **posixaio** | |
1664 | POSIX asynchronous I/O using :manpage:`aio_read(3)` and | |
1665 | :manpage:`aio_write(3)`. | |
1666 | ||
1667 | **solarisaio** | |
1668 | Solaris native asynchronous I/O. | |
1669 | ||
1670 | **windowsaio** | |
1671 | Windows native asynchronous I/O. Default on Windows. | |
1672 | ||
1673 | **mmap** | |
1674 | File is memory mapped with :manpage:`mmap(2)` and data copied | |
1675 | to/from using :manpage:`memcpy(3)`. | |
1676 | ||
1677 | **splice** | |
1678 | :manpage:`splice(2)` is used to transfer the data and | |
1679 | :manpage:`vmsplice(2)` to transfer data from user space to the | |
1680 | kernel. | |
1681 | ||
1682 | **sg** | |
1683 | SCSI generic sg v3 I/O. May either be synchronous using the SG_IO | |
1684 | ioctl, or if the target is an sg character device we use | |
1685 | :manpage:`read(2)` and :manpage:`write(2)` for asynchronous | |
1686 | I/O. Requires filename option to specify either block or character | |
1687 | devices. | |
1688 | ||
1689 | **null** | |
1690 | Doesn't transfer any data, just pretends to. This is mainly used to | |
1691 | exercise fio itself and for debugging/testing purposes. | |
1692 | ||
1693 | **net** | |
1694 | Transfer over the network to given ``host:port``. Depending on the | |
1695 | :option:`protocol` used, the :option:`hostname`, :option:`port`, | |
1696 | :option:`listen` and :option:`filename` options are used to specify | |
1697 | what sort of connection to make, while the :option:`protocol` option | |
1698 | determines which protocol will be used. This engine defines engine | |
1699 | specific options. | |
1700 | ||
1701 | **netsplice** | |
1702 | Like **net**, but uses :manpage:`splice(2)` and | |
1703 | :manpage:`vmsplice(2)` to map data and send/receive. | |
1704 | This engine defines engine specific options. | |
1705 | ||
1706 | **cpuio** | |
1707 | Doesn't transfer any data, but burns CPU cycles according to the | |
1708 | :option:`cpuload` and :option:`cpuchunks` options. Setting | |
9207a0cb | 1709 | :option:`cpuload`\=85 will cause that job to do nothing but burn 85% |
f80dba8d MT |
1710 | of the CPU. In case of SMP machines, use :option:`numjobs` |
1711 | =<no_of_cpu> to get desired CPU usage, as the cpuload only loads a | |
1712 | single CPU at the desired rate. A job never finishes unless there is | |
1713 | at least one non-cpuio job. | |
1714 | ||
1715 | **guasi** | |
1716 | The GUASI I/O engine is the Generic Userspace Asyncronous Syscall | |
1717 | Interface approach to async I/O. See | |
1718 | ||
1719 | http://www.xmailserver.org/guasi-lib.html | |
1720 | ||
1721 | for more info on GUASI. | |
1722 | ||
1723 | **rdma** | |
1724 | The RDMA I/O engine supports both RDMA memory semantics | |
1725 | (RDMA_WRITE/RDMA_READ) and channel semantics (Send/Recv) for the | |
1726 | InfiniBand, RoCE and iWARP protocols. | |
1727 | ||
1728 | **falloc** | |
1729 | I/O engine that does regular fallocate to simulate data transfer as | |
1730 | fio ioengine. | |
1731 | ||
1732 | DDIR_READ | |
1733 | does fallocate(,mode = FALLOC_FL_KEEP_SIZE,). | |
1734 | ||
1735 | DDIR_WRITE | |
1736 | does fallocate(,mode = 0). | |
1737 | ||
1738 | DDIR_TRIM | |
1739 | does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE). | |
1740 | ||
761cd093 SW |
1741 | **ftruncate** |
1742 | I/O engine that sends :manpage:`ftruncate(2)` operations in response | |
1743 | to write (DDIR_WRITE) events. Each ftruncate issued sets the file's | |
1744 | size to the current block offset. Block size is ignored. | |
1745 | ||
f80dba8d MT |
1746 | **e4defrag** |
1747 | I/O engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate | |
1748 | defragment activity in request to DDIR_WRITE event. | |
1749 | ||
1750 | **rbd** | |
1751 | I/O engine supporting direct access to Ceph Rados Block Devices | |
1752 | (RBD) via librbd without the need to use the kernel rbd driver. This | |
1753 | ioengine defines engine specific options. | |
1754 | ||
1755 | **gfapi** | |
ac8ca2af SW |
1756 | Using GlusterFS libgfapi sync interface to direct access to |
1757 | GlusterFS volumes without having to go through FUSE. This ioengine | |
f80dba8d MT |
1758 | defines engine specific options. |
1759 | ||
1760 | **gfapi_async** | |
ac8ca2af SW |
1761 | Using GlusterFS libgfapi async interface to direct access to |
1762 | GlusterFS volumes without having to go through FUSE. This ioengine | |
f80dba8d MT |
1763 | defines engine specific options. |
1764 | ||
1765 | **libhdfs** | |
1766 | Read and write through Hadoop (HDFS). The :file:`filename` option | |
1767 | is used to specify host,port of the hdfs name-node to connect. This | |
1768 | engine interprets offsets a little differently. In HDFS, files once | |
e25c0c91 SW |
1769 | created cannot be modified so random writes are not possible. To |
1770 | imitate this the libhdfs engine expects a bunch of small files to be | |
1771 | created over HDFS and will randomly pick a file from them | |
1772 | based on the offset generated by fio backend (see the example | |
f80dba8d | 1773 | job file to create such files, use ``rw=write`` option). Please |
e25c0c91 SW |
1774 | note, it may be necessary to set environment variables to work |
1775 | with HDFS/libhdfs properly. Each job uses its own connection to | |
f80dba8d MT |
1776 | HDFS. |
1777 | ||
1778 | **mtd** | |
1779 | Read, write and erase an MTD character device (e.g., | |
1780 | :file:`/dev/mtd0`). Discards are treated as erases. Depending on the | |
1781 | underlying device type, the I/O may have to go in a certain pattern, | |
1782 | e.g., on NAND, writing sequentially to erase blocks and discarding | |
c298ee71 | 1783 | before overwriting. The `trimwrite` mode works well for this |
f80dba8d MT |
1784 | constraint. |
1785 | ||
1786 | **pmemblk** | |
1787 | Read and write using filesystem DAX to a file on a filesystem | |
1788 | mounted with DAX on a persistent memory device through the NVML | |
1789 | libpmemblk library. | |
1790 | ||
1791 | **dev-dax** | |
1792 | Read and write using device DAX to a persistent memory device (e.g., | |
1793 | /dev/dax0.0) through the NVML libpmem library. | |
1794 | ||
1795 | **external** | |
1796 | Prefix to specify loading an external I/O engine object file. Append | |
c60ebc45 | 1797 | the engine filename, e.g. ``ioengine=external:/tmp/foo.o`` to load |
f80dba8d MT |
1798 | ioengine :file:`foo.o` in :file:`/tmp`. |
1799 | ||
1800 | ||
1801 | I/O engine specific parameters | |
1802 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
1803 | ||
1804 | In addition, there are some parameters which are only valid when a specific | |
1805 | ioengine is in use. These are used identically to normal parameters, with the | |
1806 | caveat that when used on the command line, they must come after the | |
1807 | :option:`ioengine` that defines them is selected. | |
1808 | ||
1809 | .. option:: userspace_reap : [libaio] | |
1810 | ||
1811 | Normally, with the libaio engine in use, fio will use the | |
1812 | :manpage:`io_getevents(2)` system call to reap newly returned events. With | |
1813 | this flag turned on, the AIO ring will be read directly from user-space to | |
1814 | reap events. The reaping mode is only enabled when polling for a minimum of | |
c60ebc45 | 1815 | 0 events (e.g. when :option:`iodepth_batch_complete` `=0`). |
f80dba8d | 1816 | |
9d25d068 | 1817 | .. option:: hipri : [pvsync2] |
f80dba8d MT |
1818 | |
1819 | Set RWF_HIPRI on I/O, indicating to the kernel that it's of higher priority | |
1820 | than normal. | |
1821 | ||
1822 | .. option:: cpuload=int : [cpuio] | |
1823 | ||
da19cdb4 TK |
1824 | Attempt to use the specified percentage of CPU cycles. This is a mandatory |
1825 | option when using cpuio I/O engine. | |
f80dba8d MT |
1826 | |
1827 | .. option:: cpuchunks=int : [cpuio] | |
1828 | ||
1829 | Split the load into cycles of the given time. In microseconds. | |
1830 | ||
1831 | .. option:: exit_on_io_done=bool : [cpuio] | |
1832 | ||
1833 | Detect when I/O threads are done, then exit. | |
1834 | ||
1835 | .. option:: hostname=str : [netsplice] [net] | |
1836 | ||
22413915 SW |
1837 | The hostname or IP address to use for TCP or UDP based I/O. If the job is |
1838 | a TCP listener or UDP reader, the hostname is not used and must be omitted | |
f80dba8d MT |
1839 | unless it is a valid UDP multicast address. |
1840 | ||
1841 | .. option:: namenode=str : [libhdfs] | |
1842 | ||
22413915 | 1843 | The hostname or IP address of a HDFS cluster namenode to contact. |
f80dba8d MT |
1844 | |
1845 | .. option:: port=int | |
1846 | ||
1847 | [netsplice], [net] | |
1848 | ||
1849 | The TCP or UDP port to bind to or connect to. If this is used with | |
1850 | :option:`numjobs` to spawn multiple instances of the same job type, then | |
1851 | this will be the starting port number since fio will use a range of | |
1852 | ports. | |
1853 | ||
1854 | [libhdfs] | |
1855 | ||
1856 | the listening port of the HFDS cluster namenode. | |
1857 | ||
1858 | .. option:: interface=str : [netsplice] [net] | |
1859 | ||
1860 | The IP address of the network interface used to send or receive UDP | |
1861 | multicast. | |
1862 | ||
1863 | .. option:: ttl=int : [netsplice] [net] | |
1864 | ||
1865 | Time-to-live value for outgoing UDP multicast packets. Default: 1. | |
1866 | ||
1867 | .. option:: nodelay=bool : [netsplice] [net] | |
1868 | ||
1869 | Set TCP_NODELAY on TCP connections. | |
1870 | ||
1871 | .. option:: protocol=str : [netsplice] [net] | |
1872 | ||
1873 | .. option:: proto=str : [netsplice] [net] | |
1874 | ||
1875 | The network protocol to use. Accepted values are: | |
1876 | ||
1877 | **tcp** | |
1878 | Transmission control protocol. | |
1879 | **tcpv6** | |
1880 | Transmission control protocol V6. | |
1881 | **udp** | |
1882 | User datagram protocol. | |
1883 | **udpv6** | |
1884 | User datagram protocol V6. | |
1885 | **unix** | |
1886 | UNIX domain socket. | |
1887 | ||
1888 | When the protocol is TCP or UDP, the port must also be given, as well as the | |
1889 | hostname if the job is a TCP listener or UDP reader. For unix sockets, the | |
1890 | normal filename option should be used and the port is invalid. | |
1891 | ||
e9184ec1 | 1892 | .. option:: listen : [netsplice] [net] |
f80dba8d MT |
1893 | |
1894 | For TCP network connections, tell fio to listen for incoming connections | |
1895 | rather than initiating an outgoing connection. The :option:`hostname` must | |
1896 | be omitted if this option is used. | |
1897 | ||
e9184ec1 | 1898 | .. option:: pingpong : [netsplice] [net] |
f80dba8d MT |
1899 | |
1900 | Normally a network writer will just continue writing data, and a network | |
1901 | reader will just consume packages. If ``pingpong=1`` is set, a writer will | |
1902 | send its normal payload to the reader, then wait for the reader to send the | |
1903 | same payload back. This allows fio to measure network latencies. The | |
1904 | submission and completion latencies then measure local time spent sending or | |
1905 | receiving, and the completion latency measures how long it took for the | |
1906 | other end to receive and send back. For UDP multicast traffic | |
1907 | ``pingpong=1`` should only be set for a single reader when multiple readers | |
1908 | are listening to the same address. | |
1909 | ||
e9184ec1 | 1910 | .. option:: window_size : [netsplice] [net] |
f80dba8d MT |
1911 | |
1912 | Set the desired socket buffer size for the connection. | |
1913 | ||
e9184ec1 | 1914 | .. option:: mss : [netsplice] [net] |
f80dba8d MT |
1915 | |
1916 | Set the TCP maximum segment size (TCP_MAXSEG). | |
1917 | ||
1918 | .. option:: donorname=str : [e4defrag] | |
1919 | ||
730bd7d9 | 1920 | File will be used as a block donor (swap extents between files). |
f80dba8d MT |
1921 | |
1922 | .. option:: inplace=int : [e4defrag] | |
1923 | ||
1924 | Configure donor file blocks allocation strategy: | |
1925 | ||
1926 | **0** | |
1927 | Default. Preallocate donor's file on init. | |
1928 | **1** | |
1929 | Allocate space immediately inside defragment event, and free right | |
1930 | after event. | |
1931 | ||
1932 | .. option:: clustername=str : [rbd] | |
1933 | ||
1934 | Specifies the name of the Ceph cluster. | |
1935 | ||
1936 | .. option:: rbdname=str : [rbd] | |
1937 | ||
1938 | Specifies the name of the RBD. | |
1939 | ||
1940 | .. option:: pool=str : [rbd] | |
1941 | ||
1942 | Specifies the name of the Ceph pool containing RBD. | |
1943 | ||
1944 | .. option:: clientname=str : [rbd] | |
1945 | ||
1946 | Specifies the username (without the 'client.' prefix) used to access the | |
1947 | Ceph cluster. If the *clustername* is specified, the *clientname* shall be | |
1948 | the full *type.id* string. If no type. prefix is given, fio will add | |
1949 | 'client.' by default. | |
1950 | ||
1951 | .. option:: skip_bad=bool : [mtd] | |
1952 | ||
1953 | Skip operations against known bad blocks. | |
1954 | ||
1955 | .. option:: hdfsdirectory : [libhdfs] | |
1956 | ||
1957 | libhdfs will create chunk in this HDFS directory. | |
1958 | ||
1959 | .. option:: chunk_size : [libhdfs] | |
1960 | ||
1961 | the size of the chunk to use for each file. | |
1962 | ||
1963 | ||
1964 | I/O depth | |
1965 | ~~~~~~~~~ | |
1966 | ||
1967 | .. option:: iodepth=int | |
1968 | ||
1969 | Number of I/O units to keep in flight against the file. Note that | |
1970 | increasing *iodepth* beyond 1 will not affect synchronous ioengines (except | |
c60ebc45 | 1971 | for small degrees when :option:`verify_async` is in use). Even async |
f80dba8d MT |
1972 | engines may impose OS restrictions causing the desired depth not to be |
1973 | achieved. This may happen on Linux when using libaio and not setting | |
9207a0cb | 1974 | :option:`direct`\=1, since buffered I/O is not async on that OS. Keep an |
f80dba8d MT |
1975 | eye on the I/O depth distribution in the fio output to verify that the |
1976 | achieved depth is as expected. Default: 1. | |
1977 | ||
1978 | .. option:: iodepth_batch_submit=int, iodepth_batch=int | |
1979 | ||
1980 | This defines how many pieces of I/O to submit at once. It defaults to 1 | |
1981 | which means that we submit each I/O as soon as it is available, but can be | |
1982 | raised to submit bigger batches of I/O at the time. If it is set to 0 the | |
1983 | :option:`iodepth` value will be used. | |
1984 | ||
1985 | .. option:: iodepth_batch_complete_min=int, iodepth_batch_complete=int | |
1986 | ||
1987 | This defines how many pieces of I/O to retrieve at once. It defaults to 1 | |
1988 | which means that we'll ask for a minimum of 1 I/O in the retrieval process | |
1989 | from the kernel. The I/O retrieval will go on until we hit the limit set by | |
1990 | :option:`iodepth_low`. If this variable is set to 0, then fio will always | |
1991 | check for completed events before queuing more I/O. This helps reduce I/O | |
1992 | latency, at the cost of more retrieval system calls. | |
1993 | ||
1994 | .. option:: iodepth_batch_complete_max=int | |
1995 | ||
1996 | This defines maximum pieces of I/O to retrieve at once. This variable should | |
9207a0cb | 1997 | be used along with :option:`iodepth_batch_complete_min`\=int variable, |
f80dba8d | 1998 | specifying the range of min and max amount of I/O which should be |
730bd7d9 | 1999 | retrieved. By default it is equal to the :option:`iodepth_batch_complete_min` |
f80dba8d MT |
2000 | value. |
2001 | ||
2002 | Example #1:: | |
2003 | ||
2004 | iodepth_batch_complete_min=1 | |
2005 | iodepth_batch_complete_max=<iodepth> | |
2006 | ||
2007 | which means that we will retrieve at least 1 I/O and up to the whole | |
2008 | submitted queue depth. If none of I/O has been completed yet, we will wait. | |
2009 | ||
2010 | Example #2:: | |
2011 | ||
2012 | iodepth_batch_complete_min=0 | |
2013 | iodepth_batch_complete_max=<iodepth> | |
2014 | ||
2015 | which means that we can retrieve up to the whole submitted queue depth, but | |
2016 | if none of I/O has been completed yet, we will NOT wait and immediately exit | |
2017 | the system call. In this example we simply do polling. | |
2018 | ||
2019 | .. option:: iodepth_low=int | |
2020 | ||
2021 | The low water mark indicating when to start filling the queue | |
2022 | again. Defaults to the same as :option:`iodepth`, meaning that fio will | |
2023 | attempt to keep the queue full at all times. If :option:`iodepth` is set to | |
c60ebc45 | 2024 | e.g. 16 and *iodepth_low* is set to 4, then after fio has filled the queue of |
f80dba8d MT |
2025 | 16 requests, it will let the depth drain down to 4 before starting to fill |
2026 | it again. | |
2027 | ||
2028 | .. option:: io_submit_mode=str | |
2029 | ||
2030 | This option controls how fio submits the I/O to the I/O engine. The default | |
2031 | is `inline`, which means that the fio job threads submit and reap I/O | |
2032 | directly. If set to `offload`, the job threads will offload I/O submission | |
2033 | to a dedicated pool of I/O threads. This requires some coordination and thus | |
2034 | has a bit of extra overhead, especially for lower queue depth I/O where it | |
2035 | can increase latencies. The benefit is that fio can manage submission rates | |
2036 | independently of the device completion rates. This avoids skewed latency | |
730bd7d9 | 2037 | reporting if I/O gets backed up on the device side (the coordinated omission |
f80dba8d MT |
2038 | problem). |
2039 | ||
2040 | ||
2041 | I/O rate | |
2042 | ~~~~~~~~ | |
2043 | ||
a881438b | 2044 | .. option:: thinktime=time |
f80dba8d | 2045 | |
f75ede1d SW |
2046 | Stall the job for the specified period of time after an I/O has completed before issuing the |
2047 | next. May be used to simulate processing being done by an application. | |
947e0fe0 | 2048 | When the unit is omitted, the value is interpreted in microseconds. See |
f80dba8d MT |
2049 | :option:`thinktime_blocks` and :option:`thinktime_spin`. |
2050 | ||
a881438b | 2051 | .. option:: thinktime_spin=time |
f80dba8d MT |
2052 | |
2053 | Only valid if :option:`thinktime` is set - pretend to spend CPU time doing | |
2054 | something with the data received, before falling back to sleeping for the | |
f75ede1d | 2055 | rest of the period specified by :option:`thinktime`. When the unit is |
947e0fe0 | 2056 | omitted, the value is interpreted in microseconds. |
f80dba8d MT |
2057 | |
2058 | .. option:: thinktime_blocks=int | |
2059 | ||
2060 | Only valid if :option:`thinktime` is set - control how many blocks to issue, | |
2061 | before waiting `thinktime` usecs. If not set, defaults to 1 which will make | |
2062 | fio wait `thinktime` usecs after every block. This effectively makes any | |
2063 | queue depth setting redundant, since no more than 1 I/O will be queued | |
2064 | before we have to complete it and do our thinktime. In other words, this | |
2065 | setting effectively caps the queue depth if the latter is larger. | |
71bfa161 | 2066 | |
f80dba8d | 2067 | .. option:: rate=int[,int][,int] |
71bfa161 | 2068 | |
f80dba8d MT |
2069 | Cap the bandwidth used by this job. The number is in bytes/sec, the normal |
2070 | suffix rules apply. Comma-separated values may be specified for reads, | |
2071 | writes, and trims as described in :option:`blocksize`. | |
71bfa161 | 2072 | |
b25b3464 SW |
2073 | For example, using `rate=1m,500k` would limit reads to 1MiB/sec and writes to |
2074 | 500KiB/sec. Capping only reads or writes can be done with `rate=,500k` or | |
2075 | `rate=500k,` where the former will only limit writes (to 500KiB/sec) and the | |
2076 | latter will only limit reads. | |
2077 | ||
f80dba8d | 2078 | .. option:: rate_min=int[,int][,int] |
71bfa161 | 2079 | |
f80dba8d MT |
2080 | Tell fio to do whatever it can to maintain at least this bandwidth. Failing |
2081 | to meet this requirement will cause the job to exit. Comma-separated values | |
2082 | may be specified for reads, writes, and trims as described in | |
2083 | :option:`blocksize`. | |
71bfa161 | 2084 | |
f80dba8d | 2085 | .. option:: rate_iops=int[,int][,int] |
71bfa161 | 2086 | |
f80dba8d MT |
2087 | Cap the bandwidth to this number of IOPS. Basically the same as |
2088 | :option:`rate`, just specified independently of bandwidth. If the job is | |
2089 | given a block size range instead of a fixed value, the smallest block size | |
2090 | is used as the metric. Comma-separated values may be specified for reads, | |
2091 | writes, and trims as described in :option:`blocksize`. | |
71bfa161 | 2092 | |
f80dba8d | 2093 | .. option:: rate_iops_min=int[,int][,int] |
71bfa161 | 2094 | |
f80dba8d MT |
2095 | If fio doesn't meet this rate of I/O, it will cause the job to exit. |
2096 | Comma-separated values may be specified for reads, writes, and trims as | |
2097 | described in :option:`blocksize`. | |
71bfa161 | 2098 | |
f80dba8d | 2099 | .. option:: rate_process=str |
66c098b8 | 2100 | |
f80dba8d MT |
2101 | This option controls how fio manages rated I/O submissions. The default is |
2102 | `linear`, which submits I/O in a linear fashion with fixed delays between | |
c60ebc45 | 2103 | I/Os that gets adjusted based on I/O completion rates. If this is set to |
f80dba8d MT |
2104 | `poisson`, fio will submit I/O based on a more real world random request |
2105 | flow, known as the Poisson process | |
2106 | (https://en.wikipedia.org/wiki/Poisson_point_process). The lambda will be | |
2107 | 10^6 / IOPS for the given workload. | |
71bfa161 JA |
2108 | |
2109 | ||
f80dba8d MT |
2110 | I/O latency |
2111 | ~~~~~~~~~~~ | |
71bfa161 | 2112 | |
a881438b | 2113 | .. option:: latency_target=time |
71bfa161 | 2114 | |
f80dba8d | 2115 | If set, fio will attempt to find the max performance point that the given |
f75ede1d | 2116 | workload will run at while maintaining a latency below this target. When |
947e0fe0 | 2117 | the unit is omitted, the value is interpreted in microseconds. See |
f75ede1d | 2118 | :option:`latency_window` and :option:`latency_percentile`. |
71bfa161 | 2119 | |
a881438b | 2120 | .. option:: latency_window=time |
71bfa161 | 2121 | |
f80dba8d | 2122 | Used with :option:`latency_target` to specify the sample window that the job |
f75ede1d | 2123 | is run at varying queue depths to test the performance. When the unit is |
947e0fe0 | 2124 | omitted, the value is interpreted in microseconds. |
b4692828 | 2125 | |
f80dba8d | 2126 | .. option:: latency_percentile=float |
71bfa161 | 2127 | |
c60ebc45 | 2128 | The percentage of I/Os that must fall within the criteria specified by |
f80dba8d | 2129 | :option:`latency_target` and :option:`latency_window`. If not set, this |
c60ebc45 | 2130 | defaults to 100.0, meaning that all I/Os must be equal or below to the value |
f80dba8d | 2131 | set by :option:`latency_target`. |
71bfa161 | 2132 | |
a881438b | 2133 | .. option:: max_latency=time |
71bfa161 | 2134 | |
f75ede1d | 2135 | If set, fio will exit the job with an ETIMEDOUT error if it exceeds this |
947e0fe0 | 2136 | maximum latency. When the unit is omitted, the value is interpreted in |
f75ede1d | 2137 | microseconds. |
71bfa161 | 2138 | |
f80dba8d | 2139 | .. option:: rate_cycle=int |
71bfa161 | 2140 | |
f80dba8d | 2141 | Average bandwidth for :option:`rate` and :option:`rate_min` over this number |
a47b697c | 2142 | of milliseconds. Defaults to 1000. |
71bfa161 | 2143 | |
71bfa161 | 2144 | |
f80dba8d MT |
2145 | I/O replay |
2146 | ~~~~~~~~~~ | |
71bfa161 | 2147 | |
f80dba8d | 2148 | .. option:: write_iolog=str |
c2b1e753 | 2149 | |
f80dba8d MT |
2150 | Write the issued I/O patterns to the specified file. See |
2151 | :option:`read_iolog`. Specify a separate file for each job, otherwise the | |
2152 | iologs will be interspersed and the file may be corrupt. | |
c2b1e753 | 2153 | |
f80dba8d | 2154 | .. option:: read_iolog=str |
71bfa161 | 2155 | |
22413915 | 2156 | Open an iolog with the specified filename and replay the I/O patterns it |
f80dba8d MT |
2157 | contains. This can be used to store a workload and replay it sometime |
2158 | later. The iolog given may also be a blktrace binary file, which allows fio | |
2159 | to replay a workload captured by :command:`blktrace`. See | |
2160 | :manpage:`blktrace(8)` for how to capture such logging data. For blktrace | |
2161 | replay, the file needs to be turned into a blkparse binary data file first | |
2162 | (``blkparse <device> -o /dev/null -d file_for_fio.bin``). | |
71bfa161 | 2163 | |
f80dba8d | 2164 | .. option:: replay_no_stall=int |
71bfa161 | 2165 | |
f80dba8d | 2166 | When replaying I/O with :option:`read_iolog` the default behavior is to |
22413915 | 2167 | attempt to respect the timestamps within the log and replay them with the |
f80dba8d MT |
2168 | appropriate delay between IOPS. By setting this variable fio will not |
2169 | respect the timestamps and attempt to replay them as fast as possible while | |
2170 | still respecting ordering. The result is the same I/O pattern to a given | |
2171 | device, but different timings. | |
71bfa161 | 2172 | |
f80dba8d | 2173 | .. option:: replay_redirect=str |
b4692828 | 2174 | |
f80dba8d MT |
2175 | While replaying I/O patterns using :option:`read_iolog` the default behavior |
2176 | is to replay the IOPS onto the major/minor device that each IOP was recorded | |
2177 | from. This is sometimes undesirable because on a different machine those | |
2178 | major/minor numbers can map to a different device. Changing hardware on the | |
2179 | same system can also result in a different major/minor mapping. | |
730bd7d9 | 2180 | ``replay_redirect`` causes all I/Os to be replayed onto the single specified |
f80dba8d | 2181 | device regardless of the device it was recorded |
9207a0cb | 2182 | from. i.e. :option:`replay_redirect`\= :file:`/dev/sdc` would cause all I/O |
f80dba8d MT |
2183 | in the blktrace or iolog to be replayed onto :file:`/dev/sdc`. This means |
2184 | multiple devices will be replayed onto a single device, if the trace | |
2185 | contains multiple devices. If you want multiple devices to be replayed | |
2186 | concurrently to multiple redirected devices you must blkparse your trace | |
2187 | into separate traces and replay them with independent fio invocations. | |
2188 | Unfortunately this also breaks the strict time ordering between multiple | |
2189 | device accesses. | |
71bfa161 | 2190 | |
f80dba8d | 2191 | .. option:: replay_align=int |
74929ac2 | 2192 | |
f80dba8d MT |
2193 | Force alignment of I/O offsets and lengths in a trace to this power of 2 |
2194 | value. | |
3c54bc46 | 2195 | |
f80dba8d | 2196 | .. option:: replay_scale=int |
3c54bc46 | 2197 | |
f80dba8d | 2198 | Scale sector offsets down by this factor when replaying traces. |
3c54bc46 | 2199 | |
3c54bc46 | 2200 | |
f80dba8d MT |
2201 | Threads, processes and job synchronization |
2202 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
3c54bc46 | 2203 | |
f80dba8d | 2204 | .. option:: thread |
3c54bc46 | 2205 | |
730bd7d9 SW |
2206 | Fio defaults to creating jobs by using fork, however if this option is |
2207 | given, fio will create jobs by using POSIX Threads' function | |
2208 | :manpage:`pthread_create(3)` to create threads instead. | |
71bfa161 | 2209 | |
f80dba8d | 2210 | .. option:: wait_for=str |
74929ac2 | 2211 | |
730bd7d9 SW |
2212 | If set, the current job won't be started until all workers of the specified |
2213 | waitee job are done. | |
74929ac2 | 2214 | |
f80dba8d MT |
2215 | ``wait_for`` operates on the job name basis, so there are a few |
2216 | limitations. First, the waitee must be defined prior to the waiter job | |
2217 | (meaning no forward references). Second, if a job is being referenced as a | |
2218 | waitee, it must have a unique name (no duplicate waitees). | |
74929ac2 | 2219 | |
f80dba8d | 2220 | .. option:: nice=int |
892a6ffc | 2221 | |
f80dba8d | 2222 | Run the job with the given nice value. See man :manpage:`nice(2)`. |
892a6ffc | 2223 | |
f80dba8d MT |
2224 | On Windows, values less than -15 set the process class to "High"; -1 through |
2225 | -15 set "Above Normal"; 1 through 15 "Below Normal"; and above 15 "Idle" | |
2226 | priority class. | |
74929ac2 | 2227 | |
f80dba8d | 2228 | .. option:: prio=int |
71bfa161 | 2229 | |
f80dba8d MT |
2230 | Set the I/O priority value of this job. Linux limits us to a positive value |
2231 | between 0 and 7, with 0 being the highest. See man | |
2232 | :manpage:`ionice(1)`. Refer to an appropriate manpage for other operating | |
2233 | systems since meaning of priority may differ. | |
71bfa161 | 2234 | |
f80dba8d | 2235 | .. option:: prioclass=int |
d59aa780 | 2236 | |
f80dba8d | 2237 | Set the I/O priority class. See man :manpage:`ionice(1)`. |
d59aa780 | 2238 | |
f80dba8d | 2239 | .. option:: cpumask=int |
71bfa161 | 2240 | |
22413915 SW |
2241 | Set the CPU affinity of this job. The parameter given is a bit mask of |
2242 | allowed CPUs the job may run on. So if you want the allowed CPUs to be 1 | |
f80dba8d MT |
2243 | and 5, you would pass the decimal value of (1 << 1 | 1 << 5), or 34. See man |
2244 | :manpage:`sched_setaffinity(2)`. This may not work on all supported | |
2245 | operating systems or kernel versions. This option doesn't work well for a | |
2246 | higher CPU count than what you can store in an integer mask, so it can only | |
2247 | control cpus 1-32. For boxes with larger CPU counts, use | |
2248 | :option:`cpus_allowed`. | |
6d500c2e | 2249 | |
f80dba8d | 2250 | .. option:: cpus_allowed=str |
6d500c2e | 2251 | |
730bd7d9 SW |
2252 | Controls the same options as :option:`cpumask`, but accepts a textual |
2253 | specification of the permitted CPUs instead. So to use CPUs 1 and 5 you | |
2254 | would specify ``cpus_allowed=1,5``. This option also allows a range of CPUs | |
2255 | to be specified -- say you wanted a binding to CPUs 1, 5, and 8 to 15, you | |
2256 | would set ``cpus_allowed=1,5,8-15``. | |
6d500c2e | 2257 | |
f80dba8d | 2258 | .. option:: cpus_allowed_policy=str |
6d500c2e | 2259 | |
f80dba8d | 2260 | Set the policy of how fio distributes the CPUs specified by |
730bd7d9 | 2261 | :option:`cpus_allowed` or :option:`cpumask`. Two policies are supported: |
6d500c2e | 2262 | |
f80dba8d MT |
2263 | **shared** |
2264 | All jobs will share the CPU set specified. | |
2265 | **split** | |
2266 | Each job will get a unique CPU from the CPU set. | |
6d500c2e | 2267 | |
22413915 | 2268 | **shared** is the default behavior, if the option isn't specified. If |
f80dba8d MT |
2269 | **split** is specified, then fio will will assign one cpu per job. If not |
2270 | enough CPUs are given for the jobs listed, then fio will roundrobin the CPUs | |
2271 | in the set. | |
6d500c2e | 2272 | |
f80dba8d | 2273 | .. option:: numa_cpu_nodes=str |
6d500c2e | 2274 | |
f80dba8d MT |
2275 | Set this job running on specified NUMA nodes' CPUs. The arguments allow |
2276 | comma delimited list of cpu numbers, A-B ranges, or `all`. Note, to enable | |
ac8ca2af | 2277 | NUMA options support, fio must be built on a system with libnuma-dev(el) |
f80dba8d | 2278 | installed. |
61b9861d | 2279 | |
f80dba8d | 2280 | .. option:: numa_mem_policy=str |
61b9861d | 2281 | |
f80dba8d MT |
2282 | Set this job's memory policy and corresponding NUMA nodes. Format of the |
2283 | arguments:: | |
5c94b008 | 2284 | |
f80dba8d | 2285 | <mode>[:<nodelist>] |
ce35b1ec | 2286 | |
730bd7d9 SW |
2287 | ``mode`` is one of the following memory poicies: ``default``, ``prefer``, |
2288 | ``bind``, ``interleave`` or ``local``. For ``default`` and ``local`` memory | |
2289 | policies, no node needs to be specified. For ``prefer``, only one node is | |
2290 | allowed. For ``bind`` and ``interleave`` the ``nodelist`` may be as | |
2291 | follows: a comma delimited list of numbers, A-B ranges, or `all`. | |
71bfa161 | 2292 | |
f80dba8d | 2293 | .. option:: cgroup=str |
390b1537 | 2294 | |
f80dba8d MT |
2295 | Add job to this control group. If it doesn't exist, it will be created. The |
2296 | system must have a mounted cgroup blkio mount point for this to work. If | |
2297 | your system doesn't have it mounted, you can do so with:: | |
5af1c6f3 | 2298 | |
f80dba8d | 2299 | # mount -t cgroup -o blkio none /cgroup |
5af1c6f3 | 2300 | |
f80dba8d | 2301 | .. option:: cgroup_weight=int |
5af1c6f3 | 2302 | |
f80dba8d MT |
2303 | Set the weight of the cgroup to this value. See the documentation that comes |
2304 | with the kernel, allowed values are in the range of 100..1000. | |
a086c257 | 2305 | |
f80dba8d | 2306 | .. option:: cgroup_nodelete=bool |
8c07860d | 2307 | |
f80dba8d MT |
2308 | Normally fio will delete the cgroups it has created after the job |
2309 | completion. To override this behavior and to leave cgroups around after the | |
2310 | job completion, set ``cgroup_nodelete=1``. This can be useful if one wants | |
2311 | to inspect various cgroup files after job completion. Default: false. | |
8c07860d | 2312 | |
f80dba8d | 2313 | .. option:: flow_id=int |
8c07860d | 2314 | |
f80dba8d MT |
2315 | The ID of the flow. If not specified, it defaults to being a global |
2316 | flow. See :option:`flow`. | |
1907dbc6 | 2317 | |
f80dba8d | 2318 | .. option:: flow=int |
71bfa161 | 2319 | |
f80dba8d MT |
2320 | Weight in token-based flow control. If this value is used, then there is a |
2321 | 'flow counter' which is used to regulate the proportion of activity between | |
2322 | two or more jobs. Fio attempts to keep this flow counter near zero. The | |
2323 | ``flow`` parameter stands for how much should be added or subtracted to the | |
2324 | flow counter on each iteration of the main I/O loop. That is, if one job has | |
2325 | ``flow=8`` and another job has ``flow=-1``, then there will be a roughly 1:8 | |
2326 | ratio in how much one runs vs the other. | |
71bfa161 | 2327 | |
f80dba8d | 2328 | .. option:: flow_watermark=int |
a31041ea | 2329 | |
f80dba8d MT |
2330 | The maximum value that the absolute value of the flow counter is allowed to |
2331 | reach before the job must wait for a lower value of the counter. | |
82407585 | 2332 | |
f80dba8d | 2333 | .. option:: flow_sleep=int |
82407585 | 2334 | |
f80dba8d MT |
2335 | The period of time, in microseconds, to wait after the flow watermark has |
2336 | been exceeded before retrying operations. | |
82407585 | 2337 | |
f80dba8d | 2338 | .. option:: stonewall, wait_for_previous |
82407585 | 2339 | |
f80dba8d MT |
2340 | Wait for preceding jobs in the job file to exit, before starting this |
2341 | one. Can be used to insert serialization points in the job file. A stone | |
2342 | wall also implies starting a new reporting group, see | |
2343 | :option:`group_reporting`. | |
2344 | ||
2345 | .. option:: exitall | |
2346 | ||
730bd7d9 SW |
2347 | By default, fio will continue running all other jobs when one job finishes |
2348 | but sometimes this is not the desired action. Setting ``exitall`` will | |
2349 | instead make fio terminate all other jobs when one job finishes. | |
f80dba8d MT |
2350 | |
2351 | .. option:: exec_prerun=str | |
2352 | ||
2353 | Before running this job, issue the command specified through | |
2354 | :manpage:`system(3)`. Output is redirected in a file called | |
2355 | :file:`jobname.prerun.txt`. | |
2356 | ||
2357 | .. option:: exec_postrun=str | |
2358 | ||
2359 | After the job completes, issue the command specified though | |
2360 | :manpage:`system(3)`. Output is redirected in a file called | |
2361 | :file:`jobname.postrun.txt`. | |
2362 | ||
2363 | .. option:: uid=int | |
2364 | ||
2365 | Instead of running as the invoking user, set the user ID to this value | |
2366 | before the thread/process does any work. | |
2367 | ||
2368 | .. option:: gid=int | |
2369 | ||
2370 | Set group ID, see :option:`uid`. | |
2371 | ||
2372 | ||
2373 | Verification | |
2374 | ~~~~~~~~~~~~ | |
2375 | ||
2376 | .. option:: verify_only | |
2377 | ||
2378 | Do not perform specified workload, only verify data still matches previous | |
2379 | invocation of this workload. This option allows one to check data multiple | |
2380 | times at a later date without overwriting it. This option makes sense only | |
2381 | for workloads that write data, and does not support workloads with the | |
2382 | :option:`time_based` option set. | |
2383 | ||
2384 | .. option:: do_verify=bool | |
2385 | ||
2386 | Run the verify phase after a write phase. Only valid if :option:`verify` is | |
2387 | set. Default: true. | |
2388 | ||
2389 | .. option:: verify=str | |
2390 | ||
2391 | If writing to a file, fio can verify the file contents after each iteration | |
2392 | of the job. Each verification method also implies verification of special | |
2393 | header, which is written to the beginning of each block. This header also | |
2394 | includes meta information, like offset of the block, block number, timestamp | |
2395 | when block was written, etc. :option:`verify` can be combined with | |
2396 | :option:`verify_pattern` option. The allowed values are: | |
2397 | ||
2398 | **md5** | |
2399 | Use an md5 sum of the data area and store it in the header of | |
2400 | each block. | |
2401 | ||
2402 | **crc64** | |
2403 | Use an experimental crc64 sum of the data area and store it in the | |
2404 | header of each block. | |
2405 | ||
2406 | **crc32c** | |
a5896300 SW |
2407 | Use a crc32c sum of the data area and store it in the header of |
2408 | each block. This will automatically use hardware acceleration | |
2409 | (e.g. SSE4.2 on an x86 or CRC crypto extensions on ARM64) but will | |
2410 | fall back to software crc32c if none is found. Generally the | |
2411 | fatest checksum fio supports when hardware accelerated. | |
f80dba8d MT |
2412 | |
2413 | **crc32c-intel** | |
a5896300 | 2414 | Synonym for crc32c. |
f80dba8d MT |
2415 | |
2416 | **crc32** | |
2417 | Use a crc32 sum of the data area and store it in the header of each | |
2418 | block. | |
2419 | ||
2420 | **crc16** | |
2421 | Use a crc16 sum of the data area and store it in the header of each | |
2422 | block. | |
2423 | ||
2424 | **crc7** | |
2425 | Use a crc7 sum of the data area and store it in the header of each | |
2426 | block. | |
2427 | ||
2428 | **xxhash** | |
2429 | Use xxhash as the checksum function. Generally the fastest software | |
2430 | checksum that fio supports. | |
2431 | ||
2432 | **sha512** | |
2433 | Use sha512 as the checksum function. | |
2434 | ||
2435 | **sha256** | |
2436 | Use sha256 as the checksum function. | |
2437 | ||
2438 | **sha1** | |
2439 | Use optimized sha1 as the checksum function. | |
82407585 | 2440 | |
ae3a5acc JA |
2441 | **sha3-224** |
2442 | Use optimized sha3-224 as the checksum function. | |
2443 | ||
2444 | **sha3-256** | |
2445 | Use optimized sha3-256 as the checksum function. | |
2446 | ||
2447 | **sha3-384** | |
2448 | Use optimized sha3-384 as the checksum function. | |
2449 | ||
2450 | **sha3-512** | |
2451 | Use optimized sha3-512 as the checksum function. | |
2452 | ||
f80dba8d MT |
2453 | **meta** |
2454 | This option is deprecated, since now meta information is included in | |
2455 | generic verification header and meta verification happens by | |
2456 | default. For detailed information see the description of the | |
2457 | :option:`verify` setting. This option is kept because of | |
2458 | compatibility's sake with old configurations. Do not use it. | |
2459 | ||
2460 | **pattern** | |
2461 | Verify a strict pattern. Normally fio includes a header with some | |
2462 | basic information and checksumming, but if this option is set, only | |
2463 | the specific pattern set with :option:`verify_pattern` is verified. | |
2464 | ||
2465 | **null** | |
2466 | Only pretend to verify. Useful for testing internals with | |
9207a0cb | 2467 | :option:`ioengine`\=null, not for much else. |
f80dba8d MT |
2468 | |
2469 | This option can be used for repeated burn-in tests of a system to make sure | |
2470 | that the written data is also correctly read back. If the data direction | |
2471 | given is a read or random read, fio will assume that it should verify a | |
2472 | previously written file. If the data direction includes any form of write, | |
2473 | the verify will be of the newly written data. | |
2474 | ||
2475 | .. option:: verifysort=bool | |
2476 | ||
2477 | If true, fio will sort written verify blocks when it deems it faster to read | |
2478 | them back in a sorted manner. This is often the case when overwriting an | |
2479 | existing file, since the blocks are already laid out in the file system. You | |
2480 | can ignore this option unless doing huge amounts of really fast I/O where | |
2481 | the red-black tree sorting CPU time becomes significant. Default: true. | |
2482 | ||
2483 | .. option:: verifysort_nr=int | |
2484 | ||
2485 | Pre-load and sort verify blocks for a read workload. | |
2486 | ||
2487 | .. option:: verify_offset=int | |
2488 | ||
2489 | Swap the verification header with data somewhere else in the block before | |
2490 | writing. It is swapped back before verifying. | |
2491 | ||
2492 | .. option:: verify_interval=int | |
2493 | ||
2494 | Write the verification header at a finer granularity than the | |
2495 | :option:`blocksize`. It will be written for chunks the size of | |
2496 | ``verify_interval``. :option:`blocksize` should divide this evenly. | |
2497 | ||
2498 | .. option:: verify_pattern=str | |
2499 | ||
2500 | If set, fio will fill the I/O buffers with this pattern. Fio defaults to | |
2501 | filling with totally random bytes, but sometimes it's interesting to fill | |
2502 | with a known pattern for I/O verification purposes. Depending on the width | |
730bd7d9 | 2503 | of the pattern, fio will fill 1/2/3/4 bytes of the buffer at the time (it can |
f80dba8d MT |
2504 | be either a decimal or a hex number). The ``verify_pattern`` if larger than |
2505 | a 32-bit quantity has to be a hex number that starts with either "0x" or | |
2506 | "0X". Use with :option:`verify`. Also, ``verify_pattern`` supports %o | |
2507 | format, which means that for each block offset will be written and then | |
2508 | verified back, e.g.:: | |
61b9861d RP |
2509 | |
2510 | verify_pattern=%o | |
2511 | ||
f80dba8d MT |
2512 | Or use combination of everything:: |
2513 | ||
61b9861d | 2514 | verify_pattern=0xff%o"abcd"-12 |
e28218f3 | 2515 | |
f80dba8d MT |
2516 | .. option:: verify_fatal=bool |
2517 | ||
2518 | Normally fio will keep checking the entire contents before quitting on a | |
2519 | block verification failure. If this option is set, fio will exit the job on | |
2520 | the first observed failure. Default: false. | |
2521 | ||
2522 | .. option:: verify_dump=bool | |
2523 | ||
2524 | If set, dump the contents of both the original data block and the data block | |
2525 | we read off disk to files. This allows later analysis to inspect just what | |
2526 | kind of data corruption occurred. Off by default. | |
2527 | ||
2528 | .. option:: verify_async=int | |
2529 | ||
2530 | Fio will normally verify I/O inline from the submitting thread. This option | |
2531 | takes an integer describing how many async offload threads to create for I/O | |
2532 | verification instead, causing fio to offload the duty of verifying I/O | |
2533 | contents to one or more separate threads. If using this offload option, even | |
2534 | sync I/O engines can benefit from using an :option:`iodepth` setting higher | |
2535 | than 1, as it allows them to have I/O in flight while verifies are running. | |
d7e6ea1c | 2536 | Defaults to 0 async threads, i.e. verification is not asynchronous. |
f80dba8d MT |
2537 | |
2538 | .. option:: verify_async_cpus=str | |
2539 | ||
2540 | Tell fio to set the given CPU affinity on the async I/O verification | |
2541 | threads. See :option:`cpus_allowed` for the format used. | |
2542 | ||
2543 | .. option:: verify_backlog=int | |
2544 | ||
2545 | Fio will normally verify the written contents of a job that utilizes verify | |
2546 | once that job has completed. In other words, everything is written then | |
2547 | everything is read back and verified. You may want to verify continually | |
2548 | instead for a variety of reasons. Fio stores the meta data associated with | |
2549 | an I/O block in memory, so for large verify workloads, quite a bit of memory | |
2550 | would be used up holding this meta data. If this option is enabled, fio will | |
2551 | write only N blocks before verifying these blocks. | |
2552 | ||
2553 | .. option:: verify_backlog_batch=int | |
2554 | ||
2555 | Control how many blocks fio will verify if :option:`verify_backlog` is | |
2556 | set. If not set, will default to the value of :option:`verify_backlog` | |
2557 | (meaning the entire queue is read back and verified). If | |
2558 | ``verify_backlog_batch`` is less than :option:`verify_backlog` then not all | |
2559 | blocks will be verified, if ``verify_backlog_batch`` is larger than | |
2560 | :option:`verify_backlog`, some blocks will be verified more than once. | |
2561 | ||
2562 | .. option:: verify_state_save=bool | |
2563 | ||
2564 | When a job exits during the write phase of a verify workload, save its | |
2565 | current state. This allows fio to replay up until that point, if the verify | |
2566 | state is loaded for the verify read phase. The format of the filename is, | |
2567 | roughly:: | |
2568 | ||
2569 | <type>-<jobname>-<jobindex>-verify.state. | |
2570 | ||
2571 | <type> is "local" for a local run, "sock" for a client/server socket | |
2572 | connection, and "ip" (192.168.0.1, for instance) for a networked | |
d7e6ea1c | 2573 | client/server connection. Defaults to true. |
f80dba8d MT |
2574 | |
2575 | .. option:: verify_state_load=bool | |
2576 | ||
2577 | If a verify termination trigger was used, fio stores the current write state | |
2578 | of each thread. This can be used at verification time so that fio knows how | |
2579 | far it should verify. Without this information, fio will run a full | |
a47b697c SW |
2580 | verification pass, according to the settings in the job file used. Default |
2581 | false. | |
f80dba8d MT |
2582 | |
2583 | .. option:: trim_percentage=int | |
2584 | ||
2585 | Number of verify blocks to discard/trim. | |
2586 | ||
2587 | .. option:: trim_verify_zero=bool | |
2588 | ||
22413915 | 2589 | Verify that trim/discarded blocks are returned as zeros. |
f80dba8d MT |
2590 | |
2591 | .. option:: trim_backlog=int | |
2592 | ||
22413915 | 2593 | Verify that trim/discarded blocks are returned as zeros. |
f80dba8d MT |
2594 | |
2595 | .. option:: trim_backlog_batch=int | |
2596 | ||
2597 | Trim this number of I/O blocks. | |
2598 | ||
2599 | .. option:: experimental_verify=bool | |
2600 | ||
2601 | Enable experimental verification. | |
2602 | ||
2603 | ||
2604 | Steady state | |
2605 | ~~~~~~~~~~~~ | |
2606 | ||
2607 | .. option:: steadystate=str:float, ss=str:float | |
2608 | ||
2609 | Define the criterion and limit for assessing steady state performance. The | |
2610 | first parameter designates the criterion whereas the second parameter sets | |
2611 | the threshold. When the criterion falls below the threshold for the | |
2612 | specified duration, the job will stop. For example, `iops_slope:0.1%` will | |
2613 | direct fio to terminate the job when the least squares regression slope | |
2614 | falls below 0.1% of the mean IOPS. If :option:`group_reporting` is enabled | |
2615 | this will apply to all jobs in the group. Below is the list of available | |
2616 | steady state assessment criteria. All assessments are carried out using only | |
2617 | data from the rolling collection window. Threshold limits can be expressed | |
2618 | as a fixed value or as a percentage of the mean in the collection window. | |
2619 | ||
2620 | **iops** | |
2621 | Collect IOPS data. Stop the job if all individual IOPS measurements | |
2622 | are within the specified limit of the mean IOPS (e.g., ``iops:2`` | |
2623 | means that all individual IOPS values must be within 2 of the mean, | |
2624 | whereas ``iops:0.2%`` means that all individual IOPS values must be | |
2625 | within 0.2% of the mean IOPS to terminate the job). | |
2626 | ||
2627 | **iops_slope** | |
2628 | Collect IOPS data and calculate the least squares regression | |
2629 | slope. Stop the job if the slope falls below the specified limit. | |
2630 | ||
2631 | **bw** | |
2632 | Collect bandwidth data. Stop the job if all individual bandwidth | |
2633 | measurements are within the specified limit of the mean bandwidth. | |
2634 | ||
2635 | **bw_slope** | |
2636 | Collect bandwidth data and calculate the least squares regression | |
2637 | slope. Stop the job if the slope falls below the specified limit. | |
2638 | ||
2639 | .. option:: steadystate_duration=time, ss_dur=time | |
2640 | ||
2641 | A rolling window of this duration will be used to judge whether steady state | |
2642 | has been reached. Data will be collected once per second. The default is 0 | |
f75ede1d | 2643 | which disables steady state detection. When the unit is omitted, the |
947e0fe0 | 2644 | value is interpreted in seconds. |
f80dba8d MT |
2645 | |
2646 | .. option:: steadystate_ramp_time=time, ss_ramp=time | |
2647 | ||
2648 | Allow the job to run for the specified duration before beginning data | |
2649 | collection for checking the steady state job termination criterion. The | |
947e0fe0 | 2650 | default is 0. When the unit is omitted, the value is interpreted in seconds. |
f80dba8d MT |
2651 | |
2652 | ||
2653 | Measurements and reporting | |
2654 | ~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
2655 | ||
2656 | .. option:: per_job_logs=bool | |
2657 | ||
2658 | If set, this generates bw/clat/iops log with per file private filenames. If | |
2659 | not set, jobs with identical names will share the log filename. Default: | |
2660 | true. | |
2661 | ||
2662 | .. option:: group_reporting | |
2663 | ||
2664 | It may sometimes be interesting to display statistics for groups of jobs as | |
2665 | a whole instead of for each individual job. This is especially true if | |
2666 | :option:`numjobs` is used; looking at individual thread/process output | |
2667 | quickly becomes unwieldy. To see the final report per-group instead of | |
2668 | per-job, use :option:`group_reporting`. Jobs in a file will be part of the | |
2669 | same reporting group, unless if separated by a :option:`stonewall`, or by | |
2670 | using :option:`new_group`. | |
2671 | ||
2672 | .. option:: new_group | |
2673 | ||
2674 | Start a new reporting group. See: :option:`group_reporting`. If not given, | |
2675 | all jobs in a file will be part of the same reporting group, unless | |
2676 | separated by a :option:`stonewall`. | |
2677 | ||
8243be59 JA |
2678 | .. option:: stats |
2679 | ||
2680 | By default, fio collects and shows final output results for all jobs | |
2681 | that run. If this option is set to 0, then fio will ignore it in | |
2682 | the final stat output. | |
2683 | ||
f80dba8d MT |
2684 | .. option:: write_bw_log=str |
2685 | ||
2686 | If given, write a bandwidth log for this job. Can be used to store data of | |
2687 | the bandwidth of the jobs in their lifetime. The included | |
2688 | :command:`fio_generate_plots` script uses :command:`gnuplot` to turn these | |
22413915 | 2689 | text files into nice graphs. See :option:`write_lat_log` for behavior of |
f80dba8d MT |
2690 | given filename. For this option, the postfix is :file:`_bw.x.log`, where `x` |
2691 | is the index of the job (`1..N`, where `N` is the number of jobs). If | |
2692 | :option:`per_job_logs` is false, then the filename will not include the job | |
2693 | index. See `Log File Formats`_. | |
2694 | ||
2695 | .. option:: write_lat_log=str | |
2696 | ||
2697 | Same as :option:`write_bw_log`, except that this option stores I/O | |
2698 | submission, completion, and total latencies instead. If no filename is given | |
2699 | with this option, the default filename of :file:`jobname_type.log` is | |
2700 | used. Even if the filename is given, fio will still append the type of | |
2701 | log. So if one specifies:: | |
e3cedca7 JA |
2702 | |
2703 | write_lat_log=foo | |
2704 | ||
f80dba8d MT |
2705 | The actual log names will be :file:`foo_slat.x.log`, :file:`foo_clat.x.log`, |
2706 | and :file:`foo_lat.x.log`, where `x` is the index of the job (1..N, where N | |
2707 | is the number of jobs). This helps :command:`fio_generate_plot` find the | |
2708 | logs automatically. If :option:`per_job_logs` is false, then the filename | |
2709 | will not include the job index. See `Log File Formats`_. | |
be4ecfdf | 2710 | |
f80dba8d | 2711 | .. option:: write_hist_log=str |
06842027 | 2712 | |
f80dba8d MT |
2713 | Same as :option:`write_lat_log`, but writes I/O completion latency |
2714 | histograms. If no filename is given with this option, the default filename | |
2715 | of :file:`jobname_clat_hist.x.log` is used, where `x` is the index of the | |
2716 | job (1..N, where `N` is the number of jobs). Even if the filename is given, | |
2717 | fio will still append the type of log. If :option:`per_job_logs` is false, | |
2718 | then the filename will not include the job index. See `Log File Formats`_. | |
06842027 | 2719 | |
f80dba8d | 2720 | .. option:: write_iops_log=str |
06842027 | 2721 | |
f80dba8d MT |
2722 | Same as :option:`write_bw_log`, but writes IOPS. If no filename is given |
2723 | with this option, the default filename of :file:`jobname_type.x.log` is | |
2724 | used,where `x` is the index of the job (1..N, where `N` is the number of | |
2725 | jobs). Even if the filename is given, fio will still append the type of | |
2726 | log. If :option:`per_job_logs` is false, then the filename will not include | |
2727 | the job index. See `Log File Formats`_. | |
06842027 | 2728 | |
f80dba8d | 2729 | .. option:: log_avg_msec=int |
06842027 | 2730 | |
f80dba8d MT |
2731 | By default, fio will log an entry in the iops, latency, or bw log for every |
2732 | I/O that completes. When writing to the disk log, that can quickly grow to a | |
2733 | very large size. Setting this option makes fio average the each log entry | |
2734 | over the specified period of time, reducing the resolution of the log. See | |
2735 | :option:`log_max_value` as well. Defaults to 0, logging all entries. | |
6fc82095 | 2736 | Also see `Log File Formats`_. |
06842027 | 2737 | |
f80dba8d | 2738 | .. option:: log_hist_msec=int |
06842027 | 2739 | |
f80dba8d MT |
2740 | Same as :option:`log_avg_msec`, but logs entries for completion latency |
2741 | histograms. Computing latency percentiles from averages of intervals using | |
c60ebc45 | 2742 | :option:`log_avg_msec` is inaccurate. Setting this option makes fio log |
f80dba8d MT |
2743 | histogram entries over the specified period of time, reducing log sizes for |
2744 | high IOPS devices while retaining percentile accuracy. See | |
2745 | :option:`log_hist_coarseness` as well. Defaults to 0, meaning histogram | |
2746 | logging is disabled. | |
06842027 | 2747 | |
f80dba8d | 2748 | .. option:: log_hist_coarseness=int |
06842027 | 2749 | |
f80dba8d MT |
2750 | Integer ranging from 0 to 6, defining the coarseness of the resolution of |
2751 | the histogram logs enabled with :option:`log_hist_msec`. For each increment | |
2752 | in coarseness, fio outputs half as many bins. Defaults to 0, for which | |
2753 | histogram logs contain 1216 latency bins. See `Log File Formats`_. | |
8b28bd41 | 2754 | |
f80dba8d | 2755 | .. option:: log_max_value=bool |
66c098b8 | 2756 | |
f80dba8d MT |
2757 | If :option:`log_avg_msec` is set, fio logs the average over that window. If |
2758 | you instead want to log the maximum value, set this option to 1. Defaults to | |
2759 | 0, meaning that averaged values are logged. | |
a696fa2a | 2760 | |
f80dba8d | 2761 | .. option:: log_offset=int |
a696fa2a | 2762 | |
f80dba8d MT |
2763 | If this is set, the iolog options will include the byte offset for the I/O |
2764 | entry as well as the other data values. | |
71bfa161 | 2765 | |
f80dba8d | 2766 | .. option:: log_compression=int |
7de87099 | 2767 | |
f80dba8d MT |
2768 | If this is set, fio will compress the I/O logs as it goes, to keep the |
2769 | memory footprint lower. When a log reaches the specified size, that chunk is | |
2770 | removed and compressed in the background. Given that I/O logs are fairly | |
2771 | highly compressible, this yields a nice memory savings for longer runs. The | |
2772 | downside is that the compression will consume some background CPU cycles, so | |
2773 | it may impact the run. This, however, is also true if the logging ends up | |
2774 | consuming most of the system memory. So pick your poison. The I/O logs are | |
2775 | saved normally at the end of a run, by decompressing the chunks and storing | |
2776 | them in the specified log file. This feature depends on the availability of | |
2777 | zlib. | |
e0b0d892 | 2778 | |
f80dba8d | 2779 | .. option:: log_compression_cpus=str |
e0b0d892 | 2780 | |
f80dba8d MT |
2781 | Define the set of CPUs that are allowed to handle online log compression for |
2782 | the I/O jobs. This can provide better isolation between performance | |
2783 | sensitive jobs, and background compression work. | |
9e684a49 | 2784 | |
f80dba8d | 2785 | .. option:: log_store_compressed=bool |
9e684a49 | 2786 | |
f80dba8d MT |
2787 | If set, fio will store the log files in a compressed format. They can be |
2788 | decompressed with fio, using the :option:`--inflate-log` command line | |
2789 | parameter. The files will be stored with a :file:`.fz` suffix. | |
9e684a49 | 2790 | |
f80dba8d | 2791 | .. option:: log_unix_epoch=bool |
9e684a49 | 2792 | |
f80dba8d MT |
2793 | If set, fio will log Unix timestamps to the log files produced by enabling |
2794 | write_type_log for each log type, instead of the default zero-based | |
2795 | timestamps. | |
2796 | ||
2797 | .. option:: block_error_percentiles=bool | |
2798 | ||
2799 | If set, record errors in trim block-sized units from writes and trims and | |
2800 | output a histogram of how many trims it took to get to errors, and what kind | |
2801 | of error was encountered. | |
2802 | ||
2803 | .. option:: bwavgtime=int | |
2804 | ||
2805 | Average the calculated bandwidth over the given time. Value is specified in | |
2806 | milliseconds. If the job also does bandwidth logging through | |
2807 | :option:`write_bw_log`, then the minimum of this option and | |
2808 | :option:`log_avg_msec` will be used. Default: 500ms. | |
2809 | ||
2810 | .. option:: iopsavgtime=int | |
2811 | ||
2812 | Average the calculated IOPS over the given time. Value is specified in | |
2813 | milliseconds. If the job also does IOPS logging through | |
2814 | :option:`write_iops_log`, then the minimum of this option and | |
2815 | :option:`log_avg_msec` will be used. Default: 500ms. | |
2816 | ||
2817 | .. option:: disk_util=bool | |
2818 | ||
2819 | Generate disk utilization statistics, if the platform supports it. | |
2820 | Default: true. | |
2821 | ||
2822 | .. option:: disable_lat=bool | |
2823 | ||
2824 | Disable measurements of total latency numbers. Useful only for cutting back | |
2825 | the number of calls to :manpage:`gettimeofday(2)`, as that does impact | |
2826 | performance at really high IOPS rates. Note that to really get rid of a | |
2827 | large amount of these calls, this option must be used with | |
f75ede1d | 2828 | :option:`disable_slat` and :option:`disable_bw_measurement` as well. |
f80dba8d MT |
2829 | |
2830 | .. option:: disable_clat=bool | |
2831 | ||
2832 | Disable measurements of completion latency numbers. See | |
2833 | :option:`disable_lat`. | |
2834 | ||
2835 | .. option:: disable_slat=bool | |
2836 | ||
2837 | Disable measurements of submission latency numbers. See | |
2838 | :option:`disable_slat`. | |
2839 | ||
f75ede1d | 2840 | .. option:: disable_bw_measurement=bool, disable_bw=bool |
f80dba8d MT |
2841 | |
2842 | Disable measurements of throughput/bandwidth numbers. See | |
2843 | :option:`disable_lat`. | |
2844 | ||
2845 | .. option:: clat_percentiles=bool | |
2846 | ||
2847 | Enable the reporting of percentiles of completion latencies. | |
2848 | ||
2849 | .. option:: percentile_list=float_list | |
2850 | ||
2851 | Overwrite the default list of percentiles for completion latencies and the | |
2852 | block error histogram. Each number is a floating number in the range | |
2853 | (0,100], and the maximum length of the list is 20. Use ``:`` to separate the | |
2854 | numbers, and list the numbers in ascending order. For example, | |
2855 | ``--percentile_list=99.5:99.9`` will cause fio to report the values of | |
2856 | completion latency below which 99.5% and 99.9% of the observed latencies | |
2857 | fell, respectively. | |
2858 | ||
2859 | ||
2860 | Error handling | |
2861 | ~~~~~~~~~~~~~~ | |
2862 | ||
2863 | .. option:: exitall_on_error | |
2864 | ||
2865 | When one job finishes in error, terminate the rest. The default is to wait | |
2866 | for each job to finish. | |
2867 | ||
2868 | .. option:: continue_on_error=str | |
2869 | ||
2870 | Normally fio will exit the job on the first observed failure. If this option | |
2871 | is set, fio will continue the job when there is a 'non-fatal error' (EIO or | |
2872 | EILSEQ) until the runtime is exceeded or the I/O size specified is | |
2873 | completed. If this option is used, there are two more stats that are | |
2874 | appended, the total error count and the first error. The error field given | |
2875 | in the stats is the first error that was hit during the run. | |
2876 | ||
2877 | The allowed values are: | |
2878 | ||
2879 | **none** | |
2880 | Exit on any I/O or verify errors. | |
2881 | ||
2882 | **read** | |
2883 | Continue on read errors, exit on all others. | |
2884 | ||
2885 | **write** | |
2886 | Continue on write errors, exit on all others. | |
2887 | ||
2888 | **io** | |
2889 | Continue on any I/O error, exit on all others. | |
2890 | ||
2891 | **verify** | |
2892 | Continue on verify errors, exit on all others. | |
2893 | ||
2894 | **all** | |
2895 | Continue on all errors. | |
2896 | ||
2897 | **0** | |
2898 | Backward-compatible alias for 'none'. | |
2899 | ||
2900 | **1** | |
2901 | Backward-compatible alias for 'all'. | |
2902 | ||
2903 | .. option:: ignore_error=str | |
2904 | ||
2905 | Sometimes you want to ignore some errors during test in that case you can | |
a35ef7cb TK |
2906 | specify error list for each error type, instead of only being able to |
2907 | ignore the default 'non-fatal error' using :option:`continue_on_error`. | |
f80dba8d MT |
2908 | ``ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST`` errors for |
2909 | given error type is separated with ':'. Error may be symbol ('ENOSPC', | |
2910 | 'ENOMEM') or integer. Example:: | |
2911 | ||
2912 | ignore_error=EAGAIN,ENOSPC:122 | |
2913 | ||
2914 | This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from | |
a35ef7cb TK |
2915 | WRITE. This option works by overriding :option:`continue_on_error` with |
2916 | the list of errors for each error type if any. | |
f80dba8d MT |
2917 | |
2918 | .. option:: error_dump=bool | |
2919 | ||
2920 | If set dump every error even if it is non fatal, true by default. If | |
2921 | disabled only fatal error will be dumped. | |
2922 | ||
f75ede1d SW |
2923 | Running predefined workloads |
2924 | ---------------------------- | |
2925 | ||
2926 | Fio includes predefined profiles that mimic the I/O workloads generated by | |
2927 | other tools. | |
2928 | ||
2929 | .. option:: profile=str | |
2930 | ||
2931 | The predefined workload to run. Current profiles are: | |
2932 | ||
2933 | **tiobench** | |
2934 | Threaded I/O bench (tiotest/tiobench) like workload. | |
2935 | ||
2936 | **act** | |
2937 | Aerospike Certification Tool (ACT) like workload. | |
2938 | ||
2939 | To view a profile's additional options use :option:`--cmdhelp` after specifying | |
2940 | the profile. For example:: | |
2941 | ||
2942 | $ fio --profile=act --cmdhelp | |
2943 | ||
2944 | Act profile options | |
2945 | ~~~~~~~~~~~~~~~~~~~ | |
2946 | ||
2947 | .. option:: device-names=str | |
2948 | :noindex: | |
2949 | ||
2950 | Devices to use. | |
2951 | ||
2952 | .. option:: load=int | |
2953 | :noindex: | |
2954 | ||
2955 | ACT load multiplier. Default: 1. | |
2956 | ||
2957 | .. option:: test-duration=time | |
2958 | :noindex: | |
2959 | ||
947e0fe0 SW |
2960 | How long the entire test takes to run. When the unit is omitted, the value |
2961 | is given in seconds. Default: 24h. | |
f75ede1d SW |
2962 | |
2963 | .. option:: threads-per-queue=int | |
2964 | :noindex: | |
2965 | ||
2966 | Number of read IO threads per device. Default: 8. | |
2967 | ||
2968 | .. option:: read-req-num-512-blocks=int | |
2969 | :noindex: | |
2970 | ||
2971 | Number of 512B blocks to read at the time. Default: 3. | |
2972 | ||
2973 | .. option:: large-block-op-kbytes=int | |
2974 | :noindex: | |
2975 | ||
2976 | Size of large block ops in KiB (writes). Default: 131072. | |
2977 | ||
2978 | .. option:: prep | |
2979 | :noindex: | |
2980 | ||
2981 | Set to run ACT prep phase. | |
2982 | ||
2983 | Tiobench profile options | |
2984 | ~~~~~~~~~~~~~~~~~~~~~~~~ | |
2985 | ||
2986 | .. option:: size=str | |
2987 | :noindex: | |
2988 | ||
2989 | Size in MiB | |
2990 | ||
2991 | .. option:: block=int | |
2992 | :noindex: | |
2993 | ||
2994 | Block size in bytes. Default: 4096. | |
2995 | ||
2996 | .. option:: numruns=int | |
2997 | :noindex: | |
2998 | ||
2999 | Number of runs. | |
3000 | ||
3001 | .. option:: dir=str | |
3002 | :noindex: | |
3003 | ||
3004 | Test directory. | |
3005 | ||
3006 | .. option:: threads=int | |
3007 | :noindex: | |
3008 | ||
3009 | Number of threads. | |
f80dba8d MT |
3010 | |
3011 | Interpreting the output | |
3012 | ----------------------- | |
3013 | ||
36214730 SW |
3014 | .. |
3015 | Example output was based on the following: | |
3016 | TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --time_based \ | |
3017 | --rate=1256k --bs=14K --name=quick --runtime=1s --name=mixed \ | |
3018 | --runtime=2m --rw=rw | |
3019 | ||
f80dba8d MT |
3020 | Fio spits out a lot of output. While running, fio will display the status of the |
3021 | jobs created. An example of that would be:: | |
3022 | ||
9d25d068 | 3023 | Jobs: 1 (f=1): [_(1),M(1)][24.8%][r=20.5MiB/s,w=23.5MiB/s][r=82,w=94 IOPS][eta 01m:31s] |
f80dba8d | 3024 | |
36214730 SW |
3025 | The characters inside the first set of square brackets denote the current status of |
3026 | each thread. The first character is the first job defined in the job file, and so | |
3027 | forth. The possible values (in typical life cycle order) are: | |
f80dba8d MT |
3028 | |
3029 | +------+-----+-----------------------------------------------------------+ | |
3030 | | Idle | Run | | | |
3031 | +======+=====+===========================================================+ | |
3032 | | P | | Thread setup, but not started. | | |
3033 | +------+-----+-----------------------------------------------------------+ | |
3034 | | C | | Thread created. | | |
3035 | +------+-----+-----------------------------------------------------------+ | |
3036 | | I | | Thread initialized, waiting or generating necessary data. | | |
3037 | +------+-----+-----------------------------------------------------------+ | |
3038 | | | p | Thread running pre-reading file(s). | | |
3039 | +------+-----+-----------------------------------------------------------+ | |
36214730 SW |
3040 | | | / | Thread is in ramp period. | |
3041 | +------+-----+-----------------------------------------------------------+ | |
f80dba8d MT |
3042 | | | R | Running, doing sequential reads. | |
3043 | +------+-----+-----------------------------------------------------------+ | |
3044 | | | r | Running, doing random reads. | | |
3045 | +------+-----+-----------------------------------------------------------+ | |
3046 | | | W | Running, doing sequential writes. | | |
3047 | +------+-----+-----------------------------------------------------------+ | |
3048 | | | w | Running, doing random writes. | | |
3049 | +------+-----+-----------------------------------------------------------+ | |
3050 | | | M | Running, doing mixed sequential reads/writes. | | |
3051 | +------+-----+-----------------------------------------------------------+ | |
3052 | | | m | Running, doing mixed random reads/writes. | | |
3053 | +------+-----+-----------------------------------------------------------+ | |
36214730 SW |
3054 | | | D | Running, doing sequential trims. | |
3055 | +------+-----+-----------------------------------------------------------+ | |
3056 | | | d | Running, doing random trims. | | |
3057 | +------+-----+-----------------------------------------------------------+ | |
3058 | | | F | Running, currently waiting for :manpage:`fsync(2)`. | | |
f80dba8d MT |
3059 | +------+-----+-----------------------------------------------------------+ |
3060 | | | V | Running, doing verification of written data. | | |
3061 | +------+-----+-----------------------------------------------------------+ | |
36214730 SW |
3062 | | f | | Thread finishing. | |
3063 | +------+-----+-----------------------------------------------------------+ | |
f80dba8d MT |
3064 | | E | | Thread exited, not reaped by main thread yet. | |
3065 | +------+-----+-----------------------------------------------------------+ | |
36214730 | 3066 | | _ | | Thread reaped. | |
f80dba8d MT |
3067 | +------+-----+-----------------------------------------------------------+ |
3068 | | X | | Thread reaped, exited with an error. | | |
3069 | +------+-----+-----------------------------------------------------------+ | |
3070 | | K | | Thread reaped, exited due to signal. | | |
3071 | +------+-----+-----------------------------------------------------------+ | |
3072 | ||
36214730 SW |
3073 | .. |
3074 | Example output was based on the following: | |
3075 | TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --runtime=58m \ | |
3076 | --time_based --rate=2512k --bs=256K --numjobs=10 \ | |
3077 | --name=readers --rw=read --name=writers --rw=write | |
3078 | ||
f80dba8d | 3079 | Fio will condense the thread string as not to take up more space on the command |
36214730 | 3080 | line than needed. For instance, if you have 10 readers and 10 writers running, |
f80dba8d MT |
3081 | the output would look like this:: |
3082 | ||
9d25d068 | 3083 | Jobs: 20 (f=20): [R(10),W(10)][4.0%][r=20.5MiB/s,w=23.5MiB/s][r=82,w=94 IOPS][eta 57m:36s] |
f80dba8d | 3084 | |
36214730 SW |
3085 | Note that the status string is displayed in order, so it's possible to tell which of |
3086 | the jobs are currently doing what. In the example above this means that jobs 1--10 | |
3087 | are readers and 11--20 are writers. | |
f80dba8d MT |
3088 | |
3089 | The other values are fairly self explanatory -- number of threads currently | |
36214730 SW |
3090 | running and doing I/O, the number of currently open files (f=), the estimated |
3091 | completion percentage, the rate of I/O since last check (read speed listed first, | |
3092 | then write speed and optionally trim speed) in terms of bandwidth and IOPS, and time to completion for the current | |
f80dba8d | 3093 | running group. It's impossible to estimate runtime of the following groups (if |
36214730 SW |
3094 | any). |
3095 | ||
3096 | .. | |
3097 | Example output was based on the following: | |
3098 | TZ=UTC fio --iodepth=16 --ioengine=posixaio --filename=/tmp/fiofile \ | |
3099 | --direct=1 --size=100M --time_based --runtime=50s --rate_iops=89 \ | |
3100 | --bs=7K --name=Client1 --rw=write | |
3101 | ||
3102 | When fio is done (or interrupted by :kbd:`Ctrl-C`), it will show the data for | |
3103 | each thread, group of threads, and disks in that order. For each overall thread (or | |
3104 | group) the output looks like:: | |
3105 | ||
3106 | Client1: (groupid=0, jobs=1): err= 0: pid=16109: Sat Jun 24 12:07:54 2017 | |
3107 | write: IOPS=88, BW=623KiB/s (638kB/s)(30.4MiB/50032msec) | |
3108 | slat (nsec): min=500, max=145500, avg=8318.00, stdev=4781.50 | |
3109 | clat (usec): min=170, max=78367, avg=4019.02, stdev=8293.31 | |
3110 | lat (usec): min=174, max=78375, avg=4027.34, stdev=8291.79 | |
3111 | clat percentiles (usec): | |
3112 | | 1.00th=[ 302], 5.00th=[ 326], 10.00th=[ 343], 20.00th=[ 363], | |
3113 | | 30.00th=[ 392], 40.00th=[ 404], 50.00th=[ 416], 60.00th=[ 445], | |
3114 | | 70.00th=[ 816], 80.00th=[ 6718], 90.00th=[12911], 95.00th=[21627], | |
3115 | | 99.00th=[43779], 99.50th=[51643], 99.90th=[68682], 99.95th=[72877], | |
3116 | | 99.99th=[78119] | |
3117 | bw ( KiB/s): min= 532, max= 686, per=0.10%, avg=622.87, stdev=24.82, samples= 100 | |
3118 | iops : min= 76, max= 98, avg=88.98, stdev= 3.54, samples= 100 | |
3119 | lat (usec) : 250=0.04%, 500=64.11%, 750=4.81%, 1000=2.79% | |
3120 | lat (msec) : 2=4.16%, 4=1.84%, 10=4.90%, 20=11.33%, 50=5.37% | |
3121 | lat (msec) : 100=0.65% | |
3122 | cpu : usr=0.27%, sys=0.18%, ctx=12072, majf=0, minf=21 | |
3123 | IO depths : 1=85.0%, 2=13.1%, 4=1.8%, 8=0.1%, 16=0.0%, 32=0.0%, >=64=0.0% | |
3124 | submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% | |
3125 | complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% | |
3126 | issued rwt: total=0,4450,0, short=0,0,0, dropped=0,0,0 | |
3127 | latency : target=0, window=0, percentile=100.00%, depth=8 | |
3128 | ||
3129 | The job name (or first job's name when using :option:`group_reporting`) is printed, | |
3130 | along with the group id, count of jobs being aggregated, last error id seen (which | |
3131 | is 0 when there are no errors), pid/tid of that thread and the time the job/group | |
3132 | completed. Below are the I/O statistics for each data direction performed (showing | |
3133 | writes in the example above). In the order listed, they denote: | |
3134 | ||
3135 | **read/write/trim** | |
3136 | The string before the colon shows the I/O direction the statistics | |
3137 | are for. **IOPS** is the average I/Os performed per second. **BW** | |
3138 | is the average bandwidth rate shown as: value in power of 2 format | |
3139 | (value in power of 10 format). The last two values show: (**total | |
3140 | I/O performed** in power of 2 format / **runtime** of that thread). | |
f80dba8d MT |
3141 | |
3142 | **slat** | |
36214730 SW |
3143 | Submission latency (**min** being the minimum, **max** being the |
3144 | maximum, **avg** being the average, **stdev** being the standard | |
3145 | deviation). This is the time it took to submit the I/O. For | |
3146 | sync I/O this row is not displayed as the slat is really the | |
3147 | completion latency (since queue/complete is one operation there). | |
3148 | This value can be in nanoseconds, microseconds or milliseconds --- | |
3149 | fio will choose the most appropriate base and print that (in the | |
3150 | example above nanoseconds was the best scale). Note: in :option:`--minimal` mode | |
0d237712 | 3151 | latencies are always expressed in microseconds. |
f80dba8d MT |
3152 | |
3153 | **clat** | |
3154 | Completion latency. Same names as slat, this denotes the time from | |
3155 | submission to completion of the I/O pieces. For sync I/O, clat will | |
3156 | usually be equal (or very close) to 0, as the time from submit to | |
3157 | complete is basically just CPU time (I/O has already been done, see slat | |
3158 | explanation). | |
3159 | ||
3160 | **bw** | |
36214730 SW |
3161 | Bandwidth statistics based on samples. Same names as the xlat stats, |
3162 | but also includes the number of samples taken (**samples**) and an | |
3163 | approximate percentage of total aggregate bandwidth this thread | |
3164 | received in its group (**per**). This last value is only really | |
3165 | useful if the threads in this group are on the same disk, since they | |
3166 | are then competing for disk access. | |
3167 | ||
3168 | **iops** | |
3169 | IOPS statistics based on samples. Same names as bw. | |
f80dba8d MT |
3170 | |
3171 | **cpu** | |
3172 | CPU usage. User and system time, along with the number of context | |
3173 | switches this thread went through, usage of system and user time, and | |
3174 | finally the number of major and minor page faults. The CPU utilization | |
3175 | numbers are averages for the jobs in that reporting group, while the | |
23a8e176 | 3176 | context and fault counters are summed. |
f80dba8d MT |
3177 | |
3178 | **IO depths** | |
a2140525 SW |
3179 | The distribution of I/O depths over the job lifetime. The numbers are |
3180 | divided into powers of 2 and each entry covers depths from that value | |
3181 | up to those that are lower than the next entry -- e.g., 16= covers | |
3182 | depths from 16 to 31. Note that the range covered by a depth | |
3183 | distribution entry can be different to the range covered by the | |
3184 | equivalent submit/complete distribution entry. | |
f80dba8d MT |
3185 | |
3186 | **IO submit** | |
3187 | How many pieces of I/O were submitting in a single submit call. Each | |
c60ebc45 | 3188 | entry denotes that amount and below, until the previous entry -- e.g., |
a2140525 SW |
3189 | 16=100% means that we submitted anywhere between 9 to 16 I/Os per submit |
3190 | call. Note that the range covered by a submit distribution entry can | |
3191 | be different to the range covered by the equivalent depth distribution | |
3192 | entry. | |
f80dba8d MT |
3193 | |
3194 | **IO complete** | |
3195 | Like the above submit number, but for completions instead. | |
3196 | ||
36214730 SW |
3197 | **IO issued rwt** |
3198 | The number of read/write/trim requests issued, and how many of them were | |
3199 | short or dropped. | |
f80dba8d MT |
3200 | |
3201 | **IO latencies** | |
3202 | The distribution of I/O completion latencies. This is the time from when | |
3203 | I/O leaves fio and when it gets completed. The numbers follow the same | |
3204 | pattern as the I/O depths, meaning that 2=1.6% means that 1.6% of the | |
3205 | I/O completed within 2 msecs, 20=12.8% means that 12.8% of the I/O took | |
3206 | more than 10 msecs, but less than (or equal to) 20 msecs. | |
71bfa161 | 3207 | |
36214730 SW |
3208 | .. |
3209 | Example output was based on the following: | |
3210 | TZ=UTC fio --ioengine=null --iodepth=2 --size=100M --numjobs=2 \ | |
3211 | --rate_process=poisson --io_limit=32M --name=read --bs=128k \ | |
3212 | --rate=11M --name=write --rw=write --bs=2k --rate=700k | |
3213 | ||
71bfa161 | 3214 | After each client has been listed, the group statistics are printed. They |
f80dba8d | 3215 | will look like this:: |
71bfa161 | 3216 | |
f80dba8d | 3217 | Run status group 0 (all jobs): |
36214730 SW |
3218 | READ: bw=20.9MiB/s (21.9MB/s), 10.4MiB/s-10.8MiB/s (10.9MB/s-11.3MB/s), io=64.0MiB (67.1MB), run=2973-3069msec |
3219 | WRITE: bw=1231KiB/s (1261kB/s), 616KiB/s-621KiB/s (630kB/s-636kB/s), io=64.0MiB (67.1MB), run=52747-53223msec | |
71bfa161 | 3220 | |
36214730 | 3221 | For each data direction it prints: |
71bfa161 | 3222 | |
36214730 SW |
3223 | **bw** |
3224 | Aggregate bandwidth of threads in this group followed by the | |
3225 | minimum and maximum bandwidth of all the threads in this group. | |
3226 | Values outside of brackets are power-of-2 format and those | |
3227 | within are the equivalent value in a power-of-10 format. | |
f80dba8d | 3228 | **io** |
36214730 SW |
3229 | Aggregate I/O performed of all threads in this group. The |
3230 | format is the same as bw. | |
3231 | **run** | |
3232 | The smallest and longest runtimes of the threads in this group. | |
71bfa161 | 3233 | |
f80dba8d | 3234 | And finally, the disk statistics are printed. They will look like this:: |
71bfa161 | 3235 | |
f80dba8d MT |
3236 | Disk stats (read/write): |
3237 | sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00% | |
71bfa161 JA |
3238 | |
3239 | Each value is printed for both reads and writes, with reads first. The | |
3240 | numbers denote: | |
3241 | ||
f80dba8d | 3242 | **ios** |
c60ebc45 | 3243 | Number of I/Os performed by all groups. |
f80dba8d MT |
3244 | **merge** |
3245 | Number of merges I/O the I/O scheduler. | |
3246 | **ticks** | |
3247 | Number of ticks we kept the disk busy. | |
36214730 | 3248 | **in_queue** |
f80dba8d MT |
3249 | Total time spent in the disk queue. |
3250 | **util** | |
3251 | The disk utilization. A value of 100% means we kept the disk | |
71bfa161 JA |
3252 | busy constantly, 50% would be a disk idling half of the time. |
3253 | ||
f80dba8d MT |
3254 | It is also possible to get fio to dump the current output while it is running, |
3255 | without terminating the job. To do that, send fio the **USR1** signal. You can | |
3256 | also get regularly timed dumps by using the :option:`--status-interval` | |
3257 | parameter, or by creating a file in :file:`/tmp` named | |
3258 | :file:`fio-dump-status`. If fio sees this file, it will unlink it and dump the | |
3259 | current output status. | |
8423bd11 | 3260 | |
71bfa161 | 3261 | |
f80dba8d MT |
3262 | Terse output |
3263 | ------------ | |
71bfa161 | 3264 | |
f80dba8d MT |
3265 | For scripted usage where you typically want to generate tables or graphs of the |
3266 | results, fio can output the results in a semicolon separated format. The format | |
3267 | is one long line of values, such as:: | |
71bfa161 | 3268 | |
f80dba8d MT |
3269 | 2;card0;0;0;7139336;121836;60004;1;10109;27.932460;116.933948;220;126861;3495.446807;1085.368601;226;126864;3523.635629;1089.012448;24063;99944;50.275485%;59818.274627;5540.657370;7155060;122104;60004;1;8338;29.086342;117.839068;388;128077;5032.488518;1234.785715;391;128085;5061.839412;1236.909129;23436;100928;50.287926%;59964.832030;5644.844189;14.595833%;19.394167%;123706;0;7313;0.1%;0.1%;0.1%;0.1%;0.1%;0.1%;100.0%;0.00%;0.00%;0.00%;0.00%;0.00%;0.00%;0.01%;0.02%;0.05%;0.16%;6.04%;40.40%;52.68%;0.64%;0.01%;0.00%;0.01%;0.00%;0.00%;0.00%;0.00%;0.00% |
3270 | A description of this job goes here. | |
562c2d2f DN |
3271 | |
3272 | The job description (if provided) follows on a second line. | |
71bfa161 | 3273 | |
a7f77fa6 SW |
3274 | To enable terse output, use the :option:`--minimal` or |
3275 | :option:`--output-format`\=terse command line options. The | |
f80dba8d MT |
3276 | first value is the version of the terse output format. If the output has to be |
3277 | changed for some reason, this number will be incremented by 1 to signify that | |
3278 | change. | |
6820cb3b | 3279 | |
a2c95580 AH |
3280 | Split up, the format is as follows (comments in brackets denote when a |
3281 | field was introduced or whether its specific to some terse version): | |
71bfa161 | 3282 | |
f80dba8d MT |
3283 | :: |
3284 | ||
a2c95580 | 3285 | terse version, fio version [v3], jobname, groupid, error |
f80dba8d MT |
3286 | |
3287 | READ status:: | |
3288 | ||
3289 | Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec) | |
3290 | Submission latency: min, max, mean, stdev (usec) | |
3291 | Completion latency: min, max, mean, stdev (usec) | |
3292 | Completion latency percentiles: 20 fields (see below) | |
3293 | Total latency: min, max, mean, stdev (usec) | |
a2c95580 AH |
3294 | Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5] |
3295 | IOPS [v5]: min, max, mean, stdev, number of samples | |
f80dba8d MT |
3296 | |
3297 | WRITE status: | |
3298 | ||
3299 | :: | |
3300 | ||
3301 | Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec) | |
3302 | Submission latency: min, max, mean, stdev (usec) | |
247823cc | 3303 | Completion latency: min, max, mean, stdev (usec) |
f80dba8d MT |
3304 | Completion latency percentiles: 20 fields (see below) |
3305 | Total latency: min, max, mean, stdev (usec) | |
a2c95580 AH |
3306 | Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5] |
3307 | IOPS [v5]: min, max, mean, stdev, number of samples | |
3308 | ||
3309 | TRIM status [all but version 3]: | |
3310 | ||
3311 | Fields are similar to READ/WRITE status. | |
f80dba8d MT |
3312 | |
3313 | CPU usage:: | |
3314 | ||
3315 | user, system, context switches, major faults, minor faults | |
3316 | ||
3317 | I/O depths:: | |
3318 | ||
3319 | <=1, 2, 4, 8, 16, 32, >=64 | |
3320 | ||
3321 | I/O latencies microseconds:: | |
3322 | ||
3323 | <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000 | |
3324 | ||
3325 | I/O latencies milliseconds:: | |
3326 | ||
3327 | <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000 | |
3328 | ||
a2c95580 | 3329 | Disk utilization [v3]:: |
f80dba8d MT |
3330 | |
3331 | Disk name, Read ios, write ios, | |
3332 | Read merges, write merges, | |
3333 | Read ticks, write ticks, | |
3334 | Time spent in queue, disk utilization percentage | |
3335 | ||
3336 | Additional Info (dependent on continue_on_error, default off):: | |
3337 | ||
3338 | total # errors, first error code | |
3339 | ||
3340 | Additional Info (dependent on description being set):: | |
3341 | ||
3342 | Text description | |
3343 | ||
3344 | Completion latency percentiles can be a grouping of up to 20 sets, so for the | |
3345 | terse output fio writes all of them. Each field will look like this:: | |
1db92cb6 JA |
3346 | |
3347 | 1.00%=6112 | |
3348 | ||
f80dba8d | 3349 | which is the Xth percentile, and the `usec` latency associated with it. |
1db92cb6 | 3350 | |
f80dba8d MT |
3351 | For disk utilization, all disks used by fio are shown. So for each disk there |
3352 | will be a disk utilization section. | |
f2f788dd | 3353 | |
2fc26c3d | 3354 | Below is a single line containing short names for each of the fields in the |
2831be97 | 3355 | minimal output v3, separated by semicolons:: |
2fc26c3d | 3356 | |
586631b8 | 3357 | terse_version_3;fio_version;jobname;groupid;error;read_kb;read_bandwidth;read_iops;read_runtime_ms;read_slat_min;read_slat_max;read_slat_mean;read_slat_dev;read_clat_min;read_clat_max;read_clat_mean;read_clat_dev;read_clat_pct01;read_clat_pct02;read_clat_pct03;read_clat_pct04;read_clat_pct05;read_clat_pct06;read_clat_pct07;read_clat_pct08;read_clat_pct09;read_clat_pct10;read_clat_pct11;read_clat_pct12;read_clat_pct13;read_clat_pct14;read_clat_pct15;read_clat_pct16;read_clat_pct17;read_clat_pct18;read_clat_pct19;read_clat_pct20;read_tlat_min;read_lat_max;read_lat_mean;read_lat_dev;read_bw_min;read_bw_max;read_bw_agg_pct;read_bw_mean;read_bw_dev;write_kb;write_bandwidth;write_iops;write_runtime_ms;write_slat_min;write_slat_max;write_slat_mean;write_slat_dev;write_clat_min;write_clat_max;write_clat_mean;write_clat_dev;write_clat_pct01;write_clat_pct02;write_clat_pct03;write_clat_pct04;write_clat_pct05;write_clat_pct06;write_clat_pct07;write_clat_pct08;write_clat_pct09;write_clat_pct10;write_clat_pct11;write_clat_pct12;write_clat_pct13;write_clat_pct14;write_clat_pct15;write_clat_pct16;write_clat_pct17;write_clat_pct18;write_clat_pct19;write_clat_pct20;write_tlat_min;write_lat_max;write_lat_mean;write_lat_dev;write_bw_min;write_bw_max;write_bw_agg_pct;write_bw_mean;write_bw_dev;cpu_user;cpu_sys;cpu_csw;cpu_mjf;cpu_minf;iodepth_1;iodepth_2;iodepth_4;iodepth_8;iodepth_16;iodepth_32;iodepth_64;lat_2us;lat_4us;lat_10us;lat_20us;lat_50us;lat_100us;lat_250us;lat_500us;lat_750us;lat_1000us;lat_2ms;lat_4ms;lat_10ms;lat_20ms;lat_50ms;lat_100ms;lat_250ms;lat_500ms;lat_750ms;lat_1000ms;lat_2000ms;lat_over_2000ms;disk_name;disk_read_iops;disk_write_iops;disk_read_merges;disk_write_merges;disk_read_ticks;write_ticks;disk_queue_time;disk_util |
2fc26c3d | 3358 | |
25c8b9d7 | 3359 | |
f80dba8d MT |
3360 | Trace file format |
3361 | ----------------- | |
3362 | ||
3363 | There are two trace file format that you can encounter. The older (v1) format is | |
3364 | unsupported since version 1.20-rc3 (March 2008). It will still be described | |
25c8b9d7 PD |
3365 | below in case that you get an old trace and want to understand it. |
3366 | ||
3367 | In any case the trace is a simple text file with a single action per line. | |
3368 | ||
3369 | ||
f80dba8d MT |
3370 | Trace file format v1 |
3371 | ~~~~~~~~~~~~~~~~~~~~ | |
3372 | ||
3373 | Each line represents a single I/O action in the following format:: | |
3374 | ||
3375 | rw, offset, length | |
25c8b9d7 | 3376 | |
f80dba8d | 3377 | where `rw=0/1` for read/write, and the offset and length entries being in bytes. |
25c8b9d7 | 3378 | |
22413915 | 3379 | This format is not supported in fio versions >= 1.20-rc3. |
25c8b9d7 | 3380 | |
25c8b9d7 | 3381 | |
f80dba8d MT |
3382 | Trace file format v2 |
3383 | ~~~~~~~~~~~~~~~~~~~~ | |
25c8b9d7 | 3384 | |
f80dba8d MT |
3385 | The second version of the trace file format was added in fio version 1.17. It |
3386 | allows to access more then one file per trace and has a bigger set of possible | |
3387 | file actions. | |
25c8b9d7 | 3388 | |
f80dba8d | 3389 | The first line of the trace file has to be:: |
25c8b9d7 | 3390 | |
f80dba8d | 3391 | fio version 2 iolog |
25c8b9d7 PD |
3392 | |
3393 | Following this can be lines in two different formats, which are described below. | |
3394 | ||
f80dba8d | 3395 | The file management format:: |
25c8b9d7 | 3396 | |
f80dba8d | 3397 | filename action |
25c8b9d7 PD |
3398 | |
3399 | The filename is given as an absolute path. The action can be one of these: | |
3400 | ||
f80dba8d MT |
3401 | **add** |
3402 | Add the given filename to the trace. | |
3403 | **open** | |
3404 | Open the file with the given filename. The filename has to have | |
3405 | been added with the **add** action before. | |
3406 | **close** | |
3407 | Close the file with the given filename. The file has to have been | |
3408 | opened before. | |
3409 | ||
3410 | ||
3411 | The file I/O action format:: | |
3412 | ||
3413 | filename action offset length | |
3414 | ||
3415 | The `filename` is given as an absolute path, and has to have been added and | |
3416 | opened before it can be used with this format. The `offset` and `length` are | |
3417 | given in bytes. The `action` can be one of these: | |
3418 | ||
3419 | **wait** | |
3420 | Wait for `offset` microseconds. Everything below 100 is discarded. | |
3421 | The time is relative to the previous `wait` statement. | |
3422 | **read** | |
3423 | Read `length` bytes beginning from `offset`. | |
3424 | **write** | |
3425 | Write `length` bytes beginning from `offset`. | |
3426 | **sync** | |
3427 | :manpage:`fsync(2)` the file. | |
3428 | **datasync** | |
3429 | :manpage:`fdatasync(2)` the file. | |
3430 | **trim** | |
3431 | Trim the given file from the given `offset` for `length` bytes. | |
3432 | ||
3433 | CPU idleness profiling | |
3434 | ---------------------- | |
3435 | ||
3436 | In some cases, we want to understand CPU overhead in a test. For example, we | |
3437 | test patches for the specific goodness of whether they reduce CPU usage. | |
3438 | Fio implements a balloon approach to create a thread per CPU that runs at idle | |
3439 | priority, meaning that it only runs when nobody else needs the cpu. | |
3440 | By measuring the amount of work completed by the thread, idleness of each CPU | |
3441 | can be derived accordingly. | |
3442 | ||
3443 | An unit work is defined as touching a full page of unsigned characters. Mean and | |
3444 | standard deviation of time to complete an unit work is reported in "unit work" | |
3445 | section. Options can be chosen to report detailed percpu idleness or overall | |
3446 | system idleness by aggregating percpu stats. | |
3447 | ||
3448 | ||
3449 | Verification and triggers | |
3450 | ------------------------- | |
3451 | ||
3452 | Fio is usually run in one of two ways, when data verification is done. The first | |
3453 | is a normal write job of some sort with verify enabled. When the write phase has | |
3454 | completed, fio switches to reads and verifies everything it wrote. The second | |
3455 | model is running just the write phase, and then later on running the same job | |
3456 | (but with reads instead of writes) to repeat the same I/O patterns and verify | |
3457 | the contents. Both of these methods depend on the write phase being completed, | |
3458 | as fio otherwise has no idea how much data was written. | |
3459 | ||
3460 | With verification triggers, fio supports dumping the current write state to | |
3461 | local files. Then a subsequent read verify workload can load this state and know | |
3462 | exactly where to stop. This is useful for testing cases where power is cut to a | |
3463 | server in a managed fashion, for instance. | |
99b9a85a JA |
3464 | |
3465 | A verification trigger consists of two things: | |
3466 | ||
f80dba8d MT |
3467 | 1) Storing the write state of each job. |
3468 | 2) Executing a trigger command. | |
99b9a85a | 3469 | |
f80dba8d MT |
3470 | The write state is relatively small, on the order of hundreds of bytes to single |
3471 | kilobytes. It contains information on the number of completions done, the last X | |
3472 | completions, etc. | |
99b9a85a | 3473 | |
f80dba8d MT |
3474 | A trigger is invoked either through creation ('touch') of a specified file in |
3475 | the system, or through a timeout setting. If fio is run with | |
9207a0cb | 3476 | :option:`--trigger-file`\= :file:`/tmp/trigger-file`, then it will continually |
f80dba8d MT |
3477 | check for the existence of :file:`/tmp/trigger-file`. When it sees this file, it |
3478 | will fire off the trigger (thus saving state, and executing the trigger | |
99b9a85a JA |
3479 | command). |
3480 | ||
f80dba8d MT |
3481 | For client/server runs, there's both a local and remote trigger. If fio is |
3482 | running as a server backend, it will send the job states back to the client for | |
3483 | safe storage, then execute the remote trigger, if specified. If a local trigger | |
3484 | is specified, the server will still send back the write state, but the client | |
3485 | will then execute the trigger. | |
99b9a85a | 3486 | |
f80dba8d MT |
3487 | Verification trigger example |
3488 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
99b9a85a | 3489 | |
4502cb42 | 3490 | Let's say we want to run a powercut test on the remote machine 'server'. Our |
f80dba8d MT |
3491 | write workload is in :file:`write-test.fio`. We want to cut power to 'server' at |
3492 | some point during the run, and we'll run this test from the safety or our local | |
3493 | machine, 'localbox'. On the server, we'll start the fio backend normally:: | |
99b9a85a | 3494 | |
f80dba8d | 3495 | server# fio --server |
99b9a85a | 3496 | |
f80dba8d | 3497 | and on the client, we'll fire off the workload:: |
99b9a85a | 3498 | |
f80dba8d | 3499 | localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger-remote="bash -c \"echo b > /proc/sysrq-triger\"" |
99b9a85a | 3500 | |
f80dba8d | 3501 | We set :file:`/tmp/my-trigger` as the trigger file, and we tell fio to execute:: |
99b9a85a | 3502 | |
f80dba8d | 3503 | echo b > /proc/sysrq-trigger |
99b9a85a | 3504 | |
f80dba8d MT |
3505 | on the server once it has received the trigger and sent us the write state. This |
3506 | will work, but it's not **really** cutting power to the server, it's merely | |
3507 | abruptly rebooting it. If we have a remote way of cutting power to the server | |
3508 | through IPMI or similar, we could do that through a local trigger command | |
4502cb42 | 3509 | instead. Let's assume we have a script that does IPMI reboot of a given hostname, |
f80dba8d MT |
3510 | ipmi-reboot. On localbox, we could then have run fio with a local trigger |
3511 | instead:: | |
99b9a85a | 3512 | |
f80dba8d | 3513 | localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger="ipmi-reboot server" |
99b9a85a | 3514 | |
f80dba8d MT |
3515 | For this case, fio would wait for the server to send us the write state, then |
3516 | execute ``ipmi-reboot server`` when that happened. | |
3517 | ||
3518 | Loading verify state | |
3519 | ~~~~~~~~~~~~~~~~~~~~ | |
3520 | ||
4502cb42 | 3521 | To load stored write state, a read verification job file must contain the |
f80dba8d | 3522 | :option:`verify_state_load` option. If that is set, fio will load the previously |
99b9a85a | 3523 | stored state. For a local fio run this is done by loading the files directly, |
f80dba8d MT |
3524 | and on a client/server run, the server backend will ask the client to send the |
3525 | files over and load them from there. | |
a3ae5b05 JA |
3526 | |
3527 | ||
f80dba8d MT |
3528 | Log File Formats |
3529 | ---------------- | |
a3ae5b05 JA |
3530 | |
3531 | Fio supports a variety of log file formats, for logging latencies, bandwidth, | |
3532 | and IOPS. The logs share a common format, which looks like this: | |
3533 | ||
f80dba8d | 3534 | *time* (`msec`), *value*, *data direction*, *offset* |
a3ae5b05 | 3535 | |
f80dba8d | 3536 | Time for the log entry is always in milliseconds. The *value* logged depends |
a3ae5b05 JA |
3537 | on the type of log, it will be one of the following: |
3538 | ||
f80dba8d MT |
3539 | **Latency log** |
3540 | Value is latency in usecs | |
3541 | **Bandwidth log** | |
3542 | Value is in KiB/sec | |
3543 | **IOPS log** | |
3544 | Value is IOPS | |
3545 | ||
3546 | *Data direction* is one of the following: | |
3547 | ||
3548 | **0** | |
3549 | I/O is a READ | |
3550 | **1** | |
3551 | I/O is a WRITE | |
3552 | **2** | |
3553 | I/O is a TRIM | |
3554 | ||
3555 | The *offset* is the offset, in bytes, from the start of the file, for that | |
3556 | particular I/O. The logging of the offset can be toggled with | |
3557 | :option:`log_offset`. | |
3558 | ||
6fc82095 SW |
3559 | Fio defaults to logging every individual I/O. When IOPS are logged for individual |
3560 | I/Os the value entry will always be 1. If windowed logging is enabled through | |
3561 | :option:`log_avg_msec`, fio logs the average values over the specified period of time. | |
3562 | If windowed logging is enabled and :option:`log_max_value` is set, then fio logs | |
3563 | maximum values in that window instead of averages. Since 'data direction' and | |
3564 | 'offset' are per-I/O values, they aren't applicable if windowed logging is enabled. | |
f80dba8d | 3565 | |
b8f7e412 | 3566 | Client/Server |
f80dba8d MT |
3567 | ------------- |
3568 | ||
3569 | Normally fio is invoked as a stand-alone application on the machine where the | |
6cf30ac0 SW |
3570 | I/O workload should be generated. However, the backend and frontend of fio can |
3571 | be run separately i.e., the fio server can generate an I/O workload on the "Device | |
3572 | Under Test" while being controlled by a client on another machine. | |
f80dba8d MT |
3573 | |
3574 | Start the server on the machine which has access to the storage DUT:: | |
3575 | ||
3576 | fio --server=args | |
3577 | ||
dbb257bb | 3578 | where `args` defines what fio listens to. The arguments are of the form |
f80dba8d MT |
3579 | ``type,hostname`` or ``IP,port``. *type* is either ``ip`` (or ip4) for TCP/IP |
3580 | v4, ``ip6`` for TCP/IP v6, or ``sock`` for a local unix domain socket. | |
3581 | *hostname* is either a hostname or IP address, and *port* is the port to listen | |
3582 | to (only valid for TCP/IP, not a local socket). Some examples: | |
3583 | ||
3584 | 1) ``fio --server`` | |
3585 | ||
3586 | Start a fio server, listening on all interfaces on the default port (8765). | |
3587 | ||
3588 | 2) ``fio --server=ip:hostname,4444`` | |
3589 | ||
3590 | Start a fio server, listening on IP belonging to hostname and on port 4444. | |
3591 | ||
3592 | 3) ``fio --server=ip6:::1,4444`` | |
3593 | ||
3594 | Start a fio server, listening on IPv6 localhost ::1 and on port 4444. | |
3595 | ||
3596 | 4) ``fio --server=,4444`` | |
3597 | ||
3598 | Start a fio server, listening on all interfaces on port 4444. | |
3599 | ||
3600 | 5) ``fio --server=1.2.3.4`` | |
3601 | ||
3602 | Start a fio server, listening on IP 1.2.3.4 on the default port. | |
3603 | ||
3604 | 6) ``fio --server=sock:/tmp/fio.sock`` | |
3605 | ||
dbb257bb | 3606 | Start a fio server, listening on the local socket :file:`/tmp/fio.sock`. |
f80dba8d MT |
3607 | |
3608 | Once a server is running, a "client" can connect to the fio server with:: | |
3609 | ||
3610 | fio <local-args> --client=<server> <remote-args> <job file(s)> | |
3611 | ||
3612 | where `local-args` are arguments for the client where it is running, `server` | |
3613 | is the connect string, and `remote-args` and `job file(s)` are sent to the | |
3614 | server. The `server` string follows the same format as it does on the server | |
3615 | side, to allow IP/hostname/socket and port strings. | |
3616 | ||
3617 | Fio can connect to multiple servers this way:: | |
3618 | ||
3619 | fio --client=<server1> <job file(s)> --client=<server2> <job file(s)> | |
3620 | ||
3621 | If the job file is located on the fio server, then you can tell the server to | |
3622 | load a local file as well. This is done by using :option:`--remote-config` :: | |
3623 | ||
3624 | fio --client=server --remote-config /path/to/file.fio | |
3625 | ||
3626 | Then fio will open this local (to the server) job file instead of being passed | |
3627 | one from the client. | |
3628 | ||
3629 | If you have many servers (example: 100 VMs/containers), you can input a pathname | |
3630 | of a file containing host IPs/names as the parameter value for the | |
3631 | :option:`--client` option. For example, here is an example :file:`host.list` | |
3632 | file containing 2 hostnames:: | |
3633 | ||
3634 | host1.your.dns.domain | |
3635 | host2.your.dns.domain | |
3636 | ||
3637 | The fio command would then be:: | |
a3ae5b05 | 3638 | |
f80dba8d | 3639 | fio --client=host.list <job file(s)> |
a3ae5b05 | 3640 | |
f80dba8d MT |
3641 | In this mode, you cannot input server-specific parameters or job files -- all |
3642 | servers receive the same job file. | |
a3ae5b05 | 3643 | |
f80dba8d MT |
3644 | In order to let ``fio --client`` runs use a shared filesystem from multiple |
3645 | hosts, ``fio --client`` now prepends the IP address of the server to the | |
4502cb42 | 3646 | filename. For example, if fio is using the directory :file:`/mnt/nfs/fio` and is |
f80dba8d MT |
3647 | writing filename :file:`fileio.tmp`, with a :option:`--client` `hostfile` |
3648 | containing two hostnames ``h1`` and ``h2`` with IP addresses 192.168.10.120 and | |
3649 | 192.168.10.121, then fio will create two files:: | |
a3ae5b05 | 3650 | |
f80dba8d MT |
3651 | /mnt/nfs/fio/192.168.10.120.fileio.tmp |
3652 | /mnt/nfs/fio/192.168.10.121.fileio.tmp |