configure: stop enabling fdatasync on OSX
[fio.git] / HOWTO
CommitLineData
f80dba8d
MT
1How fio works
2-------------
3
4The first step in getting fio to simulate a desired I/O workload, is writing a
5job file describing that specific setup. A job file may contain any number of
6threads and/or files -- the typical contents of the job file is a *global*
7section defining shared parameters, and one or more job sections describing the
8jobs involved. When run, fio parses this file and sets everything up as
9described. If we break down a job from top to bottom, it contains the following
10basic 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
48The above are the basic parameters defined for a workload, in addition there's a
49multitude of parameters that modify other aspects of how this job behaves.
50
51
52Command line options
53--------------------
54
55.. option:: --debug=type
56
f50fbdda 57 Enable verbose tracing `type` of various fio actions. May be ``all`` for all types
b034c0dd
TK
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.
a02ec45a
VF
96 *steadystate*
97 Dump info related to steadystate detection.
98 *helperthread*
99 Dump info related to the helper thread.
100 *zbd*
101 Dump info related to support for zoned block devices.
b034c0dd 102 *?* or *help*
f80dba8d
MT
103 Show available debug options.
104
105.. option:: --parse-only
106
25cd4b95 107 Parse options only, don't start any I/O.
f80dba8d 108
b9921d1a
DZ
109.. option:: --merge-blktrace-only
110
111 Merge blktraces only, don't start any I/O.
112
f80dba8d
MT
113.. option:: --output=filename
114
115 Write output to file `filename`.
116
f50fbdda 117.. option:: --output-format=format
b8f7e412 118
f50fbdda 119 Set the reporting `format` to `normal`, `terse`, `json`, or `json+`. Multiple
b8f7e412
TK
120 formats can be selected, separated by a comma. `terse` is a CSV based
121 format. `json+` is like `json`, except it adds a full dump of the latency
122 buckets.
123
f80dba8d
MT
124.. option:: --bandwidth-log
125
126 Generate aggregate bandwidth logs.
127
128.. option:: --minimal
129
130 Print statistics in a terse, semicolon-delimited format.
131
132.. option:: --append-terse
133
b034c0dd
TK
134 Print statistics in selected mode AND terse, semicolon-delimited format.
135 **Deprecated**, use :option:`--output-format` instead to select multiple
136 formats.
f80dba8d 137
f50fbdda 138.. option:: --terse-version=version
f80dba8d 139
f50fbdda 140 Set terse `version` output format (default 3, or 2 or 4 or 5).
f80dba8d
MT
141
142.. option:: --version
143
b8f7e412 144 Print version information and exit.
f80dba8d
MT
145
146.. option:: --help
147
113f0e7c 148 Print a summary of the command line options and exit.
f80dba8d
MT
149
150.. option:: --cpuclock-test
151
152 Perform test and validation of internal CPU clock.
153
113f0e7c 154.. option:: --crctest=[test]
f80dba8d 155
b034c0dd
TK
156 Test the speed of the built-in checksumming functions. If no argument is
157 given, all of them are tested. Alternatively, a comma separated list can
158 be passed, in which case the given ones are tested.
f80dba8d
MT
159
160.. option:: --cmdhelp=command
161
162 Print help information for `command`. May be ``all`` for all commands.
163
164.. option:: --enghelp=[ioengine[,command]]
165
f50fbdda
TK
166 List all commands defined by `ioengine`, or print help for `command`
167 defined by `ioengine`. If no `ioengine` is given, list all
b034c0dd 168 available ioengines.
f80dba8d
MT
169
170.. option:: --showcmd=jobfile
171
b8f7e412 172 Convert `jobfile` to a set of command-line options.
f80dba8d
MT
173
174.. option:: --readonly
175
4027b2a1
VF
176 Turn on safety read-only checks, preventing writes and trims. The
177 ``--readonly`` option is an extra safety guard to prevent users from
178 accidentally starting a write or trim workload when that is not desired.
179 Fio will only modify the device under test if
180 `rw=write/randwrite/rw/randrw/trim/randtrim/trimwrite` is given. This
181 safety net can be used as an extra precaution.
f80dba8d
MT
182
183.. option:: --eta=when
184
b8f7e412 185 Specifies when real-time ETA estimate should be printed. `when` may be
db37d890
JA
186 `always`, `never` or `auto`. `auto` is the default, it prints ETA
187 when requested if the output is a TTY. `always` disregards the output
188 type, and prints ETA when requested. `never` never prints ETA.
189
190.. option:: --eta-interval=time
191
192 By default, fio requests client ETA status roughly every second. With
193 this option, the interval is configurable. Fio imposes a minimum
194 allowed time to avoid flooding the console, less than 250 msec is
195 not supported.
f80dba8d
MT
196
197.. option:: --eta-newline=time
198
947e0fe0
SW
199 Force a new line for every `time` period passed. When the unit is omitted,
200 the value is interpreted in seconds.
f80dba8d
MT
201
202.. option:: --status-interval=time
203
aa6cb459
VF
204 Force a full status dump of cumulative (from job start) values at `time`
205 intervals. This option does *not* provide per-period measurements. So
206 values such as bandwidth are running averages. When the time unit is omitted,
c1f4de8a
JA
207 `time` is interpreted in seconds. Note that using this option with
208 ``--output-format=json`` will yield output that technically isn't valid
209 json, since the output will be collated sets of valid json. It will need
210 to be split into valid sets of json after the run.
f80dba8d
MT
211
212.. option:: --section=name
213
b034c0dd
TK
214 Only run specified section `name` in job file. Multiple sections can be specified.
215 The ``--section`` option allows one to combine related jobs into one file.
216 E.g. one job file could define light, moderate, and heavy sections. Tell
217 fio to run only the "heavy" section by giving ``--section=heavy``
218 command line option. One can also specify the "write" operations in one
219 section and "verify" operation in another section. The ``--section`` option
220 only applies to job sections. The reserved *global* section is always
221 parsed and used.
f80dba8d
MT
222
223.. option:: --alloc-size=kb
224
4a419903
VF
225 Allocate additional internal smalloc pools of size `kb` in KiB. The
226 ``--alloc-size`` option increases shared memory set aside for use by fio.
b034c0dd
TK
227 If running large jobs with randommap enabled, fio can run out of memory.
228 Smalloc is an internal allocator for shared structures from a fixed size
229 memory pool and can grow to 16 pools. The pool size defaults to 16MiB.
f80dba8d 230
b034c0dd
TK
231 NOTE: While running :file:`.fio_smalloc.*` backing store files are visible
232 in :file:`/tmp`.
f80dba8d
MT
233
234.. option:: --warnings-fatal
235
b034c0dd
TK
236 All fio parser warnings are fatal, causing fio to exit with an
237 error.
f80dba8d
MT
238
239.. option:: --max-jobs=nr
240
f50fbdda 241 Set the maximum number of threads/processes to support to `nr`.
818322cc 242 NOTE: On Linux, it may be necessary to increase the shared-memory
71aa48eb 243 limit (:file:`/proc/sys/kernel/shmmax`) if fio runs into errors while
818322cc 244 creating jobs.
f80dba8d
MT
245
246.. option:: --server=args
247
b034c0dd
TK
248 Start a backend server, with `args` specifying what to listen to.
249 See `Client/Server`_ section.
f80dba8d
MT
250
251.. option:: --daemonize=pidfile
252
b034c0dd 253 Background a fio server, writing the pid to the given `pidfile` file.
f80dba8d
MT
254
255.. option:: --client=hostname
256
f50fbdda 257 Instead of running the jobs locally, send and run them on the given `hostname`
71aa48eb 258 or set of `hostname`\s. See `Client/Server`_ section.
f80dba8d
MT
259
260.. option:: --remote-config=file
261
f50fbdda 262 Tell fio server to load this local `file`.
f80dba8d
MT
263
264.. option:: --idle-prof=option
265
b8f7e412 266 Report CPU idleness. `option` is one of the following:
113f0e7c
SW
267
268 **calibrate**
269 Run unit work calibration only and exit.
270
271 **system**
272 Show aggregate system idleness and unit work.
273
274 **percpu**
275 As **system** but also show per CPU idleness.
f80dba8d
MT
276
277.. option:: --inflate-log=log
278
f50fbdda 279 Inflate and output compressed `log`.
f80dba8d
MT
280
281.. option:: --trigger-file=file
282
f50fbdda 283 Execute trigger command when `file` exists.
f80dba8d 284
f50fbdda 285.. option:: --trigger-timeout=time
f80dba8d 286
f50fbdda 287 Execute trigger at this `time`.
f80dba8d 288
f50fbdda 289.. option:: --trigger=command
f80dba8d 290
f50fbdda 291 Set this `command` as local trigger.
f80dba8d 292
f50fbdda 293.. option:: --trigger-remote=command
f80dba8d 294
f50fbdda 295 Set this `command` as remote trigger.
f80dba8d
MT
296
297.. option:: --aux-path=path
298
f4401bf8
SW
299 Use the directory specified by `path` for generated state files instead
300 of the current working directory.
f80dba8d
MT
301
302Any parameters following the options will be assumed to be job files, unless
303they match a job file parameter. Multiple job files can be listed and each job
304file will be regarded as a separate group. Fio will :option:`stonewall`
305execution between each group.
306
307
308Job file format
309---------------
310
311As previously described, fio accepts one or more job files describing what it is
312supposed to do. The job file format is the classic ini file, where the names
c60ebc45 313enclosed in [] brackets define the job name. You are free to use any ASCII name
f80dba8d
MT
314you want, except *global* which has special meaning. Following the job name is
315a sequence of zero or more parameters, one per line, that define the behavior of
316the job. If the first character in a line is a ';' or a '#', the entire line is
317discarded as a comment.
318
319A *global* section sets defaults for the jobs described in that file. A job may
320override a *global* section parameter, and a job file may even have several
321*global* sections if so desired. A job is only affected by a *global* section
322residing above it.
323
f50fbdda
TK
324The :option:`--cmdhelp` option also lists all options. If used with a `command`
325argument, :option:`--cmdhelp` will detail the given `command`.
f80dba8d
MT
326
327See the `examples/` directory for inspiration on how to write job files. Note
328the copyright and license requirements currently apply to `examples/` files.
329
330So let's look at a really simple job file that defines two processes, each
331randomly reading from a 128MiB file:
332
333.. code-block:: ini
334
335 ; -- start job file --
336 [global]
337 rw=randread
338 size=128m
339
340 [job1]
341
342 [job2]
343
344 ; -- end job file --
345
346As you can see, the job file sections themselves are empty as all the described
347parameters are shared. As no :option:`filename` option is given, fio makes up a
348`filename` for each of the jobs as it sees fit. On the command line, this job
349would look as follows::
350
351$ fio --name=global --rw=randread --size=128m --name=job1 --name=job2
352
353
354Let's look at an example that has a number of processes writing randomly to
355files:
356
357.. code-block:: ini
358
359 ; -- start job file --
360 [random-writers]
361 ioengine=libaio
362 iodepth=4
363 rw=randwrite
364 bs=32k
365 direct=0
366 size=64m
367 numjobs=4
368 ; -- end job file --
369
370Here we have no *global* section, as we only have one job defined anyway. We
371want to use async I/O here, with a depth of 4 for each file. We also increased
372the buffer size used to 32KiB and define numjobs to 4 to fork 4 identical
373jobs. The result is 4 processes each randomly writing to their own 64MiB
374file. Instead of using the above job file, you could have given the parameters
375on the command line. For this case, you would specify::
376
377$ fio --name=random-writers --ioengine=libaio --iodepth=4 --rw=randwrite --bs=32k --direct=0 --size=64m --numjobs=4
378
379When fio is utilized as a basis of any reasonably large test suite, it might be
380desirable to share a set of standardized settings across multiple job files.
381Instead of copy/pasting such settings, any section may pull in an external
382:file:`filename.fio` file with *include filename* directive, as in the following
383example::
384
385 ; -- start job file including.fio --
386 [global]
387 filename=/tmp/test
388 filesize=1m
389 include glob-include.fio
390
391 [test]
392 rw=randread
393 bs=4k
394 time_based=1
395 runtime=10
396 include test-include.fio
397 ; -- end job file including.fio --
398
399.. code-block:: ini
400
401 ; -- start job file glob-include.fio --
402 thread=1
403 group_reporting=1
404 ; -- end job file glob-include.fio --
405
406.. code-block:: ini
407
408 ; -- start job file test-include.fio --
409 ioengine=libaio
410 iodepth=4
411 ; -- end job file test-include.fio --
412
413Settings pulled into a section apply to that section only (except *global*
414section). Include directives may be nested in that any included file may contain
415further include directive(s). Include files may not contain [] sections.
416
417
418Environment variables
419~~~~~~~~~~~~~~~~~~~~~
420
421Fio also supports environment variable expansion in job files. Any sub-string of
422the form ``${VARNAME}`` as part of an option value (in other words, on the right
423of the '='), will be expanded to the value of the environment variable called
424`VARNAME`. If no such environment variable is defined, or `VARNAME` is the
425empty string, the empty string will be substituted.
426
427As an example, let's look at a sample fio invocation and job file::
428
429$ SIZE=64m NUMJOBS=4 fio jobfile.fio
430
431.. code-block:: ini
432
433 ; -- start job file --
434 [random-writers]
435 rw=randwrite
436 size=${SIZE}
437 numjobs=${NUMJOBS}
438 ; -- end job file --
439
440This will expand to the following equivalent job file at runtime:
441
442.. code-block:: ini
443
444 ; -- start job file --
445 [random-writers]
446 rw=randwrite
447 size=64m
448 numjobs=4
449 ; -- end job file --
450
451Fio ships with a few example job files, you can also look there for inspiration.
452
453Reserved keywords
454~~~~~~~~~~~~~~~~~
455
456Additionally, fio has a set of reserved keywords that will be replaced
457internally with the appropriate value. Those keywords are:
458
459**$pagesize**
460
461 The architecture page size of the running system.
462
463**$mb_memory**
464
465 Megabytes of total memory in the system.
466
467**$ncpus**
468
469 Number of online available CPUs.
470
471These can be used on the command line or in the job file, and will be
472automatically substituted with the current system values when the job is
473run. Simple math is also supported on these keywords, so you can perform actions
474like::
475
b034c0dd 476 size=8*$mb_memory
f80dba8d
MT
477
478and get that properly expanded to 8 times the size of memory in the machine.
479
480
481Job file parameters
482-------------------
483
484This section describes in details each parameter associated with a job. Some
485parameters take an option of a given type, such as an integer or a
486string. Anywhere a numeric value is required, an arithmetic expression may be
487used, provided it is surrounded by parentheses. Supported operators are:
488
489 - addition (+)
490 - subtraction (-)
491 - multiplication (*)
492 - division (/)
493 - modulus (%)
494 - exponentiation (^)
495
496For time values in expressions, units are microseconds by default. This is
497different than for time values not in expressions (not enclosed in
498parentheses). The following types are used:
499
500
501Parameter types
502~~~~~~~~~~~~~~~
503
504**str**
b034c0dd 505 String: A sequence of alphanumeric characters.
f80dba8d
MT
506
507**time**
008d0feb
SW
508 Integer with possible time suffix. Without a unit value is interpreted as
509 seconds unless otherwise specified. Accepts a suffix of 'd' for days, 'h' for
510 hours, 'm' for minutes, 's' for seconds, 'ms' (or 'msec') for milliseconds and
511 'us' (or 'usec') for microseconds. For example, use 10m for 10 minutes.
f80dba8d
MT
512
513.. _int:
514
515**int**
516 Integer. A whole number value, which may contain an integer prefix
517 and an integer suffix:
518
b034c0dd 519 [*integer prefix*] **number** [*integer suffix*]
f80dba8d
MT
520
521 The optional *integer prefix* specifies the number's base. The default
522 is decimal. *0x* specifies hexadecimal.
523
524 The optional *integer suffix* specifies the number's units, and includes an
525 optional unit prefix and an optional unit. For quantities of data, the
947e0fe0
SW
526 default unit is bytes. For quantities of time, the default unit is seconds
527 unless otherwise specified.
f80dba8d 528
9207a0cb 529 With :option:`kb_base`\=1000, fio follows international standards for unit
f80dba8d
MT
530 prefixes. To specify power-of-10 decimal values defined in the
531 International System of Units (SI):
532
eccce61a
TK
533 * *K* -- means kilo (K) or 1000
534 * *M* -- means mega (M) or 1000**2
535 * *G* -- means giga (G) or 1000**3
536 * *T* -- means tera (T) or 1000**4
537 * *P* -- means peta (P) or 1000**5
f80dba8d
MT
538
539 To specify power-of-2 binary values defined in IEC 80000-13:
540
eccce61a
TK
541 * *Ki* -- means kibi (Ki) or 1024
542 * *Mi* -- means mebi (Mi) or 1024**2
543 * *Gi* -- means gibi (Gi) or 1024**3
544 * *Ti* -- means tebi (Ti) or 1024**4
545 * *Pi* -- means pebi (Pi) or 1024**5
f80dba8d 546
9207a0cb 547 With :option:`kb_base`\=1024 (the default), the unit prefixes are opposite
f80dba8d
MT
548 from those specified in the SI and IEC 80000-13 standards to provide
549 compatibility with old scripts. For example, 4k means 4096.
550
551 For quantities of data, an optional unit of 'B' may be included
b8f7e412 552 (e.g., 'kB' is the same as 'k').
f80dba8d
MT
553
554 The *integer suffix* is not case sensitive (e.g., m/mi mean mebi/mega,
555 not milli). 'b' and 'B' both mean byte, not bit.
556
9207a0cb 557 Examples with :option:`kb_base`\=1000:
f80dba8d
MT
558
559 * *4 KiB*: 4096, 4096b, 4096B, 4ki, 4kib, 4kiB, 4Ki, 4KiB
560 * *1 MiB*: 1048576, 1mi, 1024ki
561 * *1 MB*: 1000000, 1m, 1000k
562 * *1 TiB*: 1099511627776, 1ti, 1024gi, 1048576mi
563 * *1 TB*: 1000000000, 1t, 1000m, 1000000k
564
9207a0cb 565 Examples with :option:`kb_base`\=1024 (default):
f80dba8d
MT
566
567 * *4 KiB*: 4096, 4096b, 4096B, 4k, 4kb, 4kB, 4K, 4KB
568 * *1 MiB*: 1048576, 1m, 1024k
569 * *1 MB*: 1000000, 1mi, 1000ki
570 * *1 TiB*: 1099511627776, 1t, 1024g, 1048576m
571 * *1 TB*: 1000000000, 1ti, 1000mi, 1000000ki
572
573 To specify times (units are not case sensitive):
574
575 * *D* -- means days
576 * *H* -- means hours
4502cb42 577 * *M* -- means minutes
f80dba8d
MT
578 * *s* -- or sec means seconds (default)
579 * *ms* -- or *msec* means milliseconds
580 * *us* -- or *usec* means microseconds
581
582 If the option accepts an upper and lower range, use a colon ':' or
583 minus '-' to separate such values. See :ref:`irange <irange>`.
4502cb42
SW
584 If the lower value specified happens to be larger than the upper value
585 the two values are swapped.
f80dba8d
MT
586
587.. _bool:
588
589**bool**
590 Boolean. Usually parsed as an integer, however only defined for
591 true and false (1 and 0).
592
593.. _irange:
594
595**irange**
596 Integer range with suffix. Allows value range to be given, such as
c60ebc45 597 1024-4096. A colon may also be used as the separator, e.g. 1k:4k. If the
f80dba8d
MT
598 option allows two sets of ranges, they can be specified with a ',' or '/'
599 delimiter: 1k-4k/8k-32k. Also see :ref:`int <int>`.
600
601**float_list**
602 A list of floating point numbers, separated by a ':' character.
603
f5c3bcf2
TK
604With the above in mind, here follows the complete list of fio job parameters.
605
f80dba8d
MT
606
607Units
608~~~~~
609
610.. option:: kb_base=int
611
612 Select the interpretation of unit prefixes in input parameters.
613
614 **1000**
615 Inputs comply with IEC 80000-13 and the International
616 System of Units (SI). Use:
617
618 - power-of-2 values with IEC prefixes (e.g., KiB)
619 - power-of-10 values with SI prefixes (e.g., kB)
620
621 **1024**
622 Compatibility mode (default). To avoid breaking old scripts:
623
624 - power-of-2 values with SI prefixes
625 - power-of-10 values with IEC prefixes
626
627 See :option:`bs` for more details on input parameters.
628
629 Outputs always use correct prefixes. Most outputs include both
630 side-by-side, like::
631
632 bw=2383.3kB/s (2327.4KiB/s)
633
634 If only one value is reported, then kb_base selects the one to use:
635
636 **1000** -- SI prefixes
637
638 **1024** -- IEC prefixes
639
640.. option:: unit_base=int
641
642 Base unit for reporting. Allowed values are:
643
644 **0**
645 Use auto-detection (default).
646 **8**
647 Byte based.
648 **1**
649 Bit based.
650
651
f80dba8d
MT
652Job description
653~~~~~~~~~~~~~~~
654
655.. option:: name=str
656
657 ASCII name of the job. This may be used to override the name printed by fio
658 for this job. Otherwise the job name is used. On the command line this
659 parameter has the special purpose of also signaling the start of a new job.
660
661.. option:: description=str
662
663 Text description of the job. Doesn't do anything except dump this text
664 description when this job is run. It's not parsed.
665
666.. option:: loops=int
667
668 Run the specified number of iterations of this job. Used to repeat the same
669 workload a given number of times. Defaults to 1.
670
671.. option:: numjobs=int
672
79591fa9
TK
673 Create the specified number of clones of this job. Each clone of job
674 is spawned as an independent thread or process. May be used to setup a
f80dba8d
MT
675 larger number of threads/processes doing the same thing. Each thread is
676 reported separately; to see statistics for all clones as a whole, use
677 :option:`group_reporting` in conjunction with :option:`new_group`.
a47b697c 678 See :option:`--max-jobs`. Default: 1.
f80dba8d
MT
679
680
681Time related parameters
682~~~~~~~~~~~~~~~~~~~~~~~
683
684.. option:: runtime=time
685
f75ede1d 686 Tell fio to terminate processing after the specified period of time. It
f80dba8d 687 can be quite hard to determine for how long a specified job will run, so
f75ede1d 688 this parameter is handy to cap the total runtime to a given time. When
804c0839 689 the unit is omitted, the value is interpreted in seconds.
f80dba8d
MT
690
691.. option:: time_based
692
693 If set, fio will run for the duration of the :option:`runtime` specified
694 even if the file(s) are completely read or written. It will simply loop over
695 the same workload as many times as the :option:`runtime` allows.
696
a881438b 697.. option:: startdelay=irange(time)
f80dba8d 698
947e0fe0
SW
699 Delay the start of job for the specified amount of time. Can be a single
700 value or a range. When given as a range, each thread will choose a value
701 randomly from within the range. Value is in seconds if a unit is omitted.
f80dba8d
MT
702
703.. option:: ramp_time=time
704
705 If set, fio will run the specified workload for this amount of time before
706 logging any performance numbers. Useful for letting performance settle
707 before logging results, thus minimizing the runtime required for stable
708 results. Note that the ``ramp_time`` is considered lead in time for a job,
709 thus it will increase the total runtime if a special timeout or
f75ede1d
SW
710 :option:`runtime` is specified. When the unit is omitted, the value is
711 given in seconds.
f80dba8d
MT
712
713.. option:: clocksource=str
714
715 Use the given clocksource as the base of timing. The supported options are:
716
717 **gettimeofday**
718 :manpage:`gettimeofday(2)`
719
720 **clock_gettime**
721 :manpage:`clock_gettime(2)`
722
723 **cpu**
724 Internal CPU clock source
725
726 cpu is the preferred clocksource if it is reliable, as it is very fast (and
727 fio is heavy on time calls). Fio will automatically use this clocksource if
728 it's supported and considered reliable on the system it is running on,
729 unless another clocksource is specifically set. For x86/x86-64 CPUs, this
730 means supporting TSC Invariant.
731
732.. option:: gtod_reduce=bool
733
734 Enable all of the :manpage:`gettimeofday(2)` reducing options
f75ede1d 735 (:option:`disable_clat`, :option:`disable_slat`, :option:`disable_bw_measurement`) plus
f80dba8d
MT
736 reduce precision of the timeout somewhat to really shrink the
737 :manpage:`gettimeofday(2)` call count. With this option enabled, we only do
738 about 0.4% of the :manpage:`gettimeofday(2)` calls we would have done if all
739 time keeping was enabled.
740
741.. option:: gtod_cpu=int
742
743 Sometimes it's cheaper to dedicate a single thread of execution to just
744 getting the current time. Fio (and databases, for instance) are very
745 intensive on :manpage:`gettimeofday(2)` calls. With this option, you can set
746 one CPU aside for doing nothing but logging current time to a shared memory
747 location. Then the other threads/processes that run I/O workloads need only
748 copy that segment, instead of entering the kernel with a
749 :manpage:`gettimeofday(2)` call. The CPU set aside for doing these time
750 calls will be excluded from other uses. Fio will manually clear it from the
751 CPU mask of other jobs.
752
753
754Target file/device
755~~~~~~~~~~~~~~~~~~
756
757.. option:: directory=str
758
759 Prefix filenames with this directory. Used to place files in a different
760 location than :file:`./`. You can specify a number of directories by
761 separating the names with a ':' character. These directories will be
02dd2689 762 assigned equally distributed to job clones created by :option:`numjobs` as
f80dba8d
MT
763 long as they are using generated filenames. If specific `filename(s)` are
764 set fio will use the first listed directory, and thereby matching the
f4401bf8
SW
765 `filename` semantic (which generates a file for each clone if not
766 specified, but lets all clones use the same file if set).
f80dba8d 767
02dd2689
SW
768 See the :option:`filename` option for information on how to escape "``:``" and
769 "``\``" characters within the directory path itself.
f80dba8d 770
f4401bf8
SW
771 Note: To control the directory fio will use for internal state files
772 use :option:`--aux-path`.
773
f80dba8d
MT
774.. option:: filename=str
775
776 Fio normally makes up a `filename` based on the job name, thread number, and
02dd2689
SW
777 file number (see :option:`filename_format`). If you want to share files
778 between threads in a job or several
79591fa9
TK
779 jobs with fixed file paths, specify a `filename` for each of them to override
780 the default. If the ioengine is file based, you can specify a number of files
781 by separating the names with a ':' colon. So if you wanted a job to open
782 :file:`/dev/sda` and :file:`/dev/sdb` as the two working files, you would use
783 ``filename=/dev/sda:/dev/sdb``. This also means that whenever this option is
784 specified, :option:`nrfiles` is ignored. The size of regular files specified
02dd2689 785 by this option will be :option:`size` divided by number of files unless an
79591fa9
TK
786 explicit size is specified by :option:`filesize`.
787
02dd2689
SW
788 Each colon and backslash in the wanted path must be escaped with a ``\``
789 character. For instance, if the path is :file:`/dev/dsk/foo@3,0:c` then you
790 would use ``filename=/dev/dsk/foo@3,0\:c`` and if the path is
791 :file:`F:\\filename` then you would use ``filename=F\:\\filename``.
792
f80dba8d
MT
793 On Windows, disk devices are accessed as :file:`\\\\.\\PhysicalDrive0` for
794 the first device, :file:`\\\\.\\PhysicalDrive1` for the second etc.
795 Note: Windows and FreeBSD prevent write access to areas
02dd2689
SW
796 of the disk containing in-use data (e.g. filesystems).
797
798 The filename "`-`" is a reserved name, meaning *stdin* or *stdout*. Which
799 of the two depends on the read/write direction set.
f80dba8d
MT
800
801.. option:: filename_format=str
802
803 If sharing multiple files between jobs, it is usually necessary to have fio
804 generate the exact names that you want. By default, fio will name a file
805 based on the default file format specification of
806 :file:`jobname.jobnumber.filenumber`. With this option, that can be
807 customized. Fio will recognize and replace the following keywords in this
808 string:
809
810 **$jobname**
811 The name of the worker thread or process.
812 **$jobnum**
813 The incremental number of the worker thread or process.
814 **$filenum**
815 The incremental number of the file for that worker thread or
816 process.
817
818 To have dependent jobs share a set of files, this option can be set to have
819 fio generate filenames that are shared between the two. For instance, if
820 :file:`testfiles.$filenum` is specified, file number 4 for any job will be
821 named :file:`testfiles.4`. The default of :file:`$jobname.$jobnum.$filenum`
822 will be used if no other format specifier is given.
823
645943c0
JB
824 If you specify a path then the directories will be created up to the
825 main directory for the file. So for example if you specify
826 ``filename_format=a/b/c/$jobnum`` then the directories a/b/c will be
827 created before the file setup part of the job. If you specify
828 :option:`directory` then the path will be relative that directory,
829 otherwise it is treated as the absolute path.
830
f80dba8d
MT
831.. option:: unique_filename=bool
832
833 To avoid collisions between networked clients, fio defaults to prefixing any
834 generated filenames (with a directory specified) with the source of the
835 client connecting. To disable this behavior, set this option to 0.
836
837.. option:: opendir=str
838
839 Recursively open any files below directory `str`.
840
841.. option:: lockfile=str
842
843 Fio defaults to not locking any files before it does I/O to them. If a file
844 or file descriptor is shared, fio can serialize I/O to that file to make the
845 end result consistent. This is usual for emulating real workloads that share
846 files. The lock modes are:
847
848 **none**
849 No locking. The default.
850 **exclusive**
851 Only one thread or process may do I/O at a time, excluding all
852 others.
853 **readwrite**
854 Read-write locking on the file. Many readers may
855 access the file at the same time, but writes get exclusive access.
856
857.. option:: nrfiles=int
858
79591fa9
TK
859 Number of files to use for this job. Defaults to 1. The size of files
860 will be :option:`size` divided by this unless explicit size is specified by
861 :option:`filesize`. Files are created for each thread separately, and each
862 file will have a file number within its name by default, as explained in
863 :option:`filename` section.
864
f80dba8d
MT
865
866.. option:: openfiles=int
867
868 Number of files to keep open at the same time. Defaults to the same as
869 :option:`nrfiles`, can be set smaller to limit the number simultaneous
870 opens.
871
872.. option:: file_service_type=str
873
874 Defines how fio decides which file from a job to service next. The following
875 types are defined:
876
877 **random**
878 Choose a file at random.
879
880 **roundrobin**
881 Round robin over opened files. This is the default.
882
883 **sequential**
884 Finish one file before moving on to the next. Multiple files can
f50fbdda 885 still be open depending on :option:`openfiles`.
f80dba8d
MT
886
887 **zipf**
c60ebc45 888 Use a *Zipf* distribution to decide what file to access.
f80dba8d
MT
889
890 **pareto**
c60ebc45 891 Use a *Pareto* distribution to decide what file to access.
f80dba8d 892
dd3503d3 893 **normal**
c60ebc45 894 Use a *Gaussian* (normal) distribution to decide what file to
f80dba8d
MT
895 access.
896
dd3503d3
SW
897 **gauss**
898 Alias for normal.
899
f80dba8d
MT
900 For *random*, *roundrobin*, and *sequential*, a postfix can be appended to
901 tell fio how many I/Os to issue before switching to a new file. For example,
902 specifying ``file_service_type=random:8`` would cause fio to issue
903 8 I/Os before selecting a new file at random. For the non-uniform
904 distributions, a floating point postfix can be given to influence how the
905 distribution is skewed. See :option:`random_distribution` for a description
906 of how that would work.
907
908.. option:: ioscheduler=str
909
910 Attempt to switch the device hosting the file to the specified I/O scheduler
911 before running.
912
913.. option:: create_serialize=bool
914
915 If true, serialize the file creation for the jobs. This may be handy to
916 avoid interleaving of data files, which may greatly depend on the filesystem
a47b697c 917 used and even the number of processors in the system. Default: true.
f80dba8d
MT
918
919.. option:: create_fsync=bool
920
22413915 921 :manpage:`fsync(2)` the data file after creation. This is the default.
f80dba8d
MT
922
923.. option:: create_on_open=bool
924
730bd7d9
SW
925 If true, don't pre-create files but allow the job's open() to create a file
926 when it's time to do I/O. Default: false -- pre-create all necessary files
927 when the job starts.
f80dba8d
MT
928
929.. option:: create_only=bool
930
931 If true, fio will only run the setup phase of the job. If files need to be
4502cb42 932 laid out or updated on disk, only that will be done -- the actual job contents
a47b697c 933 are not executed. Default: false.
f80dba8d
MT
934
935.. option:: allow_file_create=bool
936
730bd7d9
SW
937 If true, fio is permitted to create files as part of its workload. If this
938 option is false, then fio will error out if
f80dba8d
MT
939 the files it needs to use don't already exist. Default: true.
940
941.. option:: allow_mounted_write=bool
942
c60ebc45 943 If this isn't set, fio will abort jobs that are destructive (e.g. that write)
f80dba8d
MT
944 to what appears to be a mounted device or partition. This should help catch
945 creating inadvertently destructive tests, not realizing that the test will
b1db0375
TK
946 destroy data on the mounted file system. Note that some platforms don't allow
947 writing against a mounted device regardless of this option. Default: false.
f80dba8d
MT
948
949.. option:: pre_read=bool
950
951 If this is given, files will be pre-read into memory before starting the
952 given I/O operation. This will also clear the :option:`invalidate` flag,
953 since it is pointless to pre-read and then drop the cache. This will only
954 work for I/O engines that are seek-able, since they allow you to read the
a47b697c
SW
955 same data multiple times. Thus it will not work on non-seekable I/O engines
956 (e.g. network, splice). Default: false.
f80dba8d
MT
957
958.. option:: unlink=bool
959
960 Unlink the job files when done. Not the default, as repeated runs of that
a47b697c
SW
961 job would then waste time recreating the file set again and again. Default:
962 false.
f80dba8d
MT
963
964.. option:: unlink_each_loop=bool
965
a47b697c 966 Unlink job files after each iteration or loop. Default: false.
f80dba8d 967
7b865a2f
BVA
968.. option:: zonemode=str
969
970 Accepted values are:
971
972 **none**
973 The :option:`zonerange`, :option:`zonesize` and
974 :option:`zoneskip` parameters are ignored.
975 **strided**
976 I/O happens in a single zone until
977 :option:`zonesize` bytes have been transferred.
978 After that number of bytes has been
979 transferred processing of the next zone
980 starts.
981 **zbd**
982 Zoned block device mode. I/O happens
983 sequentially in each zone, even if random I/O
984 has been selected. Random I/O happens across
985 all zones instead of being restricted to a
986 single zone. The :option:`zoneskip` parameter
987 is ignored. :option:`zonerange` and
988 :option:`zonesize` must be identical.
989
5faddc64 990.. option:: zonerange=int
f80dba8d 991
7b865a2f
BVA
992 Size of a single zone. See also :option:`zonesize` and
993 :option:`zoneskip`.
f80dba8d 994
5faddc64 995.. option:: zonesize=int
f80dba8d 996
7b865a2f
BVA
997 For :option:`zonemode` =strided, this is the number of bytes to
998 transfer before skipping :option:`zoneskip` bytes. If this parameter
999 is smaller than :option:`zonerange` then only a fraction of each zone
1000 with :option:`zonerange` bytes will be accessed. If this parameter is
1001 larger than :option:`zonerange` then each zone will be accessed
1002 multiple times before skipping to the next zone.
1003
1004 For :option:`zonemode` =zbd, this is the size of a single zone. The
1005 :option:`zonerange` parameter is ignored in this mode.
f80dba8d
MT
1006
1007.. option:: zoneskip=int
1008
7b865a2f
BVA
1009 For :option:`zonemode` =strided, the number of bytes to skip after
1010 :option:`zonesize` bytes of data have been transferred. This parameter
1011 must be zero for :option:`zonemode` =zbd.
f80dba8d 1012
bfbdd35b
BVA
1013.. option:: read_beyond_wp=bool
1014
1015 This parameter applies to :option:`zonemode` =zbd only.
1016
1017 Zoned block devices are block devices that consist of multiple zones.
1018 Each zone has a type, e.g. conventional or sequential. A conventional
1019 zone can be written at any offset that is a multiple of the block
1020 size. Sequential zones must be written sequentially. The position at
1021 which a write must occur is called the write pointer. A zoned block
1022 device can be either drive managed, host managed or host aware. For
1023 host managed devices the host must ensure that writes happen
1024 sequentially. Fio recognizes host managed devices and serializes
1025 writes to sequential zones for these devices.
1026
1027 If a read occurs in a sequential zone beyond the write pointer then
1028 the zoned block device will complete the read without reading any data
1029 from the storage medium. Since such reads lead to unrealistically high
1030 bandwidth and IOPS numbers fio only reads beyond the write pointer if
1031 explicitly told to do so. Default: false.
1032
59b07544
BVA
1033.. option:: max_open_zones=int
1034
1035 When running a random write test across an entire drive many more
1036 zones will be open than in a typical application workload. Hence this
1037 command line option that allows to limit the number of open zones. The
1038 number of open zones is defined as the number of zones to which write
1039 commands are issued.
1040
a7c2b6fc
BVA
1041.. option:: zone_reset_threshold=float
1042
1043 A number between zero and one that indicates the ratio of logical
1044 blocks with data to the total number of logical blocks in the test
1045 above which zones should be reset periodically.
1046
1047.. option:: zone_reset_frequency=float
1048
1049 A number between zero and one that indicates how often a zone reset
1050 should be issued if the zone reset threshold has been exceeded. A zone
1051 reset is submitted after each (1 / zone_reset_frequency) write
1052 requests. This and the previous parameter can be used to simulate
1053 garbage collection activity.
1054
f80dba8d
MT
1055
1056I/O type
1057~~~~~~~~
1058
1059.. option:: direct=bool
1060
1061 If value is true, use non-buffered I/O. This is usually O_DIRECT. Note that
8e889110 1062 OpenBSD and ZFS on Solaris don't support direct I/O. On Windows the synchronous
f80dba8d
MT
1063 ioengines don't support direct I/O. Default: false.
1064
1065.. option:: atomic=bool
1066
1067 If value is true, attempt to use atomic direct I/O. Atomic writes are
1068 guaranteed to be stable once acknowledged by the operating system. Only
1069 Linux supports O_ATOMIC right now.
1070
1071.. option:: buffered=bool
1072
1073 If value is true, use buffered I/O. This is the opposite of the
1074 :option:`direct` option. Defaults to true.
1075
1076.. option:: readwrite=str, rw=str
1077
1078 Type of I/O pattern. Accepted values are:
1079
1080 **read**
1081 Sequential reads.
1082 **write**
1083 Sequential writes.
1084 **trim**
3740cfc8
VF
1085 Sequential trims (Linux block devices and SCSI
1086 character devices only).
f80dba8d
MT
1087 **randread**
1088 Random reads.
2831be97
SW
1089 **randwrite**
1090 Random writes.
f80dba8d 1091 **randtrim**
3740cfc8
VF
1092 Random trims (Linux block devices and SCSI
1093 character devices only).
f80dba8d
MT
1094 **rw,readwrite**
1095 Sequential mixed reads and writes.
1096 **randrw**
1097 Random mixed reads and writes.
1098 **trimwrite**
1099 Sequential trim+write sequences. Blocks will be trimmed first,
1100 then the same blocks will be written to.
1101
1102 Fio defaults to read if the option is not specified. For the mixed I/O
1103 types, the default is to split them 50/50. For certain types of I/O the
730bd7d9
SW
1104 result may still be skewed a bit, since the speed may be different.
1105
1106 It is possible to specify the number of I/Os to do before getting a new
1107 offset by appending ``:<nr>`` to the end of the string given. For a
f80dba8d
MT
1108 random read, it would look like ``rw=randread:8`` for passing in an offset
1109 modifier with a value of 8. If the suffix is used with a sequential I/O
730bd7d9
SW
1110 pattern, then the *<nr>* value specified will be **added** to the generated
1111 offset for each I/O turning sequential I/O into sequential I/O with holes.
1112 For instance, using ``rw=write:4k`` will skip 4k for every write. Also see
1113 the :option:`rw_sequencer` option.
f80dba8d
MT
1114
1115.. option:: rw_sequencer=str
1116
1117 If an offset modifier is given by appending a number to the ``rw=<str>``
1118 line, then this option controls how that number modifies the I/O offset
1119 being generated. Accepted values are:
1120
1121 **sequential**
1122 Generate sequential offset.
1123 **identical**
1124 Generate the same offset.
1125
1126 ``sequential`` is only useful for random I/O, where fio would normally
c60ebc45 1127 generate a new random offset for every I/O. If you append e.g. 8 to randread,
007c7be9
SW
1128 you would get a new random offset for every 8 I/Os. The result would be a
1129 seek for only every 8 I/Os, instead of for every I/O. Use ``rw=randread:8``
f80dba8d
MT
1130 to specify that. As sequential I/O is already sequential, setting
1131 ``sequential`` for that would not result in any differences. ``identical``
1132 behaves in a similar fashion, except it sends the same offset 8 number of
1133 times before generating a new offset.
1134
1135.. option:: unified_rw_reporting=bool
1136
1137 Fio normally reports statistics on a per data direction basis, meaning that
1138 reads, writes, and trims are accounted and reported separately. If this
1139 option is set fio sums the results and report them as "mixed" instead.
1140
1141.. option:: randrepeat=bool
1142
1143 Seed the random number generator used for random I/O patterns in a
1144 predictable way so the pattern is repeatable across runs. Default: true.
1145
1146.. option:: allrandrepeat=bool
1147
1148 Seed all random number generators in a predictable way so results are
1149 repeatable across runs. Default: false.
1150
1151.. option:: randseed=int
1152
1153 Seed the random number generators based on this seed value, to be able to
1154 control what sequence of output is being generated. If not set, the random
1155 sequence depends on the :option:`randrepeat` setting.
1156
1157.. option:: fallocate=str
1158
1159 Whether pre-allocation is performed when laying down files.
1160 Accepted values are:
1161
1162 **none**
1163 Do not pre-allocate space.
1164
2c3e17be
SW
1165 **native**
1166 Use a platform's native pre-allocation call but fall back to
1167 **none** behavior if it fails/is not implemented.
1168
f80dba8d
MT
1169 **posix**
1170 Pre-allocate via :manpage:`posix_fallocate(3)`.
1171
1172 **keep**
1173 Pre-allocate via :manpage:`fallocate(2)` with
1174 FALLOC_FL_KEEP_SIZE set.
1175
1176 **0**
1177 Backward-compatible alias for **none**.
1178
1179 **1**
1180 Backward-compatible alias for **posix**.
1181
1182 May not be available on all supported platforms. **keep** is only available
2c3e17be
SW
1183 on Linux. If using ZFS on Solaris this cannot be set to **posix**
1184 because ZFS doesn't support pre-allocation. Default: **native** if any
1185 pre-allocation methods are available, **none** if not.
f80dba8d
MT
1186
1187.. option:: fadvise_hint=str
1188
c712c97a
JA
1189 Use :manpage:`posix_fadvise(2)` or :manpage:`posix_fadvise(2)` to
1190 advise the kernel on what I/O patterns are likely to be issued.
1191 Accepted values are:
f80dba8d
MT
1192
1193 **0**
1194 Backwards-compatible hint for "no hint".
1195
1196 **1**
1197 Backwards compatible hint for "advise with fio workload type". This
1198 uses **FADV_RANDOM** for a random workload, and **FADV_SEQUENTIAL**
1199 for a sequential workload.
1200
1201 **sequential**
1202 Advise using **FADV_SEQUENTIAL**.
1203
1204 **random**
1205 Advise using **FADV_RANDOM**.
1206
8f4b9f24 1207.. option:: write_hint=str
f80dba8d 1208
8f4b9f24
JA
1209 Use :manpage:`fcntl(2)` to advise the kernel what life time to expect
1210 from a write. Only supported on Linux, as of version 4.13. Accepted
1211 values are:
1212
1213 **none**
1214 No particular life time associated with this file.
1215
1216 **short**
1217 Data written to this file has a short life time.
1218
1219 **medium**
1220 Data written to this file has a medium life time.
1221
1222 **long**
1223 Data written to this file has a long life time.
1224
1225 **extreme**
1226 Data written to this file has a very long life time.
1227
1228 The values are all relative to each other, and no absolute meaning
1229 should be associated with them.
f80dba8d
MT
1230
1231.. option:: offset=int
1232
82dbb8cb 1233 Start I/O at the provided offset in the file, given as either a fixed size in
83c8b093
JF
1234 bytes or a percentage. If a percentage is given, the generated offset will be
1235 aligned to the minimum ``blocksize`` or to the value of ``offset_align`` if
1236 provided. Data before the given offset will not be touched. This
89978a6b
BW
1237 effectively caps the file size at `real_size - offset`. Can be combined with
1238 :option:`size` to constrain the start and end range of the I/O workload.
44bb1142
TK
1239 A percentage can be specified by a number between 1 and 100 followed by '%',
1240 for example, ``offset=20%`` to specify 20%.
f80dba8d 1241
83c8b093
JF
1242.. option:: offset_align=int
1243
1244 If set to non-zero value, the byte offset generated by a percentage ``offset``
1245 is aligned upwards to this value. Defaults to 0 meaning that a percentage
1246 offset is aligned to the minimum block size.
1247
f80dba8d
MT
1248.. option:: offset_increment=int
1249
1250 If this is provided, then the real offset becomes `offset + offset_increment
1251 * thread_number`, where the thread number is a counter that starts at 0 and
1252 is incremented for each sub-job (i.e. when :option:`numjobs` option is
1253 specified). This option is useful if there are several jobs which are
1254 intended to operate on a file in parallel disjoint segments, with even
0b288ba1
VF
1255 spacing between the starting points. Percentages can be used for this option.
1256 If a percentage is given, the generated offset will be aligned to the minimum
1257 ``blocksize`` or to the value of ``offset_align`` if provided.
f80dba8d
MT
1258
1259.. option:: number_ios=int
1260
c60ebc45 1261 Fio will normally perform I/Os until it has exhausted the size of the region
f80dba8d
MT
1262 set by :option:`size`, or if it exhaust the allocated time (or hits an error
1263 condition). With this setting, the range/size can be set independently of
c60ebc45 1264 the number of I/Os to perform. When fio reaches this number, it will exit
f80dba8d
MT
1265 normally and report status. Note that this does not extend the amount of I/O
1266 that will be done, it will only stop fio if this condition is met before
1267 other end-of-job criteria.
1268
1269.. option:: fsync=int
1270
730bd7d9
SW
1271 If writing to a file, issue an :manpage:`fsync(2)` (or its equivalent) of
1272 the dirty data for every number of blocks given. For example, if you give 32
1273 as a parameter, fio will sync the file after every 32 writes issued. If fio is
1274 using non-buffered I/O, we may not sync the file. The exception is the sg
1275 I/O engine, which synchronizes the disk cache anyway. Defaults to 0, which
1276 means fio does not periodically issue and wait for a sync to complete. Also
1277 see :option:`end_fsync` and :option:`fsync_on_close`.
f80dba8d
MT
1278
1279.. option:: fdatasync=int
1280
1281 Like :option:`fsync` but uses :manpage:`fdatasync(2)` to only sync data and
44f668d7 1282 not metadata blocks. In Windows, FreeBSD, DragonFlyBSD or OSX there is no
730bd7d9
SW
1283 :manpage:`fdatasync(2)` so this falls back to using :manpage:`fsync(2)`.
1284 Defaults to 0, which means fio does not periodically issue and wait for a
1285 data-only sync to complete.
f80dba8d
MT
1286
1287.. option:: write_barrier=int
1288
2831be97 1289 Make every `N-th` write a barrier write.
f80dba8d 1290
f50fbdda 1291.. option:: sync_file_range=str:int
f80dba8d 1292
f50fbdda 1293 Use :manpage:`sync_file_range(2)` for every `int` number of write
f80dba8d
MT
1294 operations. Fio will track range of writes that have happened since the last
1295 :manpage:`sync_file_range(2)` call. `str` can currently be one or more of:
1296
1297 **wait_before**
1298 SYNC_FILE_RANGE_WAIT_BEFORE
1299 **write**
1300 SYNC_FILE_RANGE_WRITE
1301 **wait_after**
1302 SYNC_FILE_RANGE_WAIT_AFTER
1303
1304 So if you do ``sync_file_range=wait_before,write:8``, fio would use
1305 ``SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE`` for every 8
1306 writes. Also see the :manpage:`sync_file_range(2)` man page. This option is
1307 Linux specific.
1308
1309.. option:: overwrite=bool
1310
1311 If true, writes to a file will always overwrite existing data. If the file
1312 doesn't already exist, it will be created before the write phase begins. If
1313 the file exists and is large enough for the specified write phase, nothing
a47b697c 1314 will be done. Default: false.
f80dba8d
MT
1315
1316.. option:: end_fsync=bool
1317
a47b697c
SW
1318 If true, :manpage:`fsync(2)` file contents when a write stage has completed.
1319 Default: false.
f80dba8d
MT
1320
1321.. option:: fsync_on_close=bool
1322
1323 If true, fio will :manpage:`fsync(2)` a dirty file on close. This differs
a47b697c
SW
1324 from :option:`end_fsync` in that it will happen on every file close, not
1325 just at the end of the job. Default: false.
f80dba8d
MT
1326
1327.. option:: rwmixread=int
1328
1329 Percentage of a mixed workload that should be reads. Default: 50.
1330
1331.. option:: rwmixwrite=int
1332
1333 Percentage of a mixed workload that should be writes. If both
1334 :option:`rwmixread` and :option:`rwmixwrite` is given and the values do not
1335 add up to 100%, the latter of the two will be used to override the
1336 first. This may interfere with a given rate setting, if fio is asked to
1337 limit reads or writes to a certain rate. If that is the case, then the
1338 distribution may be skewed. Default: 50.
1339
1340.. option:: random_distribution=str:float[,str:float][,str:float]
1341
1342 By default, fio will use a completely uniform random distribution when asked
1343 to perform random I/O. Sometimes it is useful to skew the distribution in
1344 specific ways, ensuring that some parts of the data is more hot than others.
1345 fio includes the following distribution models:
1346
1347 **random**
1348 Uniform random distribution
1349
1350 **zipf**
1351 Zipf distribution
1352
1353 **pareto**
1354 Pareto distribution
1355
b2f4b559 1356 **normal**
c60ebc45 1357 Normal (Gaussian) distribution
f80dba8d
MT
1358
1359 **zoned**
1360 Zoned random distribution
1361
59466396
JA
1362 **zoned_abs**
1363 Zone absolute random distribution
1364
f80dba8d 1365 When using a **zipf** or **pareto** distribution, an input value is also
f50fbdda 1366 needed to define the access pattern. For **zipf**, this is the `Zipf
c60ebc45 1367 theta`. For **pareto**, it's the `Pareto power`. Fio includes a test
f50fbdda 1368 program, :command:`fio-genzipf`, that can be used visualize what the given input
f80dba8d
MT
1369 values will yield in terms of hit rates. If you wanted to use **zipf** with
1370 a `theta` of 1.2, you would use ``random_distribution=zipf:1.2`` as the
1371 option. If a non-uniform model is used, fio will disable use of the random
b2f4b559
SW
1372 map. For the **normal** distribution, a normal (Gaussian) deviation is
1373 supplied as a value between 0 and 100.
f80dba8d
MT
1374
1375 For a **zoned** distribution, fio supports specifying percentages of I/O
1376 access that should fall within what range of the file or device. For
1377 example, given a criteria of:
1378
f50fbdda
TK
1379 * 60% of accesses should be to the first 10%
1380 * 30% of accesses should be to the next 20%
1381 * 8% of accesses should be to the next 30%
1382 * 2% of accesses should be to the next 40%
f80dba8d
MT
1383
1384 we can define that through zoning of the random accesses. For the above
1385 example, the user would do::
1386
1387 random_distribution=zoned:60/10:30/20:8/30:2/40
1388
59466396
JA
1389 A **zoned_abs** distribution works exactly like the **zoned**, except
1390 that it takes absolute sizes. For example, let's say you wanted to
1391 define access according to the following criteria:
1392
1393 * 60% of accesses should be to the first 20G
1394 * 30% of accesses should be to the next 100G
1395 * 10% of accesses should be to the next 500G
1396
1397 we can define an absolute zoning distribution with:
1398
1399 random_distribution=zoned_abs=60/20G:30/100G:10/500g
1400
6a16ece8
JA
1401 For both **zoned** and **zoned_abs**, fio supports defining up to
1402 256 separate zones.
1403
59466396
JA
1404 Similarly to how :option:`bssplit` works for setting ranges and
1405 percentages of block sizes. Like :option:`bssplit`, it's possible to
1406 specify separate zones for reads, writes, and trims. If just one set
1407 is given, it'll apply to all of them. This goes for both **zoned**
1408 **zoned_abs** distributions.
f80dba8d
MT
1409
1410.. option:: percentage_random=int[,int][,int]
1411
1412 For a random workload, set how big a percentage should be random. This
1413 defaults to 100%, in which case the workload is fully random. It can be set
1414 from anywhere from 0 to 100. Setting it to 0 would make the workload fully
1415 sequential. Any setting in between will result in a random mix of sequential
1416 and random I/O, at the given percentages. Comma-separated values may be
1417 specified for reads, writes, and trims as described in :option:`blocksize`.
1418
1419.. option:: norandommap
1420
1421 Normally fio will cover every block of the file when doing random I/O. If
1422 this option is given, fio will just get a new random offset without looking
1423 at past I/O history. This means that some blocks may not be read or written,
1424 and that some blocks may be read/written more than once. If this option is
1425 used with :option:`verify` and multiple blocksizes (via :option:`bsrange`),
1426 only intact blocks are verified, i.e., partially-overwritten blocks are
47e6a6e5
SW
1427 ignored. With an async I/O engine and an I/O depth > 1, it is possible for
1428 the same block to be overwritten, which can cause verification errors. Either
1429 do not use norandommap in this case, or also use the lfsr random generator.
f80dba8d
MT
1430
1431.. option:: softrandommap=bool
1432
1433 See :option:`norandommap`. If fio runs with the random block map enabled and
1434 it fails to allocate the map, if this option is set it will continue without
1435 a random block map. As coverage will not be as complete as with random maps,
1436 this option is disabled by default.
1437
1438.. option:: random_generator=str
1439
f50fbdda 1440 Fio supports the following engines for generating I/O offsets for random I/O:
f80dba8d
MT
1441
1442 **tausworthe**
f50fbdda 1443 Strong 2^88 cycle random number generator.
f80dba8d 1444 **lfsr**
f50fbdda 1445 Linear feedback shift register generator.
f80dba8d 1446 **tausworthe64**
f50fbdda 1447 Strong 64-bit 2^258 cycle random number generator.
f80dba8d
MT
1448
1449 **tausworthe** is a strong random number generator, but it requires tracking
1450 on the side if we want to ensure that blocks are only read or written
f50fbdda 1451 once. **lfsr** guarantees that we never generate the same offset twice, and
f80dba8d 1452 it's also less computationally expensive. It's not a true random generator,
f50fbdda 1453 however, though for I/O purposes it's typically good enough. **lfsr** only
f80dba8d
MT
1454 works with single block sizes, not with workloads that use multiple block
1455 sizes. If used with such a workload, fio may read or write some blocks
1456 multiple times. The default value is **tausworthe**, unless the required
1457 space exceeds 2^32 blocks. If it does, then **tausworthe64** is
1458 selected automatically.
1459
1460
1461Block size
1462~~~~~~~~~~
1463
1464.. option:: blocksize=int[,int][,int], bs=int[,int][,int]
1465
1466 The block size in bytes used for I/O units. Default: 4096. A single value
1467 applies to reads, writes, and trims. Comma-separated values may be
1468 specified for reads, writes, and trims. A value not terminated in a comma
1469 applies to subsequent types.
1470
1471 Examples:
1472
1473 **bs=256k**
1474 means 256k for reads, writes and trims.
1475
1476 **bs=8k,32k**
1477 means 8k for reads, 32k for writes and trims.
1478
1479 **bs=8k,32k,**
1480 means 8k for reads, 32k for writes, and default for trims.
1481
1482 **bs=,8k**
1483 means default for reads, 8k for writes and trims.
1484
1485 **bs=,8k,**
b443ae44 1486 means default for reads, 8k for writes, and default for trims.
f80dba8d
MT
1487
1488.. option:: blocksize_range=irange[,irange][,irange], bsrange=irange[,irange][,irange]
1489
1490 A range of block sizes in bytes for I/O units. The issued I/O unit will
1491 always be a multiple of the minimum size, unless
1492 :option:`blocksize_unaligned` is set.
1493
1494 Comma-separated ranges may be specified for reads, writes, and trims as
1495 described in :option:`blocksize`.
1496
1497 Example: ``bsrange=1k-4k,2k-8k``.
1498
1499.. option:: bssplit=str[,str][,str]
1500
6a16ece8
JA
1501 Sometimes you want even finer grained control of the block sizes
1502 issued, not just an even split between them. This option allows you to
1503 weight various block sizes, so that you are able to define a specific
1504 amount of block sizes issued. The format for this option is::
f80dba8d
MT
1505
1506 bssplit=blocksize/percentage:blocksize/percentage
1507
6a16ece8
JA
1508 for as many block sizes as needed. So if you want to define a workload
1509 that has 50% 64k blocks, 10% 4k blocks, and 40% 32k blocks, you would
1510 write::
f80dba8d
MT
1511
1512 bssplit=4k/10:64k/50:32k/40
1513
6a16ece8
JA
1514 Ordering does not matter. If the percentage is left blank, fio will
1515 fill in the remaining values evenly. So a bssplit option like this one::
f80dba8d
MT
1516
1517 bssplit=4k/50:1k/:32k/
1518
6a16ece8
JA
1519 would have 50% 4k ios, and 25% 1k and 32k ios. The percentages always
1520 add up to 100, if bssplit is given a range that adds up to more, it
1521 will error out.
f80dba8d
MT
1522
1523 Comma-separated values may be specified for reads, writes, and trims as
1524 described in :option:`blocksize`.
1525
6a16ece8
JA
1526 If you want a workload that has 50% 2k reads and 50% 4k reads, while
1527 having 90% 4k writes and 10% 8k writes, you would specify::
f80dba8d 1528
cf04b906 1529 bssplit=2k/50:4k/50,4k/90:8k/10
f80dba8d 1530
6a16ece8
JA
1531 Fio supports defining up to 64 different weights for each data
1532 direction.
1533
f80dba8d
MT
1534.. option:: blocksize_unaligned, bs_unaligned
1535
1536 If set, fio will issue I/O units with any size within
1537 :option:`blocksize_range`, not just multiples of the minimum size. This
1538 typically won't work with direct I/O, as that normally requires sector
1539 alignment.
1540
589e88b7 1541.. option:: bs_is_seq_rand=bool
f80dba8d
MT
1542
1543 If this option is set, fio will use the normal read,write blocksize settings
1544 as sequential,random blocksize settings instead. Any random read or write
1545 will use the WRITE blocksize settings, and any sequential read or write will
1546 use the READ blocksize settings.
1547
1548.. option:: blockalign=int[,int][,int], ba=int[,int][,int]
1549
1550 Boundary to which fio will align random I/O units. Default:
1551 :option:`blocksize`. Minimum alignment is typically 512b for using direct
1552 I/O, though it usually depends on the hardware block size. This option is
1553 mutually exclusive with using a random map for files, so it will turn off
1554 that option. Comma-separated values may be specified for reads, writes, and
1555 trims as described in :option:`blocksize`.
1556
1557
1558Buffers and memory
1559~~~~~~~~~~~~~~~~~~
1560
1561.. option:: zero_buffers
1562
1563 Initialize buffers with all zeros. Default: fill buffers with random data.
1564
1565.. option:: refill_buffers
1566
1567 If this option is given, fio will refill the I/O buffers on every
72592780
SW
1568 submit. Only makes sense if :option:`zero_buffers` isn't specified,
1569 naturally. Defaults to being unset i.e., the buffer is only filled at
1570 init time and the data in it is reused when possible but if any of
1571 :option:`verify`, :option:`buffer_compress_percentage` or
1572 :option:`dedupe_percentage` are enabled then `refill_buffers` is also
1573 automatically enabled.
f80dba8d
MT
1574
1575.. option:: scramble_buffers=bool
1576
1577 If :option:`refill_buffers` is too costly and the target is using data
1578 deduplication, then setting this option will slightly modify the I/O buffer
1579 contents to defeat normal de-dupe attempts. This is not enough to defeat
1580 more clever block compression attempts, but it will stop naive dedupe of
1581 blocks. Default: true.
1582
1583.. option:: buffer_compress_percentage=int
1584
72592780
SW
1585 If this is set, then fio will attempt to provide I/O buffer content
1586 (on WRITEs) that compresses to the specified level. Fio does this by
1587 providing a mix of random data followed by fixed pattern data. The
1588 fixed pattern is either zeros, or the pattern specified by
1589 :option:`buffer_pattern`. If the `buffer_pattern` option is used, it
1590 might skew the compression ratio slightly. Setting
1591 `buffer_compress_percentage` to a value other than 100 will also
1592 enable :option:`refill_buffers` in order to reduce the likelihood that
1593 adjacent blocks are so similar that they over compress when seen
1594 together. See :option:`buffer_compress_chunk` for how to set a finer or
1595 coarser granularity for the random/fixed data region. Defaults to unset
1596 i.e., buffer data will not adhere to any compression level.
f80dba8d
MT
1597
1598.. option:: buffer_compress_chunk=int
1599
72592780
SW
1600 This setting allows fio to manage how big the random/fixed data region
1601 is when using :option:`buffer_compress_percentage`. When
1602 `buffer_compress_chunk` is set to some non-zero value smaller than the
1603 block size, fio can repeat the random/fixed region throughout the I/O
1604 buffer at the specified interval (which particularly useful when
1605 bigger block sizes are used for a job). When set to 0, fio will use a
1606 chunk size that matches the block size resulting in a single
1607 random/fixed region within the I/O buffer. Defaults to 512. When the
1608 unit is omitted, the value is interpreted in bytes.
f80dba8d
MT
1609
1610.. option:: buffer_pattern=str
1611
a1554f65
SB
1612 If set, fio will fill the I/O buffers with this pattern or with the contents
1613 of a file. If not set, the contents of I/O buffers are defined by the other
1614 options related to buffer contents. The setting can be any pattern of bytes,
1615 and can be prefixed with 0x for hex values. It may also be a string, where
1616 the string must then be wrapped with ``""``. Or it may also be a filename,
1617 where the filename must be wrapped with ``''`` in which case the file is
1618 opened and read. Note that not all the file contents will be read if that
1619 would cause the buffers to overflow. So, for example::
1620
1621 buffer_pattern='filename'
1622
1623 or::
f80dba8d
MT
1624
1625 buffer_pattern="abcd"
1626
1627 or::
1628
1629 buffer_pattern=-12
1630
1631 or::
1632
1633 buffer_pattern=0xdeadface
1634
1635 Also you can combine everything together in any order::
1636
a1554f65 1637 buffer_pattern=0xdeadface"abcd"-12'filename'
f80dba8d
MT
1638
1639.. option:: dedupe_percentage=int
1640
1641 If set, fio will generate this percentage of identical buffers when
1642 writing. These buffers will be naturally dedupable. The contents of the
1643 buffers depend on what other buffer compression settings have been set. It's
1644 possible to have the individual buffers either fully compressible, or not at
72592780
SW
1645 all -- this option only controls the distribution of unique buffers. Setting
1646 this option will also enable :option:`refill_buffers` to prevent every buffer
1647 being identical.
f80dba8d
MT
1648
1649.. option:: invalidate=bool
1650
730bd7d9
SW
1651 Invalidate the buffer/page cache parts of the files to be used prior to
1652 starting I/O if the platform and file type support it. Defaults to true.
21c1b29e
TK
1653 This will be ignored if :option:`pre_read` is also specified for the
1654 same job.
f80dba8d
MT
1655
1656.. option:: sync=bool
1657
1658 Use synchronous I/O for buffered writes. For the majority of I/O engines,
1659 this means using O_SYNC. Default: false.
1660
1661.. option:: iomem=str, mem=str
1662
1663 Fio can use various types of memory as the I/O unit buffer. The allowed
1664 values are:
1665
1666 **malloc**
1667 Use memory from :manpage:`malloc(3)` as the buffers. Default memory
1668 type.
1669
1670 **shm**
1671 Use shared memory as the buffers. Allocated through
1672 :manpage:`shmget(2)`.
1673
1674 **shmhuge**
1675 Same as shm, but use huge pages as backing.
1676
1677 **mmap**
22413915 1678 Use :manpage:`mmap(2)` to allocate buffers. May either be anonymous memory, or can
f80dba8d
MT
1679 be file backed if a filename is given after the option. The format
1680 is `mem=mmap:/path/to/file`.
1681
1682 **mmaphuge**
1683 Use a memory mapped huge file as the buffer backing. Append filename
1684 after mmaphuge, ala `mem=mmaphuge:/hugetlbfs/file`.
1685
1686 **mmapshared**
1687 Same as mmap, but use a MMAP_SHARED mapping.
1688
03553853
YR
1689 **cudamalloc**
1690 Use GPU memory as the buffers for GPUDirect RDMA benchmark.
f50fbdda 1691 The :option:`ioengine` must be `rdma`.
03553853 1692
f80dba8d
MT
1693 The area allocated is a function of the maximum allowed bs size for the job,
1694 multiplied by the I/O depth given. Note that for **shmhuge** and
1695 **mmaphuge** to work, the system must have free huge pages allocated. This
1696 can normally be checked and set by reading/writing
1697 :file:`/proc/sys/vm/nr_hugepages` on a Linux system. Fio assumes a huge page
1698 is 4MiB in size. So to calculate the number of huge pages you need for a
1699 given job file, add up the I/O depth of all jobs (normally one unless
1700 :option:`iodepth` is used) and multiply by the maximum bs set. Then divide
1701 that number by the huge page size. You can see the size of the huge pages in
1702 :file:`/proc/meminfo`. If no huge pages are allocated by having a non-zero
1703 number in `nr_hugepages`, using **mmaphuge** or **shmhuge** will fail. Also
1704 see :option:`hugepage-size`.
1705
1706 **mmaphuge** also needs to have hugetlbfs mounted and the file location
1707 should point there. So if it's mounted in :file:`/huge`, you would use
1708 `mem=mmaphuge:/huge/somefile`.
1709
f50fbdda 1710.. option:: iomem_align=int, mem_align=int
f80dba8d
MT
1711
1712 This indicates the memory alignment of the I/O memory buffers. Note that
1713 the given alignment is applied to the first I/O unit buffer, if using
1714 :option:`iodepth` the alignment of the following buffers are given by the
1715 :option:`bs` used. In other words, if using a :option:`bs` that is a
1716 multiple of the page sized in the system, all buffers will be aligned to
1717 this value. If using a :option:`bs` that is not page aligned, the alignment
1718 of subsequent I/O memory buffers is the sum of the :option:`iomem_align` and
1719 :option:`bs` used.
1720
1721.. option:: hugepage-size=int
1722
1723 Defines the size of a huge page. Must at least be equal to the system
1724 setting, see :file:`/proc/meminfo`. Defaults to 4MiB. Should probably
1725 always be a multiple of megabytes, so using ``hugepage-size=Xm`` is the
1726 preferred way to set this to avoid setting a non-pow-2 bad value.
1727
1728.. option:: lockmem=int
1729
1730 Pin the specified amount of memory with :manpage:`mlock(2)`. Can be used to
1731 simulate a smaller amount of memory. The amount specified is per worker.
1732
1733
1734I/O size
1735~~~~~~~~
1736
1737.. option:: size=int
1738
79591fa9
TK
1739 The total size of file I/O for each thread of this job. Fio will run until
1740 this many bytes has been transferred, unless runtime is limited by other options
1741 (such as :option:`runtime`, for instance, or increased/decreased by :option:`io_size`).
1742 Fio will divide this size between the available files determined by options
1743 such as :option:`nrfiles`, :option:`filename`, unless :option:`filesize` is
1744 specified by the job. If the result of division happens to be 0, the size is
c4aa2d08 1745 set to the physical size of the given files or devices if they exist.
79591fa9 1746 If this option is not specified, fio will use the full size of the given
f80dba8d
MT
1747 files or devices. If the files do not exist, size must be given. It is also
1748 possible to give size as a percentage between 1 and 100. If ``size=20%`` is
1749 given, fio will use 20% of the full size of the given files or devices.
9d25d068
SW
1750 Can be combined with :option:`offset` to constrain the start and end range
1751 that I/O will be done within.
f80dba8d
MT
1752
1753.. option:: io_size=int, io_limit=int
1754
1755 Normally fio operates within the region set by :option:`size`, which means
1756 that the :option:`size` option sets both the region and size of I/O to be
1757 performed. Sometimes that is not what you want. With this option, it is
1758 possible to define just the amount of I/O that fio should do. For instance,
1759 if :option:`size` is set to 20GiB and :option:`io_size` is set to 5GiB, fio
1760 will perform I/O within the first 20GiB but exit when 5GiB have been
1761 done. The opposite is also possible -- if :option:`size` is set to 20GiB,
1762 and :option:`io_size` is set to 40GiB, then fio will do 40GiB of I/O within
1763 the 0..20GiB region.
1764
7fdd97ca 1765.. option:: filesize=irange(int)
f80dba8d
MT
1766
1767 Individual file sizes. May be a range, in which case fio will select sizes
1768 for files at random within the given range and limited to :option:`size` in
1769 total (if that is given). If not given, each created file is the same size.
79591fa9
TK
1770 This option overrides :option:`size` in terms of file size, which means
1771 this value is used as a fixed size or possible range of each file.
f80dba8d
MT
1772
1773.. option:: file_append=bool
1774
1775 Perform I/O after the end of the file. Normally fio will operate within the
1776 size of a file. If this option is set, then fio will append to the file
1777 instead. This has identical behavior to setting :option:`offset` to the size
1778 of a file. This option is ignored on non-regular files.
1779
1780.. option:: fill_device=bool, fill_fs=bool
1781
1782 Sets size to something really large and waits for ENOSPC (no space left on
1783 device) as the terminating condition. Only makes sense with sequential
1784 write. For a read workload, the mount point will be filled first then I/O
1785 started on the result. This option doesn't make sense if operating on a raw
1786 device node, since the size of that is already known by the file system.
1787 Additionally, writing beyond end-of-device will not return ENOSPC there.
1788
1789
1790I/O engine
1791~~~~~~~~~~
1792
1793.. option:: ioengine=str
1794
1795 Defines how the job issues I/O to the file. The following types are defined:
1796
1797 **sync**
1798 Basic :manpage:`read(2)` or :manpage:`write(2)`
1799 I/O. :manpage:`lseek(2)` is used to position the I/O location.
54227e6b 1800 See :option:`fsync` and :option:`fdatasync` for syncing write I/Os.
f80dba8d
MT
1801
1802 **psync**
1803 Basic :manpage:`pread(2)` or :manpage:`pwrite(2)` I/O. Default on
1804 all supported operating systems except for Windows.
1805
1806 **vsync**
1807 Basic :manpage:`readv(2)` or :manpage:`writev(2)` I/O. Will emulate
c60ebc45 1808 queuing by coalescing adjacent I/Os into a single submission.
f80dba8d
MT
1809
1810 **pvsync**
1811 Basic :manpage:`preadv(2)` or :manpage:`pwritev(2)` I/O.
1812
1813 **pvsync2**
1814 Basic :manpage:`preadv2(2)` or :manpage:`pwritev2(2)` I/O.
1815
029b42ac
JA
1816 **io_uring**
1817 Fast Linux native asynchronous I/O. Supports async IO
1818 for both direct and buffered IO.
1819 This engine defines engine specific options.
1820
f80dba8d
MT
1821 **libaio**
1822 Linux native asynchronous I/O. Note that Linux may only support
22413915 1823 queued behavior with non-buffered I/O (set ``direct=1`` or
f80dba8d
MT
1824 ``buffered=0``).
1825 This engine defines engine specific options.
1826
1827 **posixaio**
1828 POSIX asynchronous I/O using :manpage:`aio_read(3)` and
1829 :manpage:`aio_write(3)`.
1830
1831 **solarisaio**
1832 Solaris native asynchronous I/O.
1833
1834 **windowsaio**
1835 Windows native asynchronous I/O. Default on Windows.
1836
1837 **mmap**
1838 File is memory mapped with :manpage:`mmap(2)` and data copied
1839 to/from using :manpage:`memcpy(3)`.
1840
1841 **splice**
1842 :manpage:`splice(2)` is used to transfer the data and
1843 :manpage:`vmsplice(2)` to transfer data from user space to the
1844 kernel.
1845
1846 **sg**
1847 SCSI generic sg v3 I/O. May either be synchronous using the SG_IO
1848 ioctl, or if the target is an sg character device we use
1849 :manpage:`read(2)` and :manpage:`write(2)` for asynchronous
f50fbdda 1850 I/O. Requires :option:`filename` option to specify either block or
3740cfc8 1851 character devices. This engine supports trim operations.
52b81b7c 1852 The sg engine includes engine specific options.
f80dba8d
MT
1853
1854 **null**
1855 Doesn't transfer any data, just pretends to. This is mainly used to
1856 exercise fio itself and for debugging/testing purposes.
1857
1858 **net**
1859 Transfer over the network to given ``host:port``. Depending on the
1860 :option:`protocol` used, the :option:`hostname`, :option:`port`,
1861 :option:`listen` and :option:`filename` options are used to specify
1862 what sort of connection to make, while the :option:`protocol` option
1863 determines which protocol will be used. This engine defines engine
1864 specific options.
1865
1866 **netsplice**
1867 Like **net**, but uses :manpage:`splice(2)` and
1868 :manpage:`vmsplice(2)` to map data and send/receive.
1869 This engine defines engine specific options.
1870
1871 **cpuio**
1872 Doesn't transfer any data, but burns CPU cycles according to the
1873 :option:`cpuload` and :option:`cpuchunks` options. Setting
9207a0cb 1874 :option:`cpuload`\=85 will cause that job to do nothing but burn 85%
71aa48eb 1875 of the CPU. In case of SMP machines, use :option:`numjobs`\=<nr_of_cpu>
f50fbdda 1876 to get desired CPU usage, as the cpuload only loads a
f80dba8d
MT
1877 single CPU at the desired rate. A job never finishes unless there is
1878 at least one non-cpuio job.
1879
1880 **guasi**
804c0839 1881 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall
f80dba8d
MT
1882 Interface approach to async I/O. See
1883
1884 http://www.xmailserver.org/guasi-lib.html
1885
1886 for more info on GUASI.
1887
1888 **rdma**
1889 The RDMA I/O engine supports both RDMA memory semantics
1890 (RDMA_WRITE/RDMA_READ) and channel semantics (Send/Recv) for the
609ac152
SB
1891 InfiniBand, RoCE and iWARP protocols. This engine defines engine
1892 specific options.
f80dba8d
MT
1893
1894 **falloc**
1895 I/O engine that does regular fallocate to simulate data transfer as
1896 fio ioengine.
1897
1898 DDIR_READ
1899 does fallocate(,mode = FALLOC_FL_KEEP_SIZE,).
1900
1901 DDIR_WRITE
1902 does fallocate(,mode = 0).
1903
1904 DDIR_TRIM
1905 does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE).
1906
761cd093
SW
1907 **ftruncate**
1908 I/O engine that sends :manpage:`ftruncate(2)` operations in response
1909 to write (DDIR_WRITE) events. Each ftruncate issued sets the file's
f50fbdda 1910 size to the current block offset. :option:`blocksize` is ignored.
761cd093 1911
f80dba8d
MT
1912 **e4defrag**
1913 I/O engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate
1914 defragment activity in request to DDIR_WRITE event.
1915
f3f96717
IF
1916 **rados**
1917 I/O engine supporting direct access to Ceph Reliable Autonomic
1918 Distributed Object Store (RADOS) via librados. This ioengine
1919 defines engine specific options.
1920
f80dba8d
MT
1921 **rbd**
1922 I/O engine supporting direct access to Ceph Rados Block Devices
1923 (RBD) via librbd without the need to use the kernel rbd driver. This
1924 ioengine defines engine specific options.
1925
c2f6a13d
LMB
1926 **http**
1927 I/O engine supporting GET/PUT requests over HTTP(S) with libcurl to
1928 a WebDAV or S3 endpoint. This ioengine defines engine specific options.
1929
1930 This engine only supports direct IO of iodepth=1; you need to scale this
1931 via numjobs. blocksize defines the size of the objects to be created.
1932
1933 TRIM is translated to object deletion.
1934
f80dba8d 1935 **gfapi**
ac8ca2af
SW
1936 Using GlusterFS libgfapi sync interface to direct access to
1937 GlusterFS volumes without having to go through FUSE. This ioengine
f80dba8d
MT
1938 defines engine specific options.
1939
1940 **gfapi_async**
ac8ca2af
SW
1941 Using GlusterFS libgfapi async interface to direct access to
1942 GlusterFS volumes without having to go through FUSE. This ioengine
f80dba8d
MT
1943 defines engine specific options.
1944
1945 **libhdfs**
f50fbdda 1946 Read and write through Hadoop (HDFS). The :option:`filename` option
f80dba8d
MT
1947 is used to specify host,port of the hdfs name-node to connect. This
1948 engine interprets offsets a little differently. In HDFS, files once
e25c0c91
SW
1949 created cannot be modified so random writes are not possible. To
1950 imitate this the libhdfs engine expects a bunch of small files to be
1951 created over HDFS and will randomly pick a file from them
1952 based on the offset generated by fio backend (see the example
f80dba8d 1953 job file to create such files, use ``rw=write`` option). Please
e25c0c91
SW
1954 note, it may be necessary to set environment variables to work
1955 with HDFS/libhdfs properly. Each job uses its own connection to
f80dba8d
MT
1956 HDFS.
1957
1958 **mtd**
1959 Read, write and erase an MTD character device (e.g.,
1960 :file:`/dev/mtd0`). Discards are treated as erases. Depending on the
1961 underlying device type, the I/O may have to go in a certain pattern,
1962 e.g., on NAND, writing sequentially to erase blocks and discarding
c298ee71 1963 before overwriting. The `trimwrite` mode works well for this
f80dba8d
MT
1964 constraint.
1965
1966 **pmemblk**
1967 Read and write using filesystem DAX to a file on a filesystem
363a5f65 1968 mounted with DAX on a persistent memory device through the PMDK
f80dba8d
MT
1969 libpmemblk library.
1970
1971 **dev-dax**
1972 Read and write using device DAX to a persistent memory device (e.g.,
363a5f65 1973 /dev/dax0.0) through the PMDK libpmem library.
f80dba8d
MT
1974
1975 **external**
1976 Prefix to specify loading an external I/O engine object file. Append
c60ebc45 1977 the engine filename, e.g. ``ioengine=external:/tmp/foo.o`` to load
d243fd6d
TK
1978 ioengine :file:`foo.o` in :file:`/tmp`. The path can be either
1979 absolute or relative. See :file:`engines/skeleton_external.c` for
1980 details of writing an external I/O engine.
f80dba8d 1981
1216cc5a 1982 **filecreate**
b71968b1 1983 Simply create the files and do no I/O to them. You still need to
1216cc5a 1984 set `filesize` so that all the accounting still occurs, but no
b71968b1 1985 actual I/O will be done other than creating the file.
f80dba8d 1986
ae0db592
TI
1987 **libpmem**
1988 Read and write using mmap I/O to a file on a filesystem
363a5f65 1989 mounted with DAX on a persistent memory device through the PMDK
ae0db592
TI
1990 libpmem library.
1991
a40e7a59
GB
1992 **ime_psync**
1993 Synchronous read and write using DDN's Infinite Memory Engine (IME).
1994 This engine is very basic and issues calls to IME whenever an IO is
1995 queued.
1996
1997 **ime_psyncv**
1998 Synchronous read and write using DDN's Infinite Memory Engine (IME).
1999 This engine uses iovecs and will try to stack as much IOs as possible
2000 (if the IOs are "contiguous" and the IO depth is not exceeded)
2001 before issuing a call to IME.
2002
2003 **ime_aio**
2004 Asynchronous read and write using DDN's Infinite Memory Engine (IME).
2005 This engine will try to stack as much IOs as possible by creating
2006 requests for IME. FIO will then decide when to commit these requests.
247ef2aa
KZ
2007 **libiscsi**
2008 Read and write iscsi lun with libiscsi.
d643a1e2 2009 **nbd**
f2d6de5d 2010 Read and write a Network Block Device (NBD).
a40e7a59 2011
f80dba8d
MT
2012I/O engine specific parameters
2013~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2014
2015In addition, there are some parameters which are only valid when a specific
f50fbdda
TK
2016:option:`ioengine` is in use. These are used identically to normal parameters,
2017with the caveat that when used on the command line, they must come after the
f80dba8d
MT
2018:option:`ioengine` that defines them is selected.
2019
029b42ac
JA
2020.. option:: hipri : [io_uring]
2021
2022 If this option is set, fio will attempt to use polled IO completions.
2023 Normal IO completions generate interrupts to signal the completion of
2024 IO, polled completions do not. Hence they are require active reaping
2025 by the application. The benefits are more efficient IO for high IOPS
2026 scenarios, and lower latencies for low queue depth IO.
2027
2028.. option:: fixedbufs : [io_uring]
2029
2030 If fio is asked to do direct IO, then Linux will map pages for each
2031 IO call, and release them when IO is done. If this option is set, the
2032 pages are pre-mapped before IO is started. This eliminates the need to
2033 map and release for each IO. This is more efficient, and reduces the
2034 IO latency as well.
2035
5ffd5626
JA
2036.. option:: registerfiles : [io_uring]
2037 With this option, fio registers the set of files being used with the
2038 kernel. This avoids the overhead of managing file counts in the kernel,
2039 making the submission and completion part more lightweight. Required
2040 for the below :option:`sqthread_poll` option.
2041
029b42ac
JA
2042.. option:: sqthread_poll : [io_uring]
2043
2044 Normally fio will submit IO by issuing a system call to notify the
2045 kernel of available items in the SQ ring. If this option is set, the
2046 act of submitting IO will be done by a polling thread in the kernel.
2047 This frees up cycles for fio, at the cost of using more CPU in the
2048 system.
2049
2050.. option:: sqthread_poll_cpu : [io_uring]
2051
2052 When :option:`sqthread_poll` is set, this option provides a way to
2053 define which CPU should be used for the polling thread.
2054
f80dba8d
MT
2055.. option:: userspace_reap : [libaio]
2056
2057 Normally, with the libaio engine in use, fio will use the
2058 :manpage:`io_getevents(2)` system call to reap newly returned events. With
2059 this flag turned on, the AIO ring will be read directly from user-space to
2060 reap events. The reaping mode is only enabled when polling for a minimum of
c60ebc45 2061 0 events (e.g. when :option:`iodepth_batch_complete` `=0`).
f80dba8d 2062
9d25d068 2063.. option:: hipri : [pvsync2]
f80dba8d
MT
2064
2065 Set RWF_HIPRI on I/O, indicating to the kernel that it's of higher priority
2066 than normal.
2067
a0679ce5
SB
2068.. option:: hipri_percentage : [pvsync2]
2069
f50fbdda 2070 When hipri is set this determines the probability of a pvsync2 I/O being high
a0679ce5
SB
2071 priority. The default is 100%.
2072
f80dba8d
MT
2073.. option:: cpuload=int : [cpuio]
2074
da19cdb4
TK
2075 Attempt to use the specified percentage of CPU cycles. This is a mandatory
2076 option when using cpuio I/O engine.
f80dba8d
MT
2077
2078.. option:: cpuchunks=int : [cpuio]
2079
2080 Split the load into cycles of the given time. In microseconds.
2081
2082.. option:: exit_on_io_done=bool : [cpuio]
2083
2084 Detect when I/O threads are done, then exit.
2085
f80dba8d
MT
2086.. option:: namenode=str : [libhdfs]
2087
22413915 2088 The hostname or IP address of a HDFS cluster namenode to contact.
f80dba8d
MT
2089
2090.. option:: port=int
2091
f50fbdda
TK
2092 [libhdfs]
2093
2094 The listening port of the HFDS cluster namenode.
2095
f80dba8d
MT
2096 [netsplice], [net]
2097
2098 The TCP or UDP port to bind to or connect to. If this is used with
2099 :option:`numjobs` to spawn multiple instances of the same job type, then
2100 this will be the starting port number since fio will use a range of
2101 ports.
2102
609ac152
SB
2103 [rdma]
2104
2105 The port to use for RDMA-CM communication. This should be the same value
2106 on the client and the server side.
2107
2108.. option:: hostname=str : [netsplice] [net] [rdma]
f80dba8d 2109
609ac152
SB
2110 The hostname or IP address to use for TCP, UDP or RDMA-CM based I/O. If the job
2111 is a TCP listener or UDP reader, the hostname is not used and must be omitted
f50fbdda 2112 unless it is a valid UDP multicast address.
f80dba8d
MT
2113
2114.. option:: interface=str : [netsplice] [net]
2115
2116 The IP address of the network interface used to send or receive UDP
2117 multicast.
2118
2119.. option:: ttl=int : [netsplice] [net]
2120
2121 Time-to-live value for outgoing UDP multicast packets. Default: 1.
2122
2123.. option:: nodelay=bool : [netsplice] [net]
2124
2125 Set TCP_NODELAY on TCP connections.
2126
f50fbdda 2127.. option:: protocol=str, proto=str : [netsplice] [net]
f80dba8d
MT
2128
2129 The network protocol to use. Accepted values are:
2130
2131 **tcp**
2132 Transmission control protocol.
2133 **tcpv6**
2134 Transmission control protocol V6.
2135 **udp**
2136 User datagram protocol.
2137 **udpv6**
2138 User datagram protocol V6.
2139 **unix**
2140 UNIX domain socket.
2141
2142 When the protocol is TCP or UDP, the port must also be given, as well as the
2143 hostname if the job is a TCP listener or UDP reader. For unix sockets, the
f50fbdda 2144 normal :option:`filename` option should be used and the port is invalid.
f80dba8d 2145
e9184ec1 2146.. option:: listen : [netsplice] [net]
f80dba8d
MT
2147
2148 For TCP network connections, tell fio to listen for incoming connections
2149 rather than initiating an outgoing connection. The :option:`hostname` must
2150 be omitted if this option is used.
2151
e9184ec1 2152.. option:: pingpong : [netsplice] [net]
f80dba8d
MT
2153
2154 Normally a network writer will just continue writing data, and a network
2155 reader will just consume packages. If ``pingpong=1`` is set, a writer will
2156 send its normal payload to the reader, then wait for the reader to send the
2157 same payload back. This allows fio to measure network latencies. The
2158 submission and completion latencies then measure local time spent sending or
2159 receiving, and the completion latency measures how long it took for the
2160 other end to receive and send back. For UDP multicast traffic
2161 ``pingpong=1`` should only be set for a single reader when multiple readers
2162 are listening to the same address.
2163
e9184ec1 2164.. option:: window_size : [netsplice] [net]
f80dba8d
MT
2165
2166 Set the desired socket buffer size for the connection.
2167
e9184ec1 2168.. option:: mss : [netsplice] [net]
f80dba8d
MT
2169
2170 Set the TCP maximum segment size (TCP_MAXSEG).
2171
2172.. option:: donorname=str : [e4defrag]
2173
730bd7d9 2174 File will be used as a block donor (swap extents between files).
f80dba8d
MT
2175
2176.. option:: inplace=int : [e4defrag]
2177
2178 Configure donor file blocks allocation strategy:
2179
2180 **0**
2181 Default. Preallocate donor's file on init.
2182 **1**
2b455dbf 2183 Allocate space immediately inside defragment event, and free right
f80dba8d
MT
2184 after event.
2185
f3f96717 2186.. option:: clustername=str : [rbd,rados]
f80dba8d
MT
2187
2188 Specifies the name of the Ceph cluster.
2189
2190.. option:: rbdname=str : [rbd]
2191
2192 Specifies the name of the RBD.
2193
f3f96717 2194.. option:: pool=str : [rbd,rados]
f80dba8d 2195
f3f96717 2196 Specifies the name of the Ceph pool containing RBD or RADOS data.
f80dba8d 2197
f3f96717 2198.. option:: clientname=str : [rbd,rados]
f80dba8d
MT
2199
2200 Specifies the username (without the 'client.' prefix) used to access the
2201 Ceph cluster. If the *clustername* is specified, the *clientname* shall be
2202 the full *type.id* string. If no type. prefix is given, fio will add
2203 'client.' by default.
2204
f3f96717
IF
2205.. option:: busy_poll=bool : [rbd,rados]
2206
2207 Poll store instead of waiting for completion. Usually this provides better
2208 throughput at cost of higher(up to 100%) CPU utilization.
2209
f80dba8d
MT
2210.. option:: skip_bad=bool : [mtd]
2211
2212 Skip operations against known bad blocks.
2213
2214.. option:: hdfsdirectory : [libhdfs]
2215
2216 libhdfs will create chunk in this HDFS directory.
2217
2218.. option:: chunk_size : [libhdfs]
2219
2b455dbf 2220 The size of the chunk to use for each file.
f80dba8d 2221
609ac152
SB
2222.. option:: verb=str : [rdma]
2223
2224 The RDMA verb to use on this side of the RDMA ioengine connection. Valid
2225 values are write, read, send and recv. These correspond to the equivalent
2226 RDMA verbs (e.g. write = rdma_write etc.). Note that this only needs to be
2227 specified on the client side of the connection. See the examples folder.
2228
2229.. option:: bindname=str : [rdma]
2230
2231 The name to use to bind the local RDMA-CM connection to a local RDMA device.
2232 This could be a hostname or an IPv4 or IPv6 address. On the server side this
2233 will be passed into the rdma_bind_addr() function and on the client site it
2234 will be used in the rdma_resolve_add() function. This can be useful when
2235 multiple paths exist between the client and the server or in certain loopback
2236 configurations.
f80dba8d 2237
52b81b7c
KD
2238.. option:: readfua=bool : [sg]
2239
2240 With readfua option set to 1, read operations include
2241 the force unit access (fua) flag. Default is 0.
2242
2243.. option:: writefua=bool : [sg]
2244
2245 With writefua option set to 1, write operations include
2246 the force unit access (fua) flag. Default is 0.
2247
2c3a9150 2248.. option:: sg_write_mode=str : [sg]
3740cfc8 2249
2c3a9150
VF
2250 Specify the type of write commands to issue. This option can take three values:
2251
2252 **write**
2253 This is the default where write opcodes are issued as usual.
2254 **verify**
2255 Issue WRITE AND VERIFY commands. The BYTCHK bit is set to 0. This
2256 directs the device to carry out a medium verification with no data
2257 comparison. The writefua option is ignored with this selection.
2258 **same**
2259 Issue WRITE SAME commands. This transfers a single block to the device
2260 and writes this same block of data to a contiguous sequence of LBAs
2261 beginning at the specified offset. fio's block size parameter specifies
2262 the amount of data written with each command. However, the amount of data
2263 actually transferred to the device is equal to the device's block
2264 (sector) size. For a device with 512 byte sectors, blocksize=8k will
2265 write 16 sectors with each command. fio will still generate 8k of data
2266 for each command but only the first 512 bytes will be used and
2267 transferred to the device. The writefua option is ignored with this
2268 selection.
52b81b7c 2269
c2f6a13d
LMB
2270.. option:: http_host=str : [http]
2271
2272 Hostname to connect to. For S3, this could be the bucket hostname.
2273 Default is **localhost**
2274
2275.. option:: http_user=str : [http]
2276
2277 Username for HTTP authentication.
2278
2279.. option:: http_pass=str : [http]
2280
2281 Password for HTTP authentication.
2282
09fd2966 2283.. option:: https=str : [http]
c2f6a13d 2284
09fd2966
LMB
2285 Enable HTTPS instead of http. *on* enables HTTPS; *insecure*
2286 will enable HTTPS, but disable SSL peer verification (use with
2287 caution!). Default is **off**
c2f6a13d 2288
09fd2966 2289.. option:: http_mode=str : [http]
c2f6a13d 2290
09fd2966
LMB
2291 Which HTTP access mode to use: *webdav*, *swift*, or *s3*.
2292 Default is **webdav**
c2f6a13d
LMB
2293
2294.. option:: http_s3_region=str : [http]
2295
2296 The S3 region/zone string.
2297 Default is **us-east-1**
2298
2299.. option:: http_s3_key=str : [http]
2300
2301 The S3 secret key.
2302
2303.. option:: http_s3_keyid=str : [http]
2304
2305 The S3 key/access id.
2306
09fd2966
LMB
2307.. option:: http_swift_auth_token=str : [http]
2308
2309 The Swift auth token. See the example configuration file on how
2310 to retrieve this.
2311
c2f6a13d
LMB
2312.. option:: http_verbose=int : [http]
2313
2314 Enable verbose requests from libcurl. Useful for debugging. 1
2315 turns on verbose logging from libcurl, 2 additionally enables
2316 HTTP IO tracing. Default is **0**
2317
f2d6de5d
RJ
2318.. option:: uri=str : [nbd]
2319
2320 Specify the NBD URI of the server to test. The string
2321 is a standard NBD URI
2322 (see https://github.com/NetworkBlockDevice/nbd/tree/master/doc).
2323 Example URIs: nbd://localhost:10809
2324 nbd+unix:///?socket=/tmp/socket
2325 nbds://tlshost/exportname
2326
f80dba8d
MT
2327I/O depth
2328~~~~~~~~~
2329
2330.. option:: iodepth=int
2331
2332 Number of I/O units to keep in flight against the file. Note that
2333 increasing *iodepth* beyond 1 will not affect synchronous ioengines (except
c60ebc45 2334 for small degrees when :option:`verify_async` is in use). Even async
f80dba8d
MT
2335 engines may impose OS restrictions causing the desired depth not to be
2336 achieved. This may happen on Linux when using libaio and not setting
9207a0cb 2337 :option:`direct`\=1, since buffered I/O is not async on that OS. Keep an
f80dba8d
MT
2338 eye on the I/O depth distribution in the fio output to verify that the
2339 achieved depth is as expected. Default: 1.
2340
2341.. option:: iodepth_batch_submit=int, iodepth_batch=int
2342
2343 This defines how many pieces of I/O to submit at once. It defaults to 1
2344 which means that we submit each I/O as soon as it is available, but can be
2345 raised to submit bigger batches of I/O at the time. If it is set to 0 the
2346 :option:`iodepth` value will be used.
2347
2348.. option:: iodepth_batch_complete_min=int, iodepth_batch_complete=int
2349
2350 This defines how many pieces of I/O to retrieve at once. It defaults to 1
2351 which means that we'll ask for a minimum of 1 I/O in the retrieval process
2352 from the kernel. The I/O retrieval will go on until we hit the limit set by
2353 :option:`iodepth_low`. If this variable is set to 0, then fio will always
2354 check for completed events before queuing more I/O. This helps reduce I/O
2355 latency, at the cost of more retrieval system calls.
2356
2357.. option:: iodepth_batch_complete_max=int
2358
2359 This defines maximum pieces of I/O to retrieve at once. This variable should
9207a0cb 2360 be used along with :option:`iodepth_batch_complete_min`\=int variable,
f80dba8d 2361 specifying the range of min and max amount of I/O which should be
730bd7d9 2362 retrieved. By default it is equal to the :option:`iodepth_batch_complete_min`
f80dba8d
MT
2363 value.
2364
2365 Example #1::
2366
2367 iodepth_batch_complete_min=1
2368 iodepth_batch_complete_max=<iodepth>
2369
2370 which means that we will retrieve at least 1 I/O and up to the whole
2371 submitted queue depth. If none of I/O has been completed yet, we will wait.
2372
2373 Example #2::
2374
2375 iodepth_batch_complete_min=0
2376 iodepth_batch_complete_max=<iodepth>
2377
2378 which means that we can retrieve up to the whole submitted queue depth, but
2379 if none of I/O has been completed yet, we will NOT wait and immediately exit
2380 the system call. In this example we simply do polling.
2381
2382.. option:: iodepth_low=int
2383
2384 The low water mark indicating when to start filling the queue
2385 again. Defaults to the same as :option:`iodepth`, meaning that fio will
2386 attempt to keep the queue full at all times. If :option:`iodepth` is set to
c60ebc45 2387 e.g. 16 and *iodepth_low* is set to 4, then after fio has filled the queue of
f80dba8d
MT
2388 16 requests, it will let the depth drain down to 4 before starting to fill
2389 it again.
2390
997b5680
SW
2391.. option:: serialize_overlap=bool
2392
2393 Serialize in-flight I/Os that might otherwise cause or suffer from data races.
2394 When two or more I/Os are submitted simultaneously, there is no guarantee that
2395 the I/Os will be processed or completed in the submitted order. Further, if
2396 two or more of those I/Os are writes, any overlapping region between them can
2397 become indeterminate/undefined on certain storage. These issues can cause
2398 verification to fail erratically when at least one of the racing I/Os is
2399 changing data and the overlapping region has a non-zero size. Setting
2400 ``serialize_overlap`` tells fio to avoid provoking this behavior by explicitly
2401 serializing in-flight I/Os that have a non-zero overlap. Note that setting
ee21ebee 2402 this option can reduce both performance and the :option:`iodepth` achieved.
3d6a6f04
VF
2403
2404 This option only applies to I/Os issued for a single job except when it is
a02ec45a 2405 enabled along with :option:`io_submit_mode`\=offload. In offload mode, fio
3d6a6f04 2406 will check for overlap among all I/Os submitted by offload jobs with :option:`serialize_overlap`
307f2246 2407 enabled.
3d6a6f04
VF
2408
2409 Default: false.
997b5680 2410
f80dba8d
MT
2411.. option:: io_submit_mode=str
2412
2413 This option controls how fio submits the I/O to the I/O engine. The default
2414 is `inline`, which means that the fio job threads submit and reap I/O
2415 directly. If set to `offload`, the job threads will offload I/O submission
2416 to a dedicated pool of I/O threads. This requires some coordination and thus
2417 has a bit of extra overhead, especially for lower queue depth I/O where it
2418 can increase latencies. The benefit is that fio can manage submission rates
2419 independently of the device completion rates. This avoids skewed latency
730bd7d9 2420 reporting if I/O gets backed up on the device side (the coordinated omission
f80dba8d
MT
2421 problem).
2422
2423
2424I/O rate
2425~~~~~~~~
2426
a881438b 2427.. option:: thinktime=time
f80dba8d 2428
f75ede1d
SW
2429 Stall the job for the specified period of time after an I/O has completed before issuing the
2430 next. May be used to simulate processing being done by an application.
947e0fe0 2431 When the unit is omitted, the value is interpreted in microseconds. See
f80dba8d
MT
2432 :option:`thinktime_blocks` and :option:`thinktime_spin`.
2433
a881438b 2434.. option:: thinktime_spin=time
f80dba8d
MT
2435
2436 Only valid if :option:`thinktime` is set - pretend to spend CPU time doing
2437 something with the data received, before falling back to sleeping for the
f75ede1d 2438 rest of the period specified by :option:`thinktime`. When the unit is
947e0fe0 2439 omitted, the value is interpreted in microseconds.
f80dba8d
MT
2440
2441.. option:: thinktime_blocks=int
2442
2443 Only valid if :option:`thinktime` is set - control how many blocks to issue,
f50fbdda
TK
2444 before waiting :option:`thinktime` usecs. If not set, defaults to 1 which will make
2445 fio wait :option:`thinktime` usecs after every block. This effectively makes any
f80dba8d 2446 queue depth setting redundant, since no more than 1 I/O will be queued
f50fbdda 2447 before we have to complete it and do our :option:`thinktime`. In other words, this
f80dba8d 2448 setting effectively caps the queue depth if the latter is larger.
71bfa161 2449
f80dba8d 2450.. option:: rate=int[,int][,int]
71bfa161 2451
f80dba8d
MT
2452 Cap the bandwidth used by this job. The number is in bytes/sec, the normal
2453 suffix rules apply. Comma-separated values may be specified for reads,
2454 writes, and trims as described in :option:`blocksize`.
71bfa161 2455
b25b3464
SW
2456 For example, using `rate=1m,500k` would limit reads to 1MiB/sec and writes to
2457 500KiB/sec. Capping only reads or writes can be done with `rate=,500k` or
2458 `rate=500k,` where the former will only limit writes (to 500KiB/sec) and the
2459 latter will only limit reads.
2460
f80dba8d 2461.. option:: rate_min=int[,int][,int]
71bfa161 2462
f80dba8d
MT
2463 Tell fio to do whatever it can to maintain at least this bandwidth. Failing
2464 to meet this requirement will cause the job to exit. Comma-separated values
2465 may be specified for reads, writes, and trims as described in
2466 :option:`blocksize`.
71bfa161 2467
f80dba8d 2468.. option:: rate_iops=int[,int][,int]
71bfa161 2469
f80dba8d
MT
2470 Cap the bandwidth to this number of IOPS. Basically the same as
2471 :option:`rate`, just specified independently of bandwidth. If the job is
2472 given a block size range instead of a fixed value, the smallest block size
2473 is used as the metric. Comma-separated values may be specified for reads,
2474 writes, and trims as described in :option:`blocksize`.
71bfa161 2475
f80dba8d 2476.. option:: rate_iops_min=int[,int][,int]
71bfa161 2477
f80dba8d
MT
2478 If fio doesn't meet this rate of I/O, it will cause the job to exit.
2479 Comma-separated values may be specified for reads, writes, and trims as
2480 described in :option:`blocksize`.
71bfa161 2481
f80dba8d 2482.. option:: rate_process=str
66c098b8 2483
f80dba8d
MT
2484 This option controls how fio manages rated I/O submissions. The default is
2485 `linear`, which submits I/O in a linear fashion with fixed delays between
c60ebc45 2486 I/Os that gets adjusted based on I/O completion rates. If this is set to
f80dba8d
MT
2487 `poisson`, fio will submit I/O based on a more real world random request
2488 flow, known as the Poisson process
2489 (https://en.wikipedia.org/wiki/Poisson_point_process). The lambda will be
2490 10^6 / IOPS for the given workload.
71bfa161 2491
1a9bf814
JA
2492.. option:: rate_ignore_thinktime=bool
2493
2494 By default, fio will attempt to catch up to the specified rate setting,
2495 if any kind of thinktime setting was used. If this option is set, then
2496 fio will ignore the thinktime and continue doing IO at the specified
2497 rate, instead of entering a catch-up mode after thinktime is done.
2498
71bfa161 2499
f80dba8d
MT
2500I/O latency
2501~~~~~~~~~~~
71bfa161 2502
a881438b 2503.. option:: latency_target=time
71bfa161 2504
f80dba8d 2505 If set, fio will attempt to find the max performance point that the given
f75ede1d 2506 workload will run at while maintaining a latency below this target. When
947e0fe0 2507 the unit is omitted, the value is interpreted in microseconds. See
f75ede1d 2508 :option:`latency_window` and :option:`latency_percentile`.
71bfa161 2509
a881438b 2510.. option:: latency_window=time
71bfa161 2511
f80dba8d 2512 Used with :option:`latency_target` to specify the sample window that the job
f75ede1d 2513 is run at varying queue depths to test the performance. When the unit is
947e0fe0 2514 omitted, the value is interpreted in microseconds.
b4692828 2515
f80dba8d 2516.. option:: latency_percentile=float
71bfa161 2517
c60ebc45 2518 The percentage of I/Os that must fall within the criteria specified by
f80dba8d 2519 :option:`latency_target` and :option:`latency_window`. If not set, this
c60ebc45 2520 defaults to 100.0, meaning that all I/Os must be equal or below to the value
f80dba8d 2521 set by :option:`latency_target`.
71bfa161 2522
a881438b 2523.. option:: max_latency=time
71bfa161 2524
f75ede1d 2525 If set, fio will exit the job with an ETIMEDOUT error if it exceeds this
947e0fe0 2526 maximum latency. When the unit is omitted, the value is interpreted in
f75ede1d 2527 microseconds.
71bfa161 2528
f80dba8d 2529.. option:: rate_cycle=int
71bfa161 2530
f80dba8d 2531 Average bandwidth for :option:`rate` and :option:`rate_min` over this number
a47b697c 2532 of milliseconds. Defaults to 1000.
71bfa161 2533
71bfa161 2534
f80dba8d
MT
2535I/O replay
2536~~~~~~~~~~
71bfa161 2537
f80dba8d 2538.. option:: write_iolog=str
c2b1e753 2539
f80dba8d
MT
2540 Write the issued I/O patterns to the specified file. See
2541 :option:`read_iolog`. Specify a separate file for each job, otherwise the
2542 iologs will be interspersed and the file may be corrupt.
c2b1e753 2543
f80dba8d 2544.. option:: read_iolog=str
71bfa161 2545
22413915 2546 Open an iolog with the specified filename and replay the I/O patterns it
f80dba8d
MT
2547 contains. This can be used to store a workload and replay it sometime
2548 later. The iolog given may also be a blktrace binary file, which allows fio
2549 to replay a workload captured by :command:`blktrace`. See
2550 :manpage:`blktrace(8)` for how to capture such logging data. For blktrace
2551 replay, the file needs to be turned into a blkparse binary data file first
2552 (``blkparse <device> -o /dev/null -d file_for_fio.bin``).
78439a18
JA
2553 You can specify a number of files by separating the names with a ':'
2554 character. See the :option:`filename` option for information on how to
2555 escape ':' and '\' characters within the file names. These files will
2556 be sequentially assigned to job clones created by :option:`numjobs`.
71bfa161 2557
77be374d
AK
2558.. option:: read_iolog_chunked=bool
2559
2560 Determines how iolog is read. If false(default) entire :option:`read_iolog`
2561 will be read at once. If selected true, input from iolog will be read
2562 gradually. Useful when iolog is very large, or it is generated.
2563
b9921d1a
DZ
2564.. option:: merge_blktrace_file=str
2565
2566 When specified, rather than replaying the logs passed to :option:`read_iolog`,
2567 the logs go through a merge phase which aggregates them into a single
2568 blktrace. The resulting file is then passed on as the :option:`read_iolog`
2569 parameter. The intention here is to make the order of events consistent.
2570 This limits the influence of the scheduler compared to replaying multiple
2571 blktraces via concurrent jobs.
2572
87a48ada
DZ
2573.. option:: merge_blktrace_scalars=float_list
2574
2575 This is a percentage based option that is index paired with the list of
2576 files passed to :option:`read_iolog`. When merging is performed, scale
2577 the time of each event by the corresponding amount. For example,
2578 ``--merge_blktrace_scalars="50:100"`` runs the first trace in halftime
2579 and the second trace in realtime. This knob is separately tunable from
2580 :option:`replay_time_scale` which scales the trace during runtime and
2581 does not change the output of the merge unlike this option.
2582
55bfd8c8
DZ
2583.. option:: merge_blktrace_iters=float_list
2584
2585 This is a whole number option that is index paired with the list of files
2586 passed to :option:`read_iolog`. When merging is performed, run each trace
2587 for the specified number of iterations. For example,
2588 ``--merge_blktrace_iters="2:1"`` runs the first trace for two iterations
2589 and the second trace for one iteration.
2590
589e88b7 2591.. option:: replay_no_stall=bool
71bfa161 2592
f80dba8d 2593 When replaying I/O with :option:`read_iolog` the default behavior is to
22413915 2594 attempt to respect the timestamps within the log and replay them with the
f80dba8d
MT
2595 appropriate delay between IOPS. By setting this variable fio will not
2596 respect the timestamps and attempt to replay them as fast as possible while
2597 still respecting ordering. The result is the same I/O pattern to a given
2598 device, but different timings.
71bfa161 2599
6dd7fa77
JA
2600.. option:: replay_time_scale=int
2601
2602 When replaying I/O with :option:`read_iolog`, fio will honor the
2603 original timing in the trace. With this option, it's possible to scale
2604 the time. It's a percentage option, if set to 50 it means run at 50%
2605 the original IO rate in the trace. If set to 200, run at twice the
2606 original IO rate. Defaults to 100.
2607
f80dba8d 2608.. option:: replay_redirect=str
b4692828 2609
f80dba8d
MT
2610 While replaying I/O patterns using :option:`read_iolog` the default behavior
2611 is to replay the IOPS onto the major/minor device that each IOP was recorded
2612 from. This is sometimes undesirable because on a different machine those
2613 major/minor numbers can map to a different device. Changing hardware on the
2614 same system can also result in a different major/minor mapping.
730bd7d9 2615 ``replay_redirect`` causes all I/Os to be replayed onto the single specified
f80dba8d 2616 device regardless of the device it was recorded
9207a0cb 2617 from. i.e. :option:`replay_redirect`\= :file:`/dev/sdc` would cause all I/O
f80dba8d
MT
2618 in the blktrace or iolog to be replayed onto :file:`/dev/sdc`. This means
2619 multiple devices will be replayed onto a single device, if the trace
2620 contains multiple devices. If you want multiple devices to be replayed
2621 concurrently to multiple redirected devices you must blkparse your trace
2622 into separate traces and replay them with independent fio invocations.
2623 Unfortunately this also breaks the strict time ordering between multiple
2624 device accesses.
71bfa161 2625
f80dba8d 2626.. option:: replay_align=int
74929ac2 2627
350a535d
DZ
2628 Force alignment of the byte offsets in a trace to this value. The value
2629 must be a power of 2.
3c54bc46 2630
f80dba8d 2631.. option:: replay_scale=int
3c54bc46 2632
350a535d
DZ
2633 Scale byte offsets down by this factor when replaying traces. Should most
2634 likely use :option:`replay_align` as well.
3c54bc46 2635
38f68906
JA
2636.. option:: replay_skip=str
2637
2638 Sometimes it's useful to skip certain IO types in a replay trace.
2639 This could be, for instance, eliminating the writes in the trace.
2640 Or not replaying the trims/discards, if you are redirecting to
2641 a device that doesn't support them. This option takes a comma
2642 separated list of read, write, trim, sync.
2643
3c54bc46 2644
f80dba8d
MT
2645Threads, processes and job synchronization
2646~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3c54bc46 2647
f80dba8d 2648.. option:: thread
3c54bc46 2649
730bd7d9
SW
2650 Fio defaults to creating jobs by using fork, however if this option is
2651 given, fio will create jobs by using POSIX Threads' function
2652 :manpage:`pthread_create(3)` to create threads instead.
71bfa161 2653
f80dba8d 2654.. option:: wait_for=str
74929ac2 2655
730bd7d9
SW
2656 If set, the current job won't be started until all workers of the specified
2657 waitee job are done.
74929ac2 2658
f80dba8d
MT
2659 ``wait_for`` operates on the job name basis, so there are a few
2660 limitations. First, the waitee must be defined prior to the waiter job
2661 (meaning no forward references). Second, if a job is being referenced as a
2662 waitee, it must have a unique name (no duplicate waitees).
74929ac2 2663
f80dba8d 2664.. option:: nice=int
892a6ffc 2665
f80dba8d 2666 Run the job with the given nice value. See man :manpage:`nice(2)`.
892a6ffc 2667
f80dba8d
MT
2668 On Windows, values less than -15 set the process class to "High"; -1 through
2669 -15 set "Above Normal"; 1 through 15 "Below Normal"; and above 15 "Idle"
2670 priority class.
74929ac2 2671
f80dba8d 2672.. option:: prio=int
71bfa161 2673
f80dba8d
MT
2674 Set the I/O priority value of this job. Linux limits us to a positive value
2675 between 0 and 7, with 0 being the highest. See man
2676 :manpage:`ionice(1)`. Refer to an appropriate manpage for other operating
2677 systems since meaning of priority may differ.
71bfa161 2678
f80dba8d 2679.. option:: prioclass=int
d59aa780 2680
f80dba8d 2681 Set the I/O priority class. See man :manpage:`ionice(1)`.
d59aa780 2682
f80dba8d 2683.. option:: cpus_allowed=str
6d500c2e 2684
730bd7d9 2685 Controls the same options as :option:`cpumask`, but accepts a textual
b570e037
SW
2686 specification of the permitted CPUs instead and CPUs are indexed from 0. So
2687 to use CPUs 0 and 5 you would specify ``cpus_allowed=0,5``. This option also
2688 allows a range of CPUs to be specified -- say you wanted a binding to CPUs
2689 0, 5, and 8 to 15, you would set ``cpus_allowed=0,5,8-15``.
2690
2691 On Windows, when ``cpus_allowed`` is unset only CPUs from fio's current
2692 processor group will be used and affinity settings are inherited from the
2693 system. An fio build configured to target Windows 7 makes options that set
2694 CPUs processor group aware and values will set both the processor group
2695 and a CPU from within that group. For example, on a system where processor
2696 group 0 has 40 CPUs and processor group 1 has 32 CPUs, ``cpus_allowed``
2697 values between 0 and 39 will bind CPUs from processor group 0 and
2698 ``cpus_allowed`` values between 40 and 71 will bind CPUs from processor
2699 group 1. When using ``cpus_allowed_policy=shared`` all CPUs specified by a
2700 single ``cpus_allowed`` option must be from the same processor group. For
2701 Windows fio builds not built for Windows 7, CPUs will only be selected from
2702 (and be relative to) whatever processor group fio happens to be running in
2703 and CPUs from other processor groups cannot be used.
6d500c2e 2704
f80dba8d 2705.. option:: cpus_allowed_policy=str
6d500c2e 2706
f80dba8d 2707 Set the policy of how fio distributes the CPUs specified by
730bd7d9 2708 :option:`cpus_allowed` or :option:`cpumask`. Two policies are supported:
6d500c2e 2709
f80dba8d
MT
2710 **shared**
2711 All jobs will share the CPU set specified.
2712 **split**
2713 Each job will get a unique CPU from the CPU set.
6d500c2e 2714
22413915 2715 **shared** is the default behavior, if the option isn't specified. If
f80dba8d
MT
2716 **split** is specified, then fio will will assign one cpu per job. If not
2717 enough CPUs are given for the jobs listed, then fio will roundrobin the CPUs
2718 in the set.
6d500c2e 2719
b570e037
SW
2720.. option:: cpumask=int
2721
2722 Set the CPU affinity of this job. The parameter given is a bit mask of
2723 allowed CPUs the job may run on. So if you want the allowed CPUs to be 1
2724 and 5, you would pass the decimal value of (1 << 1 | 1 << 5), or 34. See man
2725 :manpage:`sched_setaffinity(2)`. This may not work on all supported
2726 operating systems or kernel versions. This option doesn't work well for a
2727 higher CPU count than what you can store in an integer mask, so it can only
2728 control cpus 1-32. For boxes with larger CPU counts, use
2729 :option:`cpus_allowed`.
2730
f80dba8d 2731.. option:: numa_cpu_nodes=str
6d500c2e 2732
f80dba8d
MT
2733 Set this job running on specified NUMA nodes' CPUs. The arguments allow
2734 comma delimited list of cpu numbers, A-B ranges, or `all`. Note, to enable
ac8ca2af 2735 NUMA options support, fio must be built on a system with libnuma-dev(el)
f80dba8d 2736 installed.
61b9861d 2737
f80dba8d 2738.. option:: numa_mem_policy=str
61b9861d 2739
f80dba8d
MT
2740 Set this job's memory policy and corresponding NUMA nodes. Format of the
2741 arguments::
5c94b008 2742
f80dba8d 2743 <mode>[:<nodelist>]
ce35b1ec 2744
804c0839 2745 ``mode`` is one of the following memory policies: ``default``, ``prefer``,
730bd7d9
SW
2746 ``bind``, ``interleave`` or ``local``. For ``default`` and ``local`` memory
2747 policies, no node needs to be specified. For ``prefer``, only one node is
2748 allowed. For ``bind`` and ``interleave`` the ``nodelist`` may be as
2749 follows: a comma delimited list of numbers, A-B ranges, or `all`.
71bfa161 2750
f80dba8d 2751.. option:: cgroup=str
390b1537 2752
f80dba8d
MT
2753 Add job to this control group. If it doesn't exist, it will be created. The
2754 system must have a mounted cgroup blkio mount point for this to work. If
2755 your system doesn't have it mounted, you can do so with::
5af1c6f3 2756
f80dba8d 2757 # mount -t cgroup -o blkio none /cgroup
5af1c6f3 2758
f80dba8d 2759.. option:: cgroup_weight=int
5af1c6f3 2760
f80dba8d
MT
2761 Set the weight of the cgroup to this value. See the documentation that comes
2762 with the kernel, allowed values are in the range of 100..1000.
a086c257 2763
f80dba8d 2764.. option:: cgroup_nodelete=bool
8c07860d 2765
f80dba8d
MT
2766 Normally fio will delete the cgroups it has created after the job
2767 completion. To override this behavior and to leave cgroups around after the
2768 job completion, set ``cgroup_nodelete=1``. This can be useful if one wants
2769 to inspect various cgroup files after job completion. Default: false.
8c07860d 2770
f80dba8d 2771.. option:: flow_id=int
8c07860d 2772
f80dba8d
MT
2773 The ID of the flow. If not specified, it defaults to being a global
2774 flow. See :option:`flow`.
1907dbc6 2775
f80dba8d 2776.. option:: flow=int
71bfa161 2777
f80dba8d
MT
2778 Weight in token-based flow control. If this value is used, then there is a
2779 'flow counter' which is used to regulate the proportion of activity between
2780 two or more jobs. Fio attempts to keep this flow counter near zero. The
2781 ``flow`` parameter stands for how much should be added or subtracted to the
2782 flow counter on each iteration of the main I/O loop. That is, if one job has
2783 ``flow=8`` and another job has ``flow=-1``, then there will be a roughly 1:8
2784 ratio in how much one runs vs the other.
71bfa161 2785
f80dba8d 2786.. option:: flow_watermark=int
a31041ea 2787
f80dba8d
MT
2788 The maximum value that the absolute value of the flow counter is allowed to
2789 reach before the job must wait for a lower value of the counter.
82407585 2790
f80dba8d 2791.. option:: flow_sleep=int
82407585 2792
f80dba8d
MT
2793 The period of time, in microseconds, to wait after the flow watermark has
2794 been exceeded before retrying operations.
82407585 2795
f80dba8d 2796.. option:: stonewall, wait_for_previous
82407585 2797
f80dba8d
MT
2798 Wait for preceding jobs in the job file to exit, before starting this
2799 one. Can be used to insert serialization points in the job file. A stone
2800 wall also implies starting a new reporting group, see
2801 :option:`group_reporting`.
2802
2803.. option:: exitall
2804
730bd7d9
SW
2805 By default, fio will continue running all other jobs when one job finishes
2806 but sometimes this is not the desired action. Setting ``exitall`` will
2807 instead make fio terminate all other jobs when one job finishes.
f80dba8d
MT
2808
2809.. option:: exec_prerun=str
2810
2811 Before running this job, issue the command specified through
2812 :manpage:`system(3)`. Output is redirected in a file called
2813 :file:`jobname.prerun.txt`.
2814
2815.. option:: exec_postrun=str
2816
2817 After the job completes, issue the command specified though
2818 :manpage:`system(3)`. Output is redirected in a file called
2819 :file:`jobname.postrun.txt`.
2820
2821.. option:: uid=int
2822
2823 Instead of running as the invoking user, set the user ID to this value
2824 before the thread/process does any work.
2825
2826.. option:: gid=int
2827
2828 Set group ID, see :option:`uid`.
2829
2830
2831Verification
2832~~~~~~~~~~~~
2833
2834.. option:: verify_only
2835
2836 Do not perform specified workload, only verify data still matches previous
2837 invocation of this workload. This option allows one to check data multiple
2838 times at a later date without overwriting it. This option makes sense only
2839 for workloads that write data, and does not support workloads with the
2840 :option:`time_based` option set.
2841
2842.. option:: do_verify=bool
2843
2844 Run the verify phase after a write phase. Only valid if :option:`verify` is
2845 set. Default: true.
2846
2847.. option:: verify=str
2848
2849 If writing to a file, fio can verify the file contents after each iteration
2850 of the job. Each verification method also implies verification of special
2851 header, which is written to the beginning of each block. This header also
2852 includes meta information, like offset of the block, block number, timestamp
2853 when block was written, etc. :option:`verify` can be combined with
2854 :option:`verify_pattern` option. The allowed values are:
2855
2856 **md5**
2857 Use an md5 sum of the data area and store it in the header of
2858 each block.
2859
2860 **crc64**
2861 Use an experimental crc64 sum of the data area and store it in the
2862 header of each block.
2863
2864 **crc32c**
a5896300
SW
2865 Use a crc32c sum of the data area and store it in the header of
2866 each block. This will automatically use hardware acceleration
2867 (e.g. SSE4.2 on an x86 or CRC crypto extensions on ARM64) but will
2868 fall back to software crc32c if none is found. Generally the
804c0839 2869 fastest checksum fio supports when hardware accelerated.
f80dba8d
MT
2870
2871 **crc32c-intel**
a5896300 2872 Synonym for crc32c.
f80dba8d
MT
2873
2874 **crc32**
2875 Use a crc32 sum of the data area and store it in the header of each
2876 block.
2877
2878 **crc16**
2879 Use a crc16 sum of the data area and store it in the header of each
2880 block.
2881
2882 **crc7**
2883 Use a crc7 sum of the data area and store it in the header of each
2884 block.
2885
2886 **xxhash**
2887 Use xxhash as the checksum function. Generally the fastest software
2888 checksum that fio supports.
2889
2890 **sha512**
2891 Use sha512 as the checksum function.
2892
2893 **sha256**
2894 Use sha256 as the checksum function.
2895
2896 **sha1**
2897 Use optimized sha1 as the checksum function.
82407585 2898
ae3a5acc
JA
2899 **sha3-224**
2900 Use optimized sha3-224 as the checksum function.
2901
2902 **sha3-256**
2903 Use optimized sha3-256 as the checksum function.
2904
2905 **sha3-384**
2906 Use optimized sha3-384 as the checksum function.
2907
2908 **sha3-512**
2909 Use optimized sha3-512 as the checksum function.
2910
f80dba8d
MT
2911 **meta**
2912 This option is deprecated, since now meta information is included in
2913 generic verification header and meta verification happens by
2914 default. For detailed information see the description of the
2915 :option:`verify` setting. This option is kept because of
2916 compatibility's sake with old configurations. Do not use it.
2917
2918 **pattern**
2919 Verify a strict pattern. Normally fio includes a header with some
2920 basic information and checksumming, but if this option is set, only
2921 the specific pattern set with :option:`verify_pattern` is verified.
2922
2923 **null**
2924 Only pretend to verify. Useful for testing internals with
9207a0cb 2925 :option:`ioengine`\=null, not for much else.
f80dba8d
MT
2926
2927 This option can be used for repeated burn-in tests of a system to make sure
2928 that the written data is also correctly read back. If the data direction
2929 given is a read or random read, fio will assume that it should verify a
2930 previously written file. If the data direction includes any form of write,
2931 the verify will be of the newly written data.
2932
47e6a6e5
SW
2933 To avoid false verification errors, do not use the norandommap option when
2934 verifying data with async I/O engines and I/O depths > 1. Or use the
2935 norandommap and the lfsr random generator together to avoid writing to the
2936 same offset with muliple outstanding I/Os.
2937
f80dba8d
MT
2938.. option:: verify_offset=int
2939
2940 Swap the verification header with data somewhere else in the block before
2941 writing. It is swapped back before verifying.
2942
2943.. option:: verify_interval=int
2944
2945 Write the verification header at a finer granularity than the
2946 :option:`blocksize`. It will be written for chunks the size of
2947 ``verify_interval``. :option:`blocksize` should divide this evenly.
2948
2949.. option:: verify_pattern=str
2950
2951 If set, fio will fill the I/O buffers with this pattern. Fio defaults to
2952 filling with totally random bytes, but sometimes it's interesting to fill
2953 with a known pattern for I/O verification purposes. Depending on the width
730bd7d9 2954 of the pattern, fio will fill 1/2/3/4 bytes of the buffer at the time (it can
f80dba8d
MT
2955 be either a decimal or a hex number). The ``verify_pattern`` if larger than
2956 a 32-bit quantity has to be a hex number that starts with either "0x" or
2957 "0X". Use with :option:`verify`. Also, ``verify_pattern`` supports %o
2958 format, which means that for each block offset will be written and then
2959 verified back, e.g.::
61b9861d
RP
2960
2961 verify_pattern=%o
2962
f80dba8d
MT
2963 Or use combination of everything::
2964
61b9861d 2965 verify_pattern=0xff%o"abcd"-12
e28218f3 2966
f80dba8d
MT
2967.. option:: verify_fatal=bool
2968
2969 Normally fio will keep checking the entire contents before quitting on a
2970 block verification failure. If this option is set, fio will exit the job on
2971 the first observed failure. Default: false.
2972
2973.. option:: verify_dump=bool
2974
2975 If set, dump the contents of both the original data block and the data block
2976 we read off disk to files. This allows later analysis to inspect just what
2977 kind of data corruption occurred. Off by default.
2978
2979.. option:: verify_async=int
2980
2981 Fio will normally verify I/O inline from the submitting thread. This option
2982 takes an integer describing how many async offload threads to create for I/O
2983 verification instead, causing fio to offload the duty of verifying I/O
2984 contents to one or more separate threads. If using this offload option, even
2985 sync I/O engines can benefit from using an :option:`iodepth` setting higher
2986 than 1, as it allows them to have I/O in flight while verifies are running.
d7e6ea1c 2987 Defaults to 0 async threads, i.e. verification is not asynchronous.
f80dba8d
MT
2988
2989.. option:: verify_async_cpus=str
2990
2991 Tell fio to set the given CPU affinity on the async I/O verification
2992 threads. See :option:`cpus_allowed` for the format used.
2993
2994.. option:: verify_backlog=int
2995
2996 Fio will normally verify the written contents of a job that utilizes verify
2997 once that job has completed. In other words, everything is written then
2998 everything is read back and verified. You may want to verify continually
2999 instead for a variety of reasons. Fio stores the meta data associated with
3000 an I/O block in memory, so for large verify workloads, quite a bit of memory
3001 would be used up holding this meta data. If this option is enabled, fio will
3002 write only N blocks before verifying these blocks.
3003
3004.. option:: verify_backlog_batch=int
3005
3006 Control how many blocks fio will verify if :option:`verify_backlog` is
3007 set. If not set, will default to the value of :option:`verify_backlog`
3008 (meaning the entire queue is read back and verified). If
3009 ``verify_backlog_batch`` is less than :option:`verify_backlog` then not all
3010 blocks will be verified, if ``verify_backlog_batch`` is larger than
3011 :option:`verify_backlog`, some blocks will be verified more than once.
3012
3013.. option:: verify_state_save=bool
3014
3015 When a job exits during the write phase of a verify workload, save its
3016 current state. This allows fio to replay up until that point, if the verify
3017 state is loaded for the verify read phase. The format of the filename is,
3018 roughly::
3019
f50fbdda 3020 <type>-<jobname>-<jobindex>-verify.state.
f80dba8d
MT
3021
3022 <type> is "local" for a local run, "sock" for a client/server socket
3023 connection, and "ip" (192.168.0.1, for instance) for a networked
d7e6ea1c 3024 client/server connection. Defaults to true.
f80dba8d
MT
3025
3026.. option:: verify_state_load=bool
3027
3028 If a verify termination trigger was used, fio stores the current write state
3029 of each thread. This can be used at verification time so that fio knows how
3030 far it should verify. Without this information, fio will run a full
a47b697c
SW
3031 verification pass, according to the settings in the job file used. Default
3032 false.
f80dba8d
MT
3033
3034.. option:: trim_percentage=int
3035
3036 Number of verify blocks to discard/trim.
3037
3038.. option:: trim_verify_zero=bool
3039
22413915 3040 Verify that trim/discarded blocks are returned as zeros.
f80dba8d
MT
3041
3042.. option:: trim_backlog=int
3043
5cfd1e9a 3044 Trim after this number of blocks are written.
f80dba8d
MT
3045
3046.. option:: trim_backlog_batch=int
3047
3048 Trim this number of I/O blocks.
3049
3050.. option:: experimental_verify=bool
3051
3052 Enable experimental verification.
3053
f80dba8d
MT
3054Steady state
3055~~~~~~~~~~~~
3056
3057.. option:: steadystate=str:float, ss=str:float
3058
3059 Define the criterion and limit for assessing steady state performance. The
3060 first parameter designates the criterion whereas the second parameter sets
3061 the threshold. When the criterion falls below the threshold for the
3062 specified duration, the job will stop. For example, `iops_slope:0.1%` will
3063 direct fio to terminate the job when the least squares regression slope
3064 falls below 0.1% of the mean IOPS. If :option:`group_reporting` is enabled
3065 this will apply to all jobs in the group. Below is the list of available
3066 steady state assessment criteria. All assessments are carried out using only
3067 data from the rolling collection window. Threshold limits can be expressed
3068 as a fixed value or as a percentage of the mean in the collection window.
3069
1cb049d9
VF
3070 When using this feature, most jobs should include the :option:`time_based`
3071 and :option:`runtime` options or the :option:`loops` option so that fio does not
3072 stop running after it has covered the full size of the specified file(s) or device(s).
3073
f80dba8d
MT
3074 **iops**
3075 Collect IOPS data. Stop the job if all individual IOPS measurements
3076 are within the specified limit of the mean IOPS (e.g., ``iops:2``
3077 means that all individual IOPS values must be within 2 of the mean,
3078 whereas ``iops:0.2%`` means that all individual IOPS values must be
3079 within 0.2% of the mean IOPS to terminate the job).
3080
3081 **iops_slope**
3082 Collect IOPS data and calculate the least squares regression
3083 slope. Stop the job if the slope falls below the specified limit.
3084
3085 **bw**
3086 Collect bandwidth data. Stop the job if all individual bandwidth
3087 measurements are within the specified limit of the mean bandwidth.
3088
3089 **bw_slope**
3090 Collect bandwidth data and calculate the least squares regression
3091 slope. Stop the job if the slope falls below the specified limit.
3092
3093.. option:: steadystate_duration=time, ss_dur=time
3094
3095 A rolling window of this duration will be used to judge whether steady state
3096 has been reached. Data will be collected once per second. The default is 0
f75ede1d 3097 which disables steady state detection. When the unit is omitted, the
947e0fe0 3098 value is interpreted in seconds.
f80dba8d
MT
3099
3100.. option:: steadystate_ramp_time=time, ss_ramp=time
3101
3102 Allow the job to run for the specified duration before beginning data
3103 collection for checking the steady state job termination criterion. The
947e0fe0 3104 default is 0. When the unit is omitted, the value is interpreted in seconds.
f80dba8d
MT
3105
3106
3107Measurements and reporting
3108~~~~~~~~~~~~~~~~~~~~~~~~~~
3109
3110.. option:: per_job_logs=bool
3111
3112 If set, this generates bw/clat/iops log with per file private filenames. If
3113 not set, jobs with identical names will share the log filename. Default:
3114 true.
3115
3116.. option:: group_reporting
3117
3118 It may sometimes be interesting to display statistics for groups of jobs as
3119 a whole instead of for each individual job. This is especially true if
3120 :option:`numjobs` is used; looking at individual thread/process output
3121 quickly becomes unwieldy. To see the final report per-group instead of
3122 per-job, use :option:`group_reporting`. Jobs in a file will be part of the
3123 same reporting group, unless if separated by a :option:`stonewall`, or by
3124 using :option:`new_group`.
3125
3126.. option:: new_group
3127
3128 Start a new reporting group. See: :option:`group_reporting`. If not given,
3129 all jobs in a file will be part of the same reporting group, unless
3130 separated by a :option:`stonewall`.
3131
589e88b7 3132.. option:: stats=bool
8243be59
JA
3133
3134 By default, fio collects and shows final output results for all jobs
3135 that run. If this option is set to 0, then fio will ignore it in
3136 the final stat output.
3137
f80dba8d
MT
3138.. option:: write_bw_log=str
3139
3140 If given, write a bandwidth log for this job. Can be used to store data of
074f0817 3141 the bandwidth of the jobs in their lifetime.
f80dba8d 3142
074f0817
SW
3143 If no str argument is given, the default filename of
3144 :file:`jobname_type.x.log` is used. Even when the argument is given, fio
3145 will still append the type of log. So if one specifies::
3146
3147 write_bw_log=foo
f80dba8d 3148
074f0817
SW
3149 The actual log name will be :file:`foo_bw.x.log` where `x` is the index
3150 of the job (`1..N`, where `N` is the number of jobs). If
3151 :option:`per_job_logs` is false, then the filename will not include the
3152 `.x` job index.
e3cedca7 3153
074f0817
SW
3154 The included :command:`fio_generate_plots` script uses :command:`gnuplot` to turn these
3155 text files into nice graphs. See `Log File Formats`_ for how data is
3156 structured within the file.
3157
3158.. option:: write_lat_log=str
e3cedca7 3159
074f0817 3160 Same as :option:`write_bw_log`, except this option creates I/O
77b7e675
SW
3161 submission (e.g., :file:`name_slat.x.log`), completion (e.g.,
3162 :file:`name_clat.x.log`), and total (e.g., :file:`name_lat.x.log`)
074f0817
SW
3163 latency files instead. See :option:`write_bw_log` for details about
3164 the filename format and `Log File Formats`_ for how data is structured
3165 within the files.
be4ecfdf 3166
f80dba8d 3167.. option:: write_hist_log=str
06842027 3168
074f0817 3169 Same as :option:`write_bw_log` but writes an I/O completion latency
77b7e675 3170 histogram file (e.g., :file:`name_hist.x.log`) instead. Note that this
074f0817
SW
3171 file will be empty unless :option:`log_hist_msec` has also been set.
3172 See :option:`write_bw_log` for details about the filename format and
3173 `Log File Formats`_ for how data is structured within the file.
06842027 3174
f80dba8d 3175.. option:: write_iops_log=str
06842027 3176
074f0817 3177 Same as :option:`write_bw_log`, but writes an IOPS file (e.g.
15417073
SW
3178 :file:`name_iops.x.log`) instead. Because fio defaults to individual
3179 I/O logging, the value entry in the IOPS log will be 1 unless windowed
3180 logging (see :option:`log_avg_msec`) has been enabled. See
3181 :option:`write_bw_log` for details about the filename format and `Log
3182 File Formats`_ for how data is structured within the file.
06842027 3183
f80dba8d 3184.. option:: log_avg_msec=int
06842027 3185
f80dba8d
MT
3186 By default, fio will log an entry in the iops, latency, or bw log for every
3187 I/O that completes. When writing to the disk log, that can quickly grow to a
3188 very large size. Setting this option makes fio average the each log entry
3189 over the specified period of time, reducing the resolution of the log. See
3190 :option:`log_max_value` as well. Defaults to 0, logging all entries.
6fc82095 3191 Also see `Log File Formats`_.
06842027 3192
f80dba8d 3193.. option:: log_hist_msec=int
06842027 3194
f80dba8d
MT
3195 Same as :option:`log_avg_msec`, but logs entries for completion latency
3196 histograms. Computing latency percentiles from averages of intervals using
c60ebc45 3197 :option:`log_avg_msec` is inaccurate. Setting this option makes fio log
f80dba8d
MT
3198 histogram entries over the specified period of time, reducing log sizes for
3199 high IOPS devices while retaining percentile accuracy. See
074f0817
SW
3200 :option:`log_hist_coarseness` and :option:`write_hist_log` as well.
3201 Defaults to 0, meaning histogram logging is disabled.
06842027 3202
f80dba8d 3203.. option:: log_hist_coarseness=int
06842027 3204
f80dba8d
MT
3205 Integer ranging from 0 to 6, defining the coarseness of the resolution of
3206 the histogram logs enabled with :option:`log_hist_msec`. For each increment
3207 in coarseness, fio outputs half as many bins. Defaults to 0, for which
074f0817
SW
3208 histogram logs contain 1216 latency bins. See :option:`write_hist_log`
3209 and `Log File Formats`_.
8b28bd41 3210
f80dba8d 3211.. option:: log_max_value=bool
66c098b8 3212
f80dba8d
MT
3213 If :option:`log_avg_msec` is set, fio logs the average over that window. If
3214 you instead want to log the maximum value, set this option to 1. Defaults to
3215 0, meaning that averaged values are logged.
a696fa2a 3216
589e88b7 3217.. option:: log_offset=bool
a696fa2a 3218
f80dba8d 3219 If this is set, the iolog options will include the byte offset for the I/O
5a83478f
SW
3220 entry as well as the other data values. Defaults to 0 meaning that
3221 offsets are not present in logs. Also see `Log File Formats`_.
71bfa161 3222
f80dba8d 3223.. option:: log_compression=int
7de87099 3224
f80dba8d
MT
3225 If this is set, fio will compress the I/O logs as it goes, to keep the
3226 memory footprint lower. When a log reaches the specified size, that chunk is
3227 removed and compressed in the background. Given that I/O logs are fairly
3228 highly compressible, this yields a nice memory savings for longer runs. The
3229 downside is that the compression will consume some background CPU cycles, so
3230 it may impact the run. This, however, is also true if the logging ends up
3231 consuming most of the system memory. So pick your poison. The I/O logs are
3232 saved normally at the end of a run, by decompressing the chunks and storing
3233 them in the specified log file. This feature depends on the availability of
3234 zlib.
e0b0d892 3235
f80dba8d 3236.. option:: log_compression_cpus=str
e0b0d892 3237
f80dba8d
MT
3238 Define the set of CPUs that are allowed to handle online log compression for
3239 the I/O jobs. This can provide better isolation between performance
0cf90a62
SW
3240 sensitive jobs, and background compression work. See
3241 :option:`cpus_allowed` for the format used.
9e684a49 3242
f80dba8d 3243.. option:: log_store_compressed=bool
9e684a49 3244
f80dba8d
MT
3245 If set, fio will store the log files in a compressed format. They can be
3246 decompressed with fio, using the :option:`--inflate-log` command line
3247 parameter. The files will be stored with a :file:`.fz` suffix.
9e684a49 3248
f80dba8d 3249.. option:: log_unix_epoch=bool
9e684a49 3250
f80dba8d
MT
3251 If set, fio will log Unix timestamps to the log files produced by enabling
3252 write_type_log for each log type, instead of the default zero-based
3253 timestamps.
3254
3255.. option:: block_error_percentiles=bool
3256
3257 If set, record errors in trim block-sized units from writes and trims and
3258 output a histogram of how many trims it took to get to errors, and what kind
3259 of error was encountered.
3260
3261.. option:: bwavgtime=int
3262
3263 Average the calculated bandwidth over the given time. Value is specified in
3264 milliseconds. If the job also does bandwidth logging through
3265 :option:`write_bw_log`, then the minimum of this option and
3266 :option:`log_avg_msec` will be used. Default: 500ms.
3267
3268.. option:: iopsavgtime=int
3269
3270 Average the calculated IOPS over the given time. Value is specified in
3271 milliseconds. If the job also does IOPS logging through
3272 :option:`write_iops_log`, then the minimum of this option and
3273 :option:`log_avg_msec` will be used. Default: 500ms.
3274
3275.. option:: disk_util=bool
3276
3277 Generate disk utilization statistics, if the platform supports it.
3278 Default: true.
3279
3280.. option:: disable_lat=bool
3281
3282 Disable measurements of total latency numbers. Useful only for cutting back
3283 the number of calls to :manpage:`gettimeofday(2)`, as that does impact
3284 performance at really high IOPS rates. Note that to really get rid of a
3285 large amount of these calls, this option must be used with
f75ede1d 3286 :option:`disable_slat` and :option:`disable_bw_measurement` as well.
f80dba8d
MT
3287
3288.. option:: disable_clat=bool
3289
3290 Disable measurements of completion latency numbers. See
3291 :option:`disable_lat`.
3292
3293.. option:: disable_slat=bool
3294
3295 Disable measurements of submission latency numbers. See
f50fbdda 3296 :option:`disable_lat`.
f80dba8d 3297
f75ede1d 3298.. option:: disable_bw_measurement=bool, disable_bw=bool
f80dba8d
MT
3299
3300 Disable measurements of throughput/bandwidth numbers. See
3301 :option:`disable_lat`.
3302
3303.. option:: clat_percentiles=bool
3304
b599759b
JA
3305 Enable the reporting of percentiles of completion latencies. This
3306 option is mutually exclusive with :option:`lat_percentiles`.
3307
3308.. option:: lat_percentiles=bool
3309
b71968b1 3310 Enable the reporting of percentiles of I/O latencies. This is similar
b599759b
JA
3311 to :option:`clat_percentiles`, except that this includes the
3312 submission latency. This option is mutually exclusive with
3313 :option:`clat_percentiles`.
f80dba8d
MT
3314
3315.. option:: percentile_list=float_list
3316
c32ba107
JA
3317 Overwrite the default list of percentiles for completion latencies and
3318 the block error histogram. Each number is a floating number in the
3319 range (0,100], and the maximum length of the list is 20. Use ``:`` to
3320 separate the numbers, and list the numbers in ascending order. For
3321 example, ``--percentile_list=99.5:99.9`` will cause fio to report the
3322 values of completion latency below which 99.5% and 99.9% of the observed
3323 latencies fell, respectively.
f80dba8d 3324
e883cb35
JF
3325.. option:: significant_figures=int
3326
c32ba107
JA
3327 If using :option:`--output-format` of `normal`, set the significant
3328 figures to this value. Higher values will yield more precise IOPS and
3329 throughput units, while lower values will round. Requires a minimum
3330 value of 1 and a maximum value of 10. Defaults to 4.
e883cb35 3331
f80dba8d
MT
3332
3333Error handling
3334~~~~~~~~~~~~~~
3335
3336.. option:: exitall_on_error
3337
3338 When one job finishes in error, terminate the rest. The default is to wait
3339 for each job to finish.
3340
3341.. option:: continue_on_error=str
3342
3343 Normally fio will exit the job on the first observed failure. If this option
3344 is set, fio will continue the job when there is a 'non-fatal error' (EIO or
3345 EILSEQ) until the runtime is exceeded or the I/O size specified is
3346 completed. If this option is used, there are two more stats that are
3347 appended, the total error count and the first error. The error field given
3348 in the stats is the first error that was hit during the run.
3349
3350 The allowed values are:
3351
3352 **none**
3353 Exit on any I/O or verify errors.
3354
3355 **read**
3356 Continue on read errors, exit on all others.
3357
3358 **write**
3359 Continue on write errors, exit on all others.
3360
3361 **io**
3362 Continue on any I/O error, exit on all others.
3363
3364 **verify**
3365 Continue on verify errors, exit on all others.
3366
3367 **all**
3368 Continue on all errors.
3369
3370 **0**
3371 Backward-compatible alias for 'none'.
3372
3373 **1**
3374 Backward-compatible alias for 'all'.
3375
3376.. option:: ignore_error=str
3377
3378 Sometimes you want to ignore some errors during test in that case you can
a35ef7cb
TK
3379 specify error list for each error type, instead of only being able to
3380 ignore the default 'non-fatal error' using :option:`continue_on_error`.
f80dba8d
MT
3381 ``ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST`` errors for
3382 given error type is separated with ':'. Error may be symbol ('ENOSPC',
3383 'ENOMEM') or integer. Example::
3384
3385 ignore_error=EAGAIN,ENOSPC:122
3386
3387 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from
a35ef7cb
TK
3388 WRITE. This option works by overriding :option:`continue_on_error` with
3389 the list of errors for each error type if any.
f80dba8d
MT
3390
3391.. option:: error_dump=bool
3392
3393 If set dump every error even if it is non fatal, true by default. If
3394 disabled only fatal error will be dumped.
3395
f75ede1d
SW
3396Running predefined workloads
3397----------------------------
3398
3399Fio includes predefined profiles that mimic the I/O workloads generated by
3400other tools.
3401
3402.. option:: profile=str
3403
3404 The predefined workload to run. Current profiles are:
3405
3406 **tiobench**
3407 Threaded I/O bench (tiotest/tiobench) like workload.
3408
3409 **act**
3410 Aerospike Certification Tool (ACT) like workload.
3411
3412To view a profile's additional options use :option:`--cmdhelp` after specifying
3413the profile. For example::
3414
f50fbdda 3415 $ fio --profile=act --cmdhelp
f75ede1d
SW
3416
3417Act profile options
3418~~~~~~~~~~~~~~~~~~~
3419
3420.. option:: device-names=str
3421 :noindex:
3422
3423 Devices to use.
3424
3425.. option:: load=int
3426 :noindex:
3427
3428 ACT load multiplier. Default: 1.
3429
3430.. option:: test-duration=time
3431 :noindex:
3432
947e0fe0
SW
3433 How long the entire test takes to run. When the unit is omitted, the value
3434 is given in seconds. Default: 24h.
f75ede1d
SW
3435
3436.. option:: threads-per-queue=int
3437 :noindex:
3438
f50fbdda 3439 Number of read I/O threads per device. Default: 8.
f75ede1d
SW
3440
3441.. option:: read-req-num-512-blocks=int
3442 :noindex:
3443
3444 Number of 512B blocks to read at the time. Default: 3.
3445
3446.. option:: large-block-op-kbytes=int
3447 :noindex:
3448
3449 Size of large block ops in KiB (writes). Default: 131072.
3450
3451.. option:: prep
3452 :noindex:
3453
3454 Set to run ACT prep phase.
3455
3456Tiobench profile options
3457~~~~~~~~~~~~~~~~~~~~~~~~
3458
3459.. option:: size=str
3460 :noindex:
3461
f50fbdda 3462 Size in MiB.
f75ede1d
SW
3463
3464.. option:: block=int
3465 :noindex:
3466
3467 Block size in bytes. Default: 4096.
3468
3469.. option:: numruns=int
3470 :noindex:
3471
3472 Number of runs.
3473
3474.. option:: dir=str
3475 :noindex:
3476
3477 Test directory.
3478
3479.. option:: threads=int
3480 :noindex:
3481
3482 Number of threads.
f80dba8d
MT
3483
3484Interpreting the output
3485-----------------------
3486
36214730
SW
3487..
3488 Example output was based on the following:
3489 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --time_based \
3490 --rate=1256k --bs=14K --name=quick --runtime=1s --name=mixed \
3491 --runtime=2m --rw=rw
3492
f80dba8d
MT
3493Fio spits out a lot of output. While running, fio will display the status of the
3494jobs created. An example of that would be::
3495
9d25d068 3496 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 3497
36214730
SW
3498The characters inside the first set of square brackets denote the current status of
3499each thread. The first character is the first job defined in the job file, and so
3500forth. The possible values (in typical life cycle order) are:
f80dba8d
MT
3501
3502+------+-----+-----------------------------------------------------------+
3503| Idle | Run | |
3504+======+=====+===========================================================+
3505| P | | Thread setup, but not started. |
3506+------+-----+-----------------------------------------------------------+
3507| C | | Thread created. |
3508+------+-----+-----------------------------------------------------------+
3509| I | | Thread initialized, waiting or generating necessary data. |
3510+------+-----+-----------------------------------------------------------+
3511| | p | Thread running pre-reading file(s). |
3512+------+-----+-----------------------------------------------------------+
36214730
SW
3513| | / | Thread is in ramp period. |
3514+------+-----+-----------------------------------------------------------+
f80dba8d
MT
3515| | R | Running, doing sequential reads. |
3516+------+-----+-----------------------------------------------------------+
3517| | r | Running, doing random reads. |
3518+------+-----+-----------------------------------------------------------+
3519| | W | Running, doing sequential writes. |
3520+------+-----+-----------------------------------------------------------+
3521| | w | Running, doing random writes. |
3522+------+-----+-----------------------------------------------------------+
3523| | M | Running, doing mixed sequential reads/writes. |
3524+------+-----+-----------------------------------------------------------+
3525| | m | Running, doing mixed random reads/writes. |
3526+------+-----+-----------------------------------------------------------+
36214730
SW
3527| | D | Running, doing sequential trims. |
3528+------+-----+-----------------------------------------------------------+
3529| | d | Running, doing random trims. |
3530+------+-----+-----------------------------------------------------------+
3531| | F | Running, currently waiting for :manpage:`fsync(2)`. |
f80dba8d
MT
3532+------+-----+-----------------------------------------------------------+
3533| | V | Running, doing verification of written data. |
3534+------+-----+-----------------------------------------------------------+
36214730
SW
3535| f | | Thread finishing. |
3536+------+-----+-----------------------------------------------------------+
f80dba8d
MT
3537| E | | Thread exited, not reaped by main thread yet. |
3538+------+-----+-----------------------------------------------------------+
36214730 3539| _ | | Thread reaped. |
f80dba8d
MT
3540+------+-----+-----------------------------------------------------------+
3541| X | | Thread reaped, exited with an error. |
3542+------+-----+-----------------------------------------------------------+
3543| K | | Thread reaped, exited due to signal. |
3544+------+-----+-----------------------------------------------------------+
3545
36214730
SW
3546..
3547 Example output was based on the following:
3548 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --runtime=58m \
3549 --time_based --rate=2512k --bs=256K --numjobs=10 \
3550 --name=readers --rw=read --name=writers --rw=write
3551
f80dba8d 3552Fio will condense the thread string as not to take up more space on the command
36214730 3553line than needed. For instance, if you have 10 readers and 10 writers running,
f80dba8d
MT
3554the output would look like this::
3555
9d25d068 3556 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 3557
36214730
SW
3558Note that the status string is displayed in order, so it's possible to tell which of
3559the jobs are currently doing what. In the example above this means that jobs 1--10
3560are readers and 11--20 are writers.
f80dba8d
MT
3561
3562The other values are fairly self explanatory -- number of threads currently
36214730
SW
3563running and doing I/O, the number of currently open files (f=), the estimated
3564completion percentage, the rate of I/O since last check (read speed listed first,
f50fbdda
TK
3565then write speed and optionally trim speed) in terms of bandwidth and IOPS,
3566and time to completion for the current running group. It's impossible to estimate
3567runtime of the following groups (if any).
36214730
SW
3568
3569..
3570 Example output was based on the following:
3571 TZ=UTC fio --iodepth=16 --ioengine=posixaio --filename=/tmp/fiofile \
3572 --direct=1 --size=100M --time_based --runtime=50s --rate_iops=89 \
3573 --bs=7K --name=Client1 --rw=write
3574
3575When fio is done (or interrupted by :kbd:`Ctrl-C`), it will show the data for
3576each thread, group of threads, and disks in that order. For each overall thread (or
3577group) the output looks like::
3578
3579 Client1: (groupid=0, jobs=1): err= 0: pid=16109: Sat Jun 24 12:07:54 2017
3580 write: IOPS=88, BW=623KiB/s (638kB/s)(30.4MiB/50032msec)
3581 slat (nsec): min=500, max=145500, avg=8318.00, stdev=4781.50
3582 clat (usec): min=170, max=78367, avg=4019.02, stdev=8293.31
3583 lat (usec): min=174, max=78375, avg=4027.34, stdev=8291.79
3584 clat percentiles (usec):
3585 | 1.00th=[ 302], 5.00th=[ 326], 10.00th=[ 343], 20.00th=[ 363],
3586 | 30.00th=[ 392], 40.00th=[ 404], 50.00th=[ 416], 60.00th=[ 445],
3587 | 70.00th=[ 816], 80.00th=[ 6718], 90.00th=[12911], 95.00th=[21627],
3588 | 99.00th=[43779], 99.50th=[51643], 99.90th=[68682], 99.95th=[72877],
3589 | 99.99th=[78119]
3590 bw ( KiB/s): min= 532, max= 686, per=0.10%, avg=622.87, stdev=24.82, samples= 100
3591 iops : min= 76, max= 98, avg=88.98, stdev= 3.54, samples= 100
29092211
VF
3592 lat (usec) : 250=0.04%, 500=64.11%, 750=4.81%, 1000=2.79%
3593 lat (msec) : 2=4.16%, 4=1.84%, 10=4.90%, 20=11.33%, 50=5.37%
3594 lat (msec) : 100=0.65%
36214730
SW
3595 cpu : usr=0.27%, sys=0.18%, ctx=12072, majf=0, minf=21
3596 IO depths : 1=85.0%, 2=13.1%, 4=1.8%, 8=0.1%, 16=0.0%, 32=0.0%, >=64=0.0%
3597 submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3598 complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3599 issued rwt: total=0,4450,0, short=0,0,0, dropped=0,0,0
3600 latency : target=0, window=0, percentile=100.00%, depth=8
3601
3602The job name (or first job's name when using :option:`group_reporting`) is printed,
3603along with the group id, count of jobs being aggregated, last error id seen (which
3604is 0 when there are no errors), pid/tid of that thread and the time the job/group
3605completed. Below are the I/O statistics for each data direction performed (showing
3606writes in the example above). In the order listed, they denote:
3607
3608**read/write/trim**
3609 The string before the colon shows the I/O direction the statistics
3610 are for. **IOPS** is the average I/Os performed per second. **BW**
3611 is the average bandwidth rate shown as: value in power of 2 format
3612 (value in power of 10 format). The last two values show: (**total
3613 I/O performed** in power of 2 format / **runtime** of that thread).
f80dba8d
MT
3614
3615**slat**
36214730
SW
3616 Submission latency (**min** being the minimum, **max** being the
3617 maximum, **avg** being the average, **stdev** being the standard
3618 deviation). This is the time it took to submit the I/O. For
3619 sync I/O this row is not displayed as the slat is really the
3620 completion latency (since queue/complete is one operation there).
3621 This value can be in nanoseconds, microseconds or milliseconds ---
3622 fio will choose the most appropriate base and print that (in the
3623 example above nanoseconds was the best scale). Note: in :option:`--minimal` mode
0d237712 3624 latencies are always expressed in microseconds.
f80dba8d
MT
3625
3626**clat**
3627 Completion latency. Same names as slat, this denotes the time from
3628 submission to completion of the I/O pieces. For sync I/O, clat will
3629 usually be equal (or very close) to 0, as the time from submit to
3630 complete is basically just CPU time (I/O has already been done, see slat
3631 explanation).
3632
29092211
VF
3633**lat**
3634 Total latency. Same names as slat and clat, this denotes the time from
3635 when fio created the I/O unit to completion of the I/O operation.
3636
f80dba8d 3637**bw**
36214730
SW
3638 Bandwidth statistics based on samples. Same names as the xlat stats,
3639 but also includes the number of samples taken (**samples**) and an
3640 approximate percentage of total aggregate bandwidth this thread
3641 received in its group (**per**). This last value is only really
3642 useful if the threads in this group are on the same disk, since they
3643 are then competing for disk access.
3644
3645**iops**
3646 IOPS statistics based on samples. Same names as bw.
f80dba8d 3647
29092211
VF
3648**lat (nsec/usec/msec)**
3649 The distribution of I/O completion latencies. This is the time from when
3650 I/O leaves fio and when it gets completed. Unlike the separate
3651 read/write/trim sections above, the data here and in the remaining
3652 sections apply to all I/Os for the reporting group. 250=0.04% means that
3653 0.04% of the I/Os completed in under 250us. 500=64.11% means that 64.11%
3654 of the I/Os required 250 to 499us for completion.
3655
f80dba8d
MT
3656**cpu**
3657 CPU usage. User and system time, along with the number of context
3658 switches this thread went through, usage of system and user time, and
3659 finally the number of major and minor page faults. The CPU utilization
3660 numbers are averages for the jobs in that reporting group, while the
23a8e176 3661 context and fault counters are summed.
f80dba8d
MT
3662
3663**IO depths**
a2140525
SW
3664 The distribution of I/O depths over the job lifetime. The numbers are
3665 divided into powers of 2 and each entry covers depths from that value
3666 up to those that are lower than the next entry -- e.g., 16= covers
3667 depths from 16 to 31. Note that the range covered by a depth
3668 distribution entry can be different to the range covered by the
3669 equivalent submit/complete distribution entry.
f80dba8d
MT
3670
3671**IO submit**
3672 How many pieces of I/O were submitting in a single submit call. Each
c60ebc45 3673 entry denotes that amount and below, until the previous entry -- e.g.,
a2140525
SW
3674 16=100% means that we submitted anywhere between 9 to 16 I/Os per submit
3675 call. Note that the range covered by a submit distribution entry can
3676 be different to the range covered by the equivalent depth distribution
3677 entry.
f80dba8d
MT
3678
3679**IO complete**
3680 Like the above submit number, but for completions instead.
3681
36214730
SW
3682**IO issued rwt**
3683 The number of read/write/trim requests issued, and how many of them were
3684 short or dropped.
f80dba8d 3685
29092211 3686**IO latency**
ee21ebee 3687 These values are for :option:`latency_target` and related options. When
29092211
VF
3688 these options are engaged, this section describes the I/O depth required
3689 to meet the specified latency target.
71bfa161 3690
36214730
SW
3691..
3692 Example output was based on the following:
3693 TZ=UTC fio --ioengine=null --iodepth=2 --size=100M --numjobs=2 \
3694 --rate_process=poisson --io_limit=32M --name=read --bs=128k \
3695 --rate=11M --name=write --rw=write --bs=2k --rate=700k
3696
71bfa161 3697After each client has been listed, the group statistics are printed. They
f80dba8d 3698will look like this::
71bfa161 3699
f80dba8d 3700 Run status group 0 (all jobs):
36214730
SW
3701 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
3702 WRITE: bw=1231KiB/s (1261kB/s), 616KiB/s-621KiB/s (630kB/s-636kB/s), io=64.0MiB (67.1MB), run=52747-53223msec
71bfa161 3703
36214730 3704For each data direction it prints:
71bfa161 3705
36214730
SW
3706**bw**
3707 Aggregate bandwidth of threads in this group followed by the
3708 minimum and maximum bandwidth of all the threads in this group.
3709 Values outside of brackets are power-of-2 format and those
3710 within are the equivalent value in a power-of-10 format.
f80dba8d 3711**io**
36214730
SW
3712 Aggregate I/O performed of all threads in this group. The
3713 format is the same as bw.
3714**run**
3715 The smallest and longest runtimes of the threads in this group.
71bfa161 3716
f50fbdda 3717And finally, the disk statistics are printed. This is Linux specific. They will look like this::
71bfa161 3718
f80dba8d
MT
3719 Disk stats (read/write):
3720 sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00%
71bfa161
JA
3721
3722Each value is printed for both reads and writes, with reads first. The
3723numbers denote:
3724
f80dba8d 3725**ios**
c60ebc45 3726 Number of I/Os performed by all groups.
f80dba8d 3727**merge**
007c7be9 3728 Number of merges performed by the I/O scheduler.
f80dba8d
MT
3729**ticks**
3730 Number of ticks we kept the disk busy.
36214730 3731**in_queue**
f80dba8d
MT
3732 Total time spent in the disk queue.
3733**util**
3734 The disk utilization. A value of 100% means we kept the disk
71bfa161
JA
3735 busy constantly, 50% would be a disk idling half of the time.
3736
f80dba8d
MT
3737It is also possible to get fio to dump the current output while it is running,
3738without terminating the job. To do that, send fio the **USR1** signal. You can
3739also get regularly timed dumps by using the :option:`--status-interval`
3740parameter, or by creating a file in :file:`/tmp` named
3741:file:`fio-dump-status`. If fio sees this file, it will unlink it and dump the
3742current output status.
8423bd11 3743
71bfa161 3744
f80dba8d
MT
3745Terse output
3746------------
71bfa161 3747
f80dba8d
MT
3748For scripted usage where you typically want to generate tables or graphs of the
3749results, fio can output the results in a semicolon separated format. The format
3750is one long line of values, such as::
71bfa161 3751
f80dba8d
MT
3752 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%
3753 A description of this job goes here.
562c2d2f 3754
4e757af1
VF
3755The job description (if provided) follows on a second line for terse v2.
3756It appears on the same line for other terse versions.
71bfa161 3757
a7f77fa6
SW
3758To enable terse output, use the :option:`--minimal` or
3759:option:`--output-format`\=terse command line options. The
f80dba8d
MT
3760first value is the version of the terse output format. If the output has to be
3761changed for some reason, this number will be incremented by 1 to signify that
3762change.
6820cb3b 3763
a2c95580 3764Split up, the format is as follows (comments in brackets denote when a
007c7be9 3765field was introduced or whether it's specific to some terse version):
71bfa161 3766
f80dba8d
MT
3767 ::
3768
f50fbdda 3769 terse version, fio version [v3], jobname, groupid, error
f80dba8d
MT
3770
3771 READ status::
3772
3773 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
3774 Submission latency: min, max, mean, stdev (usec)
3775 Completion latency: min, max, mean, stdev (usec)
3776 Completion latency percentiles: 20 fields (see below)
3777 Total latency: min, max, mean, stdev (usec)
f50fbdda
TK
3778 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
3779 IOPS [v5]: min, max, mean, stdev, number of samples
f80dba8d
MT
3780
3781 WRITE status:
3782
3783 ::
3784
3785 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
3786 Submission latency: min, max, mean, stdev (usec)
247823cc 3787 Completion latency: min, max, mean, stdev (usec)
f80dba8d
MT
3788 Completion latency percentiles: 20 fields (see below)
3789 Total latency: min, max, mean, stdev (usec)
f50fbdda
TK
3790 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
3791 IOPS [v5]: min, max, mean, stdev, number of samples
a2c95580
AH
3792
3793 TRIM status [all but version 3]:
3794
f50fbdda 3795 Fields are similar to READ/WRITE status.
f80dba8d
MT
3796
3797 CPU usage::
3798
3799 user, system, context switches, major faults, minor faults
3800
3801 I/O depths::
3802
3803 <=1, 2, 4, 8, 16, 32, >=64