Merge branch 'pshared1' of https://github.com/kusumi/fio
[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
3b803fe1
SW
768 See the :option:`filename` option for information on how to escape "``:``"
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
3b803fe1 788 Each colon in the wanted path must be escaped with a ``\``
02dd2689
SW
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
3b803fe1 791 :file:`F:\\filename` then you would use ``filename=F\:\filename``.
02dd2689 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
38ca5f03
TV
1176 **truncate**
1177 Extend file to final size via :manpage:`ftruncate(2)`
1178 instead of allocating.
1179
f80dba8d
MT
1180 **0**
1181 Backward-compatible alias for **none**.
1182
1183 **1**
1184 Backward-compatible alias for **posix**.
1185
1186 May not be available on all supported platforms. **keep** is only available
2c3e17be
SW
1187 on Linux. If using ZFS on Solaris this cannot be set to **posix**
1188 because ZFS doesn't support pre-allocation. Default: **native** if any
38ca5f03
TV
1189 pre-allocation methods except **truncate** are available, **none** if not.
1190
1191 Note that using **truncate** on Windows will interact surprisingly
1192 with non-sequential write patterns. When writing to a file that has
1193 been extended by setting the end-of-file information, Windows will
1194 backfill the unwritten portion of the file up to that offset with
1195 zeroes before issuing the new write. This means that a single small
1196 write to the end of an extended file will stall until the entire
1197 file has been filled with zeroes.
f80dba8d
MT
1198
1199.. option:: fadvise_hint=str
1200
c712c97a
JA
1201 Use :manpage:`posix_fadvise(2)` or :manpage:`posix_fadvise(2)` to
1202 advise the kernel on what I/O patterns are likely to be issued.
1203 Accepted values are:
f80dba8d
MT
1204
1205 **0**
1206 Backwards-compatible hint for "no hint".
1207
1208 **1**
1209 Backwards compatible hint for "advise with fio workload type". This
1210 uses **FADV_RANDOM** for a random workload, and **FADV_SEQUENTIAL**
1211 for a sequential workload.
1212
1213 **sequential**
1214 Advise using **FADV_SEQUENTIAL**.
1215
1216 **random**
1217 Advise using **FADV_RANDOM**.
1218
8f4b9f24 1219.. option:: write_hint=str
f80dba8d 1220
8f4b9f24
JA
1221 Use :manpage:`fcntl(2)` to advise the kernel what life time to expect
1222 from a write. Only supported on Linux, as of version 4.13. Accepted
1223 values are:
1224
1225 **none**
1226 No particular life time associated with this file.
1227
1228 **short**
1229 Data written to this file has a short life time.
1230
1231 **medium**
1232 Data written to this file has a medium life time.
1233
1234 **long**
1235 Data written to this file has a long life time.
1236
1237 **extreme**
1238 Data written to this file has a very long life time.
1239
1240 The values are all relative to each other, and no absolute meaning
1241 should be associated with them.
f80dba8d
MT
1242
1243.. option:: offset=int
1244
82dbb8cb 1245 Start I/O at the provided offset in the file, given as either a fixed size in
83c8b093
JF
1246 bytes or a percentage. If a percentage is given, the generated offset will be
1247 aligned to the minimum ``blocksize`` or to the value of ``offset_align`` if
1248 provided. Data before the given offset will not be touched. This
89978a6b
BW
1249 effectively caps the file size at `real_size - offset`. Can be combined with
1250 :option:`size` to constrain the start and end range of the I/O workload.
44bb1142
TK
1251 A percentage can be specified by a number between 1 and 100 followed by '%',
1252 for example, ``offset=20%`` to specify 20%.
f80dba8d 1253
83c8b093
JF
1254.. option:: offset_align=int
1255
1256 If set to non-zero value, the byte offset generated by a percentage ``offset``
1257 is aligned upwards to this value. Defaults to 0 meaning that a percentage
1258 offset is aligned to the minimum block size.
1259
f80dba8d
MT
1260.. option:: offset_increment=int
1261
1262 If this is provided, then the real offset becomes `offset + offset_increment
1263 * thread_number`, where the thread number is a counter that starts at 0 and
1264 is incremented for each sub-job (i.e. when :option:`numjobs` option is
1265 specified). This option is useful if there are several jobs which are
1266 intended to operate on a file in parallel disjoint segments, with even
0b288ba1
VF
1267 spacing between the starting points. Percentages can be used for this option.
1268 If a percentage is given, the generated offset will be aligned to the minimum
1269 ``blocksize`` or to the value of ``offset_align`` if provided.
f80dba8d
MT
1270
1271.. option:: number_ios=int
1272
c60ebc45 1273 Fio will normally perform I/Os until it has exhausted the size of the region
f80dba8d
MT
1274 set by :option:`size`, or if it exhaust the allocated time (or hits an error
1275 condition). With this setting, the range/size can be set independently of
c60ebc45 1276 the number of I/Os to perform. When fio reaches this number, it will exit
f80dba8d
MT
1277 normally and report status. Note that this does not extend the amount of I/O
1278 that will be done, it will only stop fio if this condition is met before
1279 other end-of-job criteria.
1280
1281.. option:: fsync=int
1282
730bd7d9
SW
1283 If writing to a file, issue an :manpage:`fsync(2)` (or its equivalent) of
1284 the dirty data for every number of blocks given. For example, if you give 32
1285 as a parameter, fio will sync the file after every 32 writes issued. If fio is
1286 using non-buffered I/O, we may not sync the file. The exception is the sg
1287 I/O engine, which synchronizes the disk cache anyway. Defaults to 0, which
1288 means fio does not periodically issue and wait for a sync to complete. Also
1289 see :option:`end_fsync` and :option:`fsync_on_close`.
f80dba8d
MT
1290
1291.. option:: fdatasync=int
1292
1293 Like :option:`fsync` but uses :manpage:`fdatasync(2)` to only sync data and
44f668d7 1294 not metadata blocks. In Windows, FreeBSD, DragonFlyBSD or OSX there is no
730bd7d9
SW
1295 :manpage:`fdatasync(2)` so this falls back to using :manpage:`fsync(2)`.
1296 Defaults to 0, which means fio does not periodically issue and wait for a
1297 data-only sync to complete.
f80dba8d
MT
1298
1299.. option:: write_barrier=int
1300
2831be97 1301 Make every `N-th` write a barrier write.
f80dba8d 1302
f50fbdda 1303.. option:: sync_file_range=str:int
f80dba8d 1304
f50fbdda 1305 Use :manpage:`sync_file_range(2)` for every `int` number of write
f80dba8d
MT
1306 operations. Fio will track range of writes that have happened since the last
1307 :manpage:`sync_file_range(2)` call. `str` can currently be one or more of:
1308
1309 **wait_before**
1310 SYNC_FILE_RANGE_WAIT_BEFORE
1311 **write**
1312 SYNC_FILE_RANGE_WRITE
1313 **wait_after**
1314 SYNC_FILE_RANGE_WAIT_AFTER
1315
1316 So if you do ``sync_file_range=wait_before,write:8``, fio would use
1317 ``SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE`` for every 8
1318 writes. Also see the :manpage:`sync_file_range(2)` man page. This option is
1319 Linux specific.
1320
1321.. option:: overwrite=bool
1322
1323 If true, writes to a file will always overwrite existing data. If the file
1324 doesn't already exist, it will be created before the write phase begins. If
1325 the file exists and is large enough for the specified write phase, nothing
a47b697c 1326 will be done. Default: false.
f80dba8d
MT
1327
1328.. option:: end_fsync=bool
1329
a47b697c
SW
1330 If true, :manpage:`fsync(2)` file contents when a write stage has completed.
1331 Default: false.
f80dba8d
MT
1332
1333.. option:: fsync_on_close=bool
1334
1335 If true, fio will :manpage:`fsync(2)` a dirty file on close. This differs
a47b697c
SW
1336 from :option:`end_fsync` in that it will happen on every file close, not
1337 just at the end of the job. Default: false.
f80dba8d
MT
1338
1339.. option:: rwmixread=int
1340
1341 Percentage of a mixed workload that should be reads. Default: 50.
1342
1343.. option:: rwmixwrite=int
1344
1345 Percentage of a mixed workload that should be writes. If both
1346 :option:`rwmixread` and :option:`rwmixwrite` is given and the values do not
1347 add up to 100%, the latter of the two will be used to override the
1348 first. This may interfere with a given rate setting, if fio is asked to
1349 limit reads or writes to a certain rate. If that is the case, then the
1350 distribution may be skewed. Default: 50.
1351
1352.. option:: random_distribution=str:float[,str:float][,str:float]
1353
1354 By default, fio will use a completely uniform random distribution when asked
1355 to perform random I/O. Sometimes it is useful to skew the distribution in
1356 specific ways, ensuring that some parts of the data is more hot than others.
1357 fio includes the following distribution models:
1358
1359 **random**
1360 Uniform random distribution
1361
1362 **zipf**
1363 Zipf distribution
1364
1365 **pareto**
1366 Pareto distribution
1367
b2f4b559 1368 **normal**
c60ebc45 1369 Normal (Gaussian) distribution
f80dba8d
MT
1370
1371 **zoned**
1372 Zoned random distribution
1373
59466396
JA
1374 **zoned_abs**
1375 Zone absolute random distribution
1376
f80dba8d 1377 When using a **zipf** or **pareto** distribution, an input value is also
f50fbdda 1378 needed to define the access pattern. For **zipf**, this is the `Zipf
c60ebc45 1379 theta`. For **pareto**, it's the `Pareto power`. Fio includes a test
f50fbdda 1380 program, :command:`fio-genzipf`, that can be used visualize what the given input
f80dba8d
MT
1381 values will yield in terms of hit rates. If you wanted to use **zipf** with
1382 a `theta` of 1.2, you would use ``random_distribution=zipf:1.2`` as the
1383 option. If a non-uniform model is used, fio will disable use of the random
b2f4b559
SW
1384 map. For the **normal** distribution, a normal (Gaussian) deviation is
1385 supplied as a value between 0 and 100.
f80dba8d
MT
1386
1387 For a **zoned** distribution, fio supports specifying percentages of I/O
1388 access that should fall within what range of the file or device. For
1389 example, given a criteria of:
1390
f50fbdda
TK
1391 * 60% of accesses should be to the first 10%
1392 * 30% of accesses should be to the next 20%
1393 * 8% of accesses should be to the next 30%
1394 * 2% of accesses should be to the next 40%
f80dba8d
MT
1395
1396 we can define that through zoning of the random accesses. For the above
1397 example, the user would do::
1398
1399 random_distribution=zoned:60/10:30/20:8/30:2/40
1400
59466396
JA
1401 A **zoned_abs** distribution works exactly like the **zoned**, except
1402 that it takes absolute sizes. For example, let's say you wanted to
1403 define access according to the following criteria:
1404
1405 * 60% of accesses should be to the first 20G
1406 * 30% of accesses should be to the next 100G
1407 * 10% of accesses should be to the next 500G
1408
1409 we can define an absolute zoning distribution with:
1410
1411 random_distribution=zoned_abs=60/20G:30/100G:10/500g
1412
6a16ece8
JA
1413 For both **zoned** and **zoned_abs**, fio supports defining up to
1414 256 separate zones.
1415
59466396
JA
1416 Similarly to how :option:`bssplit` works for setting ranges and
1417 percentages of block sizes. Like :option:`bssplit`, it's possible to
1418 specify separate zones for reads, writes, and trims. If just one set
1419 is given, it'll apply to all of them. This goes for both **zoned**
1420 **zoned_abs** distributions.
f80dba8d
MT
1421
1422.. option:: percentage_random=int[,int][,int]
1423
1424 For a random workload, set how big a percentage should be random. This
1425 defaults to 100%, in which case the workload is fully random. It can be set
1426 from anywhere from 0 to 100. Setting it to 0 would make the workload fully
1427 sequential. Any setting in between will result in a random mix of sequential
1428 and random I/O, at the given percentages. Comma-separated values may be
1429 specified for reads, writes, and trims as described in :option:`blocksize`.
1430
1431.. option:: norandommap
1432
1433 Normally fio will cover every block of the file when doing random I/O. If
1434 this option is given, fio will just get a new random offset without looking
1435 at past I/O history. This means that some blocks may not be read or written,
1436 and that some blocks may be read/written more than once. If this option is
1437 used with :option:`verify` and multiple blocksizes (via :option:`bsrange`),
1438 only intact blocks are verified, i.e., partially-overwritten blocks are
47e6a6e5
SW
1439 ignored. With an async I/O engine and an I/O depth > 1, it is possible for
1440 the same block to be overwritten, which can cause verification errors. Either
1441 do not use norandommap in this case, or also use the lfsr random generator.
f80dba8d
MT
1442
1443.. option:: softrandommap=bool
1444
1445 See :option:`norandommap`. If fio runs with the random block map enabled and
1446 it fails to allocate the map, if this option is set it will continue without
1447 a random block map. As coverage will not be as complete as with random maps,
1448 this option is disabled by default.
1449
1450.. option:: random_generator=str
1451
f50fbdda 1452 Fio supports the following engines for generating I/O offsets for random I/O:
f80dba8d
MT
1453
1454 **tausworthe**
f50fbdda 1455 Strong 2^88 cycle random number generator.
f80dba8d 1456 **lfsr**
f50fbdda 1457 Linear feedback shift register generator.
f80dba8d 1458 **tausworthe64**
f50fbdda 1459 Strong 64-bit 2^258 cycle random number generator.
f80dba8d
MT
1460
1461 **tausworthe** is a strong random number generator, but it requires tracking
1462 on the side if we want to ensure that blocks are only read or written
f50fbdda 1463 once. **lfsr** guarantees that we never generate the same offset twice, and
f80dba8d 1464 it's also less computationally expensive. It's not a true random generator,
f50fbdda 1465 however, though for I/O purposes it's typically good enough. **lfsr** only
f80dba8d
MT
1466 works with single block sizes, not with workloads that use multiple block
1467 sizes. If used with such a workload, fio may read or write some blocks
1468 multiple times. The default value is **tausworthe**, unless the required
1469 space exceeds 2^32 blocks. If it does, then **tausworthe64** is
1470 selected automatically.
1471
1472
1473Block size
1474~~~~~~~~~~
1475
1476.. option:: blocksize=int[,int][,int], bs=int[,int][,int]
1477
1478 The block size in bytes used for I/O units. Default: 4096. A single value
1479 applies to reads, writes, and trims. Comma-separated values may be
1480 specified for reads, writes, and trims. A value not terminated in a comma
1481 applies to subsequent types.
1482
1483 Examples:
1484
1485 **bs=256k**
1486 means 256k for reads, writes and trims.
1487
1488 **bs=8k,32k**
1489 means 8k for reads, 32k for writes and trims.
1490
1491 **bs=8k,32k,**
1492 means 8k for reads, 32k for writes, and default for trims.
1493
1494 **bs=,8k**
1495 means default for reads, 8k for writes and trims.
1496
1497 **bs=,8k,**
b443ae44 1498 means default for reads, 8k for writes, and default for trims.
f80dba8d
MT
1499
1500.. option:: blocksize_range=irange[,irange][,irange], bsrange=irange[,irange][,irange]
1501
1502 A range of block sizes in bytes for I/O units. The issued I/O unit will
1503 always be a multiple of the minimum size, unless
1504 :option:`blocksize_unaligned` is set.
1505
1506 Comma-separated ranges may be specified for reads, writes, and trims as
1507 described in :option:`blocksize`.
1508
1509 Example: ``bsrange=1k-4k,2k-8k``.
1510
1511.. option:: bssplit=str[,str][,str]
1512
6a16ece8
JA
1513 Sometimes you want even finer grained control of the block sizes
1514 issued, not just an even split between them. This option allows you to
1515 weight various block sizes, so that you are able to define a specific
1516 amount of block sizes issued. The format for this option is::
f80dba8d
MT
1517
1518 bssplit=blocksize/percentage:blocksize/percentage
1519
6a16ece8
JA
1520 for as many block sizes as needed. So if you want to define a workload
1521 that has 50% 64k blocks, 10% 4k blocks, and 40% 32k blocks, you would
1522 write::
f80dba8d
MT
1523
1524 bssplit=4k/10:64k/50:32k/40
1525
6a16ece8
JA
1526 Ordering does not matter. If the percentage is left blank, fio will
1527 fill in the remaining values evenly. So a bssplit option like this one::
f80dba8d
MT
1528
1529 bssplit=4k/50:1k/:32k/
1530
6a16ece8
JA
1531 would have 50% 4k ios, and 25% 1k and 32k ios. The percentages always
1532 add up to 100, if bssplit is given a range that adds up to more, it
1533 will error out.
f80dba8d
MT
1534
1535 Comma-separated values may be specified for reads, writes, and trims as
1536 described in :option:`blocksize`.
1537
6a16ece8
JA
1538 If you want a workload that has 50% 2k reads and 50% 4k reads, while
1539 having 90% 4k writes and 10% 8k writes, you would specify::
f80dba8d 1540
cf04b906 1541 bssplit=2k/50:4k/50,4k/90:8k/10
f80dba8d 1542
6a16ece8
JA
1543 Fio supports defining up to 64 different weights for each data
1544 direction.
1545
f80dba8d
MT
1546.. option:: blocksize_unaligned, bs_unaligned
1547
1548 If set, fio will issue I/O units with any size within
1549 :option:`blocksize_range`, not just multiples of the minimum size. This
1550 typically won't work with direct I/O, as that normally requires sector
1551 alignment.
1552
589e88b7 1553.. option:: bs_is_seq_rand=bool
f80dba8d
MT
1554
1555 If this option is set, fio will use the normal read,write blocksize settings
1556 as sequential,random blocksize settings instead. Any random read or write
1557 will use the WRITE blocksize settings, and any sequential read or write will
1558 use the READ blocksize settings.
1559
1560.. option:: blockalign=int[,int][,int], ba=int[,int][,int]
1561
1562 Boundary to which fio will align random I/O units. Default:
1563 :option:`blocksize`. Minimum alignment is typically 512b for using direct
1564 I/O, though it usually depends on the hardware block size. This option is
1565 mutually exclusive with using a random map for files, so it will turn off
1566 that option. Comma-separated values may be specified for reads, writes, and
1567 trims as described in :option:`blocksize`.
1568
1569
1570Buffers and memory
1571~~~~~~~~~~~~~~~~~~
1572
1573.. option:: zero_buffers
1574
1575 Initialize buffers with all zeros. Default: fill buffers with random data.
1576
1577.. option:: refill_buffers
1578
1579 If this option is given, fio will refill the I/O buffers on every
72592780
SW
1580 submit. Only makes sense if :option:`zero_buffers` isn't specified,
1581 naturally. Defaults to being unset i.e., the buffer is only filled at
1582 init time and the data in it is reused when possible but if any of
1583 :option:`verify`, :option:`buffer_compress_percentage` or
1584 :option:`dedupe_percentage` are enabled then `refill_buffers` is also
1585 automatically enabled.
f80dba8d
MT
1586
1587.. option:: scramble_buffers=bool
1588
1589 If :option:`refill_buffers` is too costly and the target is using data
1590 deduplication, then setting this option will slightly modify the I/O buffer
1591 contents to defeat normal de-dupe attempts. This is not enough to defeat
1592 more clever block compression attempts, but it will stop naive dedupe of
1593 blocks. Default: true.
1594
1595.. option:: buffer_compress_percentage=int
1596
72592780
SW
1597 If this is set, then fio will attempt to provide I/O buffer content
1598 (on WRITEs) that compresses to the specified level. Fio does this by
1599 providing a mix of random data followed by fixed pattern data. The
1600 fixed pattern is either zeros, or the pattern specified by
1601 :option:`buffer_pattern`. If the `buffer_pattern` option is used, it
1602 might skew the compression ratio slightly. Setting
1603 `buffer_compress_percentage` to a value other than 100 will also
1604 enable :option:`refill_buffers` in order to reduce the likelihood that
1605 adjacent blocks are so similar that they over compress when seen
1606 together. See :option:`buffer_compress_chunk` for how to set a finer or
1607 coarser granularity for the random/fixed data region. Defaults to unset
1608 i.e., buffer data will not adhere to any compression level.
f80dba8d
MT
1609
1610.. option:: buffer_compress_chunk=int
1611
72592780
SW
1612 This setting allows fio to manage how big the random/fixed data region
1613 is when using :option:`buffer_compress_percentage`. When
1614 `buffer_compress_chunk` is set to some non-zero value smaller than the
1615 block size, fio can repeat the random/fixed region throughout the I/O
1616 buffer at the specified interval (which particularly useful when
1617 bigger block sizes are used for a job). When set to 0, fio will use a
1618 chunk size that matches the block size resulting in a single
1619 random/fixed region within the I/O buffer. Defaults to 512. When the
1620 unit is omitted, the value is interpreted in bytes.
f80dba8d
MT
1621
1622.. option:: buffer_pattern=str
1623
a1554f65
SB
1624 If set, fio will fill the I/O buffers with this pattern or with the contents
1625 of a file. If not set, the contents of I/O buffers are defined by the other
1626 options related to buffer contents. The setting can be any pattern of bytes,
1627 and can be prefixed with 0x for hex values. It may also be a string, where
1628 the string must then be wrapped with ``""``. Or it may also be a filename,
1629 where the filename must be wrapped with ``''`` in which case the file is
1630 opened and read. Note that not all the file contents will be read if that
1631 would cause the buffers to overflow. So, for example::
1632
1633 buffer_pattern='filename'
1634
1635 or::
f80dba8d
MT
1636
1637 buffer_pattern="abcd"
1638
1639 or::
1640
1641 buffer_pattern=-12
1642
1643 or::
1644
1645 buffer_pattern=0xdeadface
1646
1647 Also you can combine everything together in any order::
1648
a1554f65 1649 buffer_pattern=0xdeadface"abcd"-12'filename'
f80dba8d
MT
1650
1651.. option:: dedupe_percentage=int
1652
1653 If set, fio will generate this percentage of identical buffers when
1654 writing. These buffers will be naturally dedupable. The contents of the
1655 buffers depend on what other buffer compression settings have been set. It's
1656 possible to have the individual buffers either fully compressible, or not at
72592780
SW
1657 all -- this option only controls the distribution of unique buffers. Setting
1658 this option will also enable :option:`refill_buffers` to prevent every buffer
1659 being identical.
f80dba8d
MT
1660
1661.. option:: invalidate=bool
1662
730bd7d9
SW
1663 Invalidate the buffer/page cache parts of the files to be used prior to
1664 starting I/O if the platform and file type support it. Defaults to true.
21c1b29e
TK
1665 This will be ignored if :option:`pre_read` is also specified for the
1666 same job.
f80dba8d
MT
1667
1668.. option:: sync=bool
1669
1670 Use synchronous I/O for buffered writes. For the majority of I/O engines,
1671 this means using O_SYNC. Default: false.
1672
1673.. option:: iomem=str, mem=str
1674
1675 Fio can use various types of memory as the I/O unit buffer. The allowed
1676 values are:
1677
1678 **malloc**
1679 Use memory from :manpage:`malloc(3)` as the buffers. Default memory
1680 type.
1681
1682 **shm**
1683 Use shared memory as the buffers. Allocated through
1684 :manpage:`shmget(2)`.
1685
1686 **shmhuge**
1687 Same as shm, but use huge pages as backing.
1688
1689 **mmap**
22413915 1690 Use :manpage:`mmap(2)` to allocate buffers. May either be anonymous memory, or can
f80dba8d
MT
1691 be file backed if a filename is given after the option. The format
1692 is `mem=mmap:/path/to/file`.
1693
1694 **mmaphuge**
1695 Use a memory mapped huge file as the buffer backing. Append filename
1696 after mmaphuge, ala `mem=mmaphuge:/hugetlbfs/file`.
1697
1698 **mmapshared**
1699 Same as mmap, but use a MMAP_SHARED mapping.
1700
03553853
YR
1701 **cudamalloc**
1702 Use GPU memory as the buffers for GPUDirect RDMA benchmark.
f50fbdda 1703 The :option:`ioengine` must be `rdma`.
03553853 1704
f80dba8d
MT
1705 The area allocated is a function of the maximum allowed bs size for the job,
1706 multiplied by the I/O depth given. Note that for **shmhuge** and
1707 **mmaphuge** to work, the system must have free huge pages allocated. This
1708 can normally be checked and set by reading/writing
1709 :file:`/proc/sys/vm/nr_hugepages` on a Linux system. Fio assumes a huge page
1710 is 4MiB in size. So to calculate the number of huge pages you need for a
1711 given job file, add up the I/O depth of all jobs (normally one unless
1712 :option:`iodepth` is used) and multiply by the maximum bs set. Then divide
1713 that number by the huge page size. You can see the size of the huge pages in
1714 :file:`/proc/meminfo`. If no huge pages are allocated by having a non-zero
1715 number in `nr_hugepages`, using **mmaphuge** or **shmhuge** will fail. Also
1716 see :option:`hugepage-size`.
1717
1718 **mmaphuge** also needs to have hugetlbfs mounted and the file location
1719 should point there. So if it's mounted in :file:`/huge`, you would use
1720 `mem=mmaphuge:/huge/somefile`.
1721
f50fbdda 1722.. option:: iomem_align=int, mem_align=int
f80dba8d
MT
1723
1724 This indicates the memory alignment of the I/O memory buffers. Note that
1725 the given alignment is applied to the first I/O unit buffer, if using
1726 :option:`iodepth` the alignment of the following buffers are given by the
1727 :option:`bs` used. In other words, if using a :option:`bs` that is a
1728 multiple of the page sized in the system, all buffers will be aligned to
1729 this value. If using a :option:`bs` that is not page aligned, the alignment
1730 of subsequent I/O memory buffers is the sum of the :option:`iomem_align` and
1731 :option:`bs` used.
1732
1733.. option:: hugepage-size=int
1734
1735 Defines the size of a huge page. Must at least be equal to the system
1736 setting, see :file:`/proc/meminfo`. Defaults to 4MiB. Should probably
1737 always be a multiple of megabytes, so using ``hugepage-size=Xm`` is the
1738 preferred way to set this to avoid setting a non-pow-2 bad value.
1739
1740.. option:: lockmem=int
1741
1742 Pin the specified amount of memory with :manpage:`mlock(2)`. Can be used to
1743 simulate a smaller amount of memory. The amount specified is per worker.
1744
1745
1746I/O size
1747~~~~~~~~
1748
1749.. option:: size=int
1750
79591fa9
TK
1751 The total size of file I/O for each thread of this job. Fio will run until
1752 this many bytes has been transferred, unless runtime is limited by other options
1753 (such as :option:`runtime`, for instance, or increased/decreased by :option:`io_size`).
1754 Fio will divide this size between the available files determined by options
1755 such as :option:`nrfiles`, :option:`filename`, unless :option:`filesize` is
1756 specified by the job. If the result of division happens to be 0, the size is
c4aa2d08 1757 set to the physical size of the given files or devices if they exist.
79591fa9 1758 If this option is not specified, fio will use the full size of the given
f80dba8d
MT
1759 files or devices. If the files do not exist, size must be given. It is also
1760 possible to give size as a percentage between 1 and 100. If ``size=20%`` is
1761 given, fio will use 20% of the full size of the given files or devices.
9d25d068
SW
1762 Can be combined with :option:`offset` to constrain the start and end range
1763 that I/O will be done within.
f80dba8d
MT
1764
1765.. option:: io_size=int, io_limit=int
1766
1767 Normally fio operates within the region set by :option:`size`, which means
1768 that the :option:`size` option sets both the region and size of I/O to be
1769 performed. Sometimes that is not what you want. With this option, it is
1770 possible to define just the amount of I/O that fio should do. For instance,
1771 if :option:`size` is set to 20GiB and :option:`io_size` is set to 5GiB, fio
1772 will perform I/O within the first 20GiB but exit when 5GiB have been
1773 done. The opposite is also possible -- if :option:`size` is set to 20GiB,
1774 and :option:`io_size` is set to 40GiB, then fio will do 40GiB of I/O within
1775 the 0..20GiB region.
1776
7fdd97ca 1777.. option:: filesize=irange(int)
f80dba8d
MT
1778
1779 Individual file sizes. May be a range, in which case fio will select sizes
1780 for files at random within the given range and limited to :option:`size` in
1781 total (if that is given). If not given, each created file is the same size.
79591fa9
TK
1782 This option overrides :option:`size` in terms of file size, which means
1783 this value is used as a fixed size or possible range of each file.
f80dba8d
MT
1784
1785.. option:: file_append=bool
1786
1787 Perform I/O after the end of the file. Normally fio will operate within the
1788 size of a file. If this option is set, then fio will append to the file
1789 instead. This has identical behavior to setting :option:`offset` to the size
1790 of a file. This option is ignored on non-regular files.
1791
1792.. option:: fill_device=bool, fill_fs=bool
1793
1794 Sets size to something really large and waits for ENOSPC (no space left on
1795 device) as the terminating condition. Only makes sense with sequential
1796 write. For a read workload, the mount point will be filled first then I/O
1797 started on the result. This option doesn't make sense if operating on a raw
1798 device node, since the size of that is already known by the file system.
1799 Additionally, writing beyond end-of-device will not return ENOSPC there.
1800
1801
1802I/O engine
1803~~~~~~~~~~
1804
1805.. option:: ioengine=str
1806
1807 Defines how the job issues I/O to the file. The following types are defined:
1808
1809 **sync**
1810 Basic :manpage:`read(2)` or :manpage:`write(2)`
1811 I/O. :manpage:`lseek(2)` is used to position the I/O location.
54227e6b 1812 See :option:`fsync` and :option:`fdatasync` for syncing write I/Os.
f80dba8d
MT
1813
1814 **psync**
1815 Basic :manpage:`pread(2)` or :manpage:`pwrite(2)` I/O. Default on
1816 all supported operating systems except for Windows.
1817
1818 **vsync**
1819 Basic :manpage:`readv(2)` or :manpage:`writev(2)` I/O. Will emulate
c60ebc45 1820 queuing by coalescing adjacent I/Os into a single submission.
f80dba8d
MT
1821
1822 **pvsync**
1823 Basic :manpage:`preadv(2)` or :manpage:`pwritev(2)` I/O.
1824
1825 **pvsync2**
1826 Basic :manpage:`preadv2(2)` or :manpage:`pwritev2(2)` I/O.
1827
029b42ac
JA
1828 **io_uring**
1829 Fast Linux native asynchronous I/O. Supports async IO
1830 for both direct and buffered IO.
1831 This engine defines engine specific options.
1832
f80dba8d
MT
1833 **libaio**
1834 Linux native asynchronous I/O. Note that Linux may only support
22413915 1835 queued behavior with non-buffered I/O (set ``direct=1`` or
f80dba8d
MT
1836 ``buffered=0``).
1837 This engine defines engine specific options.
1838
1839 **posixaio**
1840 POSIX asynchronous I/O using :manpage:`aio_read(3)` and
1841 :manpage:`aio_write(3)`.
1842
1843 **solarisaio**
1844 Solaris native asynchronous I/O.
1845
1846 **windowsaio**
1847 Windows native asynchronous I/O. Default on Windows.
1848
1849 **mmap**
1850 File is memory mapped with :manpage:`mmap(2)` and data copied
1851 to/from using :manpage:`memcpy(3)`.
1852
1853 **splice**
1854 :manpage:`splice(2)` is used to transfer the data and
1855 :manpage:`vmsplice(2)` to transfer data from user space to the
1856 kernel.
1857
1858 **sg**
1859 SCSI generic sg v3 I/O. May either be synchronous using the SG_IO
1860 ioctl, or if the target is an sg character device we use
1861 :manpage:`read(2)` and :manpage:`write(2)` for asynchronous
f50fbdda 1862 I/O. Requires :option:`filename` option to specify either block or
3740cfc8 1863 character devices. This engine supports trim operations.
52b81b7c 1864 The sg engine includes engine specific options.
f80dba8d
MT
1865
1866 **null**
1867 Doesn't transfer any data, just pretends to. This is mainly used to
1868 exercise fio itself and for debugging/testing purposes.
1869
1870 **net**
1871 Transfer over the network to given ``host:port``. Depending on the
1872 :option:`protocol` used, the :option:`hostname`, :option:`port`,
1873 :option:`listen` and :option:`filename` options are used to specify
1874 what sort of connection to make, while the :option:`protocol` option
1875 determines which protocol will be used. This engine defines engine
1876 specific options.
1877
1878 **netsplice**
1879 Like **net**, but uses :manpage:`splice(2)` and
1880 :manpage:`vmsplice(2)` to map data and send/receive.
1881 This engine defines engine specific options.
1882
1883 **cpuio**
1884 Doesn't transfer any data, but burns CPU cycles according to the
1885 :option:`cpuload` and :option:`cpuchunks` options. Setting
9207a0cb 1886 :option:`cpuload`\=85 will cause that job to do nothing but burn 85%
71aa48eb 1887 of the CPU. In case of SMP machines, use :option:`numjobs`\=<nr_of_cpu>
f50fbdda 1888 to get desired CPU usage, as the cpuload only loads a
f80dba8d
MT
1889 single CPU at the desired rate. A job never finishes unless there is
1890 at least one non-cpuio job.
1891
1892 **guasi**
804c0839 1893 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall
f80dba8d
MT
1894 Interface approach to async I/O. See
1895
1896 http://www.xmailserver.org/guasi-lib.html
1897
1898 for more info on GUASI.
1899
1900 **rdma**
1901 The RDMA I/O engine supports both RDMA memory semantics
1902 (RDMA_WRITE/RDMA_READ) and channel semantics (Send/Recv) for the
609ac152
SB
1903 InfiniBand, RoCE and iWARP protocols. This engine defines engine
1904 specific options.
f80dba8d
MT
1905
1906 **falloc**
1907 I/O engine that does regular fallocate to simulate data transfer as
1908 fio ioengine.
1909
1910 DDIR_READ
1911 does fallocate(,mode = FALLOC_FL_KEEP_SIZE,).
1912
1913 DDIR_WRITE
1914 does fallocate(,mode = 0).
1915
1916 DDIR_TRIM
1917 does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE).
1918
761cd093
SW
1919 **ftruncate**
1920 I/O engine that sends :manpage:`ftruncate(2)` operations in response
1921 to write (DDIR_WRITE) events. Each ftruncate issued sets the file's
f50fbdda 1922 size to the current block offset. :option:`blocksize` is ignored.
761cd093 1923
f80dba8d
MT
1924 **e4defrag**
1925 I/O engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate
1926 defragment activity in request to DDIR_WRITE event.
1927
f3f96717
IF
1928 **rados**
1929 I/O engine supporting direct access to Ceph Reliable Autonomic
1930 Distributed Object Store (RADOS) via librados. This ioengine
1931 defines engine specific options.
1932
f80dba8d
MT
1933 **rbd**
1934 I/O engine supporting direct access to Ceph Rados Block Devices
1935 (RBD) via librbd without the need to use the kernel rbd driver. This
1936 ioengine defines engine specific options.
1937
c2f6a13d
LMB
1938 **http**
1939 I/O engine supporting GET/PUT requests over HTTP(S) with libcurl to
1940 a WebDAV or S3 endpoint. This ioengine defines engine specific options.
1941
1942 This engine only supports direct IO of iodepth=1; you need to scale this
1943 via numjobs. blocksize defines the size of the objects to be created.
1944
1945 TRIM is translated to object deletion.
1946
f80dba8d 1947 **gfapi**
ac8ca2af
SW
1948 Using GlusterFS libgfapi sync interface to direct access to
1949 GlusterFS volumes without having to go through FUSE. This ioengine
f80dba8d
MT
1950 defines engine specific options.
1951
1952 **gfapi_async**
ac8ca2af
SW
1953 Using GlusterFS libgfapi async interface to direct access to
1954 GlusterFS volumes without having to go through FUSE. This ioengine
f80dba8d
MT
1955 defines engine specific options.
1956
1957 **libhdfs**
f50fbdda 1958 Read and write through Hadoop (HDFS). The :option:`filename` option
f80dba8d
MT
1959 is used to specify host,port of the hdfs name-node to connect. This
1960 engine interprets offsets a little differently. In HDFS, files once
e25c0c91
SW
1961 created cannot be modified so random writes are not possible. To
1962 imitate this the libhdfs engine expects a bunch of small files to be
1963 created over HDFS and will randomly pick a file from them
1964 based on the offset generated by fio backend (see the example
f80dba8d 1965 job file to create such files, use ``rw=write`` option). Please
e25c0c91
SW
1966 note, it may be necessary to set environment variables to work
1967 with HDFS/libhdfs properly. Each job uses its own connection to
f80dba8d
MT
1968 HDFS.
1969
1970 **mtd**
1971 Read, write and erase an MTD character device (e.g.,
1972 :file:`/dev/mtd0`). Discards are treated as erases. Depending on the
1973 underlying device type, the I/O may have to go in a certain pattern,
1974 e.g., on NAND, writing sequentially to erase blocks and discarding
c298ee71 1975 before overwriting. The `trimwrite` mode works well for this
f80dba8d
MT
1976 constraint.
1977
1978 **pmemblk**
1979 Read and write using filesystem DAX to a file on a filesystem
363a5f65 1980 mounted with DAX on a persistent memory device through the PMDK
f80dba8d
MT
1981 libpmemblk library.
1982
1983 **dev-dax**
1984 Read and write using device DAX to a persistent memory device (e.g.,
363a5f65 1985 /dev/dax0.0) through the PMDK libpmem library.
f80dba8d
MT
1986
1987 **external**
1988 Prefix to specify loading an external I/O engine object file. Append
c60ebc45 1989 the engine filename, e.g. ``ioengine=external:/tmp/foo.o`` to load
d243fd6d
TK
1990 ioengine :file:`foo.o` in :file:`/tmp`. The path can be either
1991 absolute or relative. See :file:`engines/skeleton_external.c` for
1992 details of writing an external I/O engine.
f80dba8d 1993
1216cc5a 1994 **filecreate**
b71968b1 1995 Simply create the files and do no I/O to them. You still need to
1216cc5a 1996 set `filesize` so that all the accounting still occurs, but no
b71968b1 1997 actual I/O will be done other than creating the file.
f80dba8d 1998
73ccd14e
SF
1999 **filestat**
2000 Simply do stat() and do no I/O to the file. You need to set 'filesize'
2001 and 'nrfiles', so that files will be created.
2002 This engine is to measure file lookup and meta data access.
2003
ae0db592
TI
2004 **libpmem**
2005 Read and write using mmap I/O to a file on a filesystem
363a5f65 2006 mounted with DAX on a persistent memory device through the PMDK
ae0db592
TI
2007 libpmem library.
2008
a40e7a59
GB
2009 **ime_psync**
2010 Synchronous read and write using DDN's Infinite Memory Engine (IME).
2011 This engine is very basic and issues calls to IME whenever an IO is
2012 queued.
2013
2014 **ime_psyncv**
2015 Synchronous read and write using DDN's Infinite Memory Engine (IME).
2016 This engine uses iovecs and will try to stack as much IOs as possible
2017 (if the IOs are "contiguous" and the IO depth is not exceeded)
2018 before issuing a call to IME.
2019
2020 **ime_aio**
2021 Asynchronous read and write using DDN's Infinite Memory Engine (IME).
2022 This engine will try to stack as much IOs as possible by creating
2023 requests for IME. FIO will then decide when to commit these requests.
247ef2aa
KZ
2024 **libiscsi**
2025 Read and write iscsi lun with libiscsi.
d643a1e2 2026 **nbd**
f2d6de5d 2027 Read and write a Network Block Device (NBD).
a40e7a59 2028
f80dba8d
MT
2029I/O engine specific parameters
2030~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2031
2032In addition, there are some parameters which are only valid when a specific
f50fbdda
TK
2033:option:`ioengine` is in use. These are used identically to normal parameters,
2034with the caveat that when used on the command line, they must come after the
f80dba8d
MT
2035:option:`ioengine` that defines them is selected.
2036
b2a432bf 2037.. option:: cmdprio_percentage=int : [io_uring] [libaio]
029b42ac 2038
b2a432bf
PC
2039 Set the percentage of I/O that will be issued with higher priority by setting
2040 the priority bit. Non-read I/O is likely unaffected by ``cmdprio_percentage``.
2041 This option cannot be used with the `prio` or `prioclass` options. For this
2042 option to set the priority bit properly, NCQ priority must be supported and
7896180a
VF
2043 enabled and :option:`direct`\=1 option must be used. fio must also be run as
2044 the root user.
029b42ac
JA
2045
2046.. option:: fixedbufs : [io_uring]
2047
b2a432bf
PC
2048 If fio is asked to do direct IO, then Linux will map pages for each
2049 IO call, and release them when IO is done. If this option is set, the
2050 pages are pre-mapped before IO is started. This eliminates the need to
2051 map and release for each IO. This is more efficient, and reduces the
2052 IO latency as well.
2053
2054.. option:: hipri : [io_uring]
2055
2056 If this option is set, fio will attempt to use polled IO completions.
2057 Normal IO completions generate interrupts to signal the completion of
2058 IO, polled completions do not. Hence they are require active reaping
2059 by the application. The benefits are more efficient IO for high IOPS
2060 scenarios, and lower latencies for low queue depth IO.
029b42ac 2061
5ffd5626 2062.. option:: registerfiles : [io_uring]
2c870598 2063
5ffd5626
JA
2064 With this option, fio registers the set of files being used with the
2065 kernel. This avoids the overhead of managing file counts in the kernel,
2066 making the submission and completion part more lightweight. Required
2067 for the below :option:`sqthread_poll` option.
2068
029b42ac
JA
2069.. option:: sqthread_poll : [io_uring]
2070
2071 Normally fio will submit IO by issuing a system call to notify the
2072 kernel of available items in the SQ ring. If this option is set, the
2073 act of submitting IO will be done by a polling thread in the kernel.
2074 This frees up cycles for fio, at the cost of using more CPU in the
2075 system.
2076
2077.. option:: sqthread_poll_cpu : [io_uring]
2078
2079 When :option:`sqthread_poll` is set, this option provides a way to
2080 define which CPU should be used for the polling thread.
2081
f80dba8d
MT
2082.. option:: userspace_reap : [libaio]
2083
2084 Normally, with the libaio engine in use, fio will use the
2085 :manpage:`io_getevents(2)` system call to reap newly returned events. With
2086 this flag turned on, the AIO ring will be read directly from user-space to
2087 reap events. The reaping mode is only enabled when polling for a minimum of
c60ebc45 2088 0 events (e.g. when :option:`iodepth_batch_complete` `=0`).
f80dba8d 2089
9d25d068 2090.. option:: hipri : [pvsync2]
f80dba8d
MT
2091
2092 Set RWF_HIPRI on I/O, indicating to the kernel that it's of higher priority
2093 than normal.
2094
a0679ce5
SB
2095.. option:: hipri_percentage : [pvsync2]
2096
f50fbdda 2097 When hipri is set this determines the probability of a pvsync2 I/O being high
a0679ce5
SB
2098 priority. The default is 100%.
2099
f80dba8d
MT
2100.. option:: cpuload=int : [cpuio]
2101
da19cdb4
TK
2102 Attempt to use the specified percentage of CPU cycles. This is a mandatory
2103 option when using cpuio I/O engine.
f80dba8d
MT
2104
2105.. option:: cpuchunks=int : [cpuio]
2106
2107 Split the load into cycles of the given time. In microseconds.
2108
2109.. option:: exit_on_io_done=bool : [cpuio]
2110
2111 Detect when I/O threads are done, then exit.
2112
f80dba8d
MT
2113.. option:: namenode=str : [libhdfs]
2114
22413915 2115 The hostname or IP address of a HDFS cluster namenode to contact.
f80dba8d
MT
2116
2117.. option:: port=int
2118
f50fbdda
TK
2119 [libhdfs]
2120
2121 The listening port of the HFDS cluster namenode.
2122
f80dba8d
MT
2123 [netsplice], [net]
2124
2125 The TCP or UDP port to bind to or connect to. If this is used with
2126 :option:`numjobs` to spawn multiple instances of the same job type, then
2127 this will be the starting port number since fio will use a range of
2128 ports.
2129
609ac152
SB
2130 [rdma]
2131
2132 The port to use for RDMA-CM communication. This should be the same value
2133 on the client and the server side.
2134
2135.. option:: hostname=str : [netsplice] [net] [rdma]
f80dba8d 2136
609ac152
SB
2137 The hostname or IP address to use for TCP, UDP or RDMA-CM based I/O. If the job
2138 is a TCP listener or UDP reader, the hostname is not used and must be omitted
f50fbdda 2139 unless it is a valid UDP multicast address.
f80dba8d
MT
2140
2141.. option:: interface=str : [netsplice] [net]
2142
2143 The IP address of the network interface used to send or receive UDP
2144 multicast.
2145
2146.. option:: ttl=int : [netsplice] [net]
2147
2148 Time-to-live value for outgoing UDP multicast packets. Default: 1.
2149
2150.. option:: nodelay=bool : [netsplice] [net]
2151
2152 Set TCP_NODELAY on TCP connections.
2153
f50fbdda 2154.. option:: protocol=str, proto=str : [netsplice] [net]
f80dba8d
MT
2155
2156 The network protocol to use. Accepted values are:
2157
2158 **tcp**
2159 Transmission control protocol.
2160 **tcpv6**
2161 Transmission control protocol V6.
2162 **udp**
2163 User datagram protocol.
2164 **udpv6**
2165 User datagram protocol V6.
2166 **unix**
2167 UNIX domain socket.
2168
2169 When the protocol is TCP or UDP, the port must also be given, as well as the
2170 hostname if the job is a TCP listener or UDP reader. For unix sockets, the
f50fbdda 2171 normal :option:`filename` option should be used and the port is invalid.
f80dba8d 2172
e9184ec1 2173.. option:: listen : [netsplice] [net]
f80dba8d
MT
2174
2175 For TCP network connections, tell fio to listen for incoming connections
2176 rather than initiating an outgoing connection. The :option:`hostname` must
2177 be omitted if this option is used.
2178
e9184ec1 2179.. option:: pingpong : [netsplice] [net]
f80dba8d
MT
2180
2181 Normally a network writer will just continue writing data, and a network
2182 reader will just consume packages. If ``pingpong=1`` is set, a writer will
2183 send its normal payload to the reader, then wait for the reader to send the
2184 same payload back. This allows fio to measure network latencies. The
2185 submission and completion latencies then measure local time spent sending or
2186 receiving, and the completion latency measures how long it took for the
2187 other end to receive and send back. For UDP multicast traffic
2188 ``pingpong=1`` should only be set for a single reader when multiple readers
2189 are listening to the same address.
2190
e9184ec1 2191.. option:: window_size : [netsplice] [net]
f80dba8d
MT
2192
2193 Set the desired socket buffer size for the connection.
2194
e9184ec1 2195.. option:: mss : [netsplice] [net]
f80dba8d
MT
2196
2197 Set the TCP maximum segment size (TCP_MAXSEG).
2198
2199.. option:: donorname=str : [e4defrag]
2200
730bd7d9 2201 File will be used as a block donor (swap extents between files).
f80dba8d
MT
2202
2203.. option:: inplace=int : [e4defrag]
2204
2205 Configure donor file blocks allocation strategy:
2206
2207 **0**
2208 Default. Preallocate donor's file on init.
2209 **1**
2b455dbf 2210 Allocate space immediately inside defragment event, and free right
f80dba8d
MT
2211 after event.
2212
f3f96717 2213.. option:: clustername=str : [rbd,rados]
f80dba8d
MT
2214
2215 Specifies the name of the Ceph cluster.
2216
2217.. option:: rbdname=str : [rbd]
2218
2219 Specifies the name of the RBD.
2220
f3f96717 2221.. option:: pool=str : [rbd,rados]
f80dba8d 2222
f3f96717 2223 Specifies the name of the Ceph pool containing RBD or RADOS data.
f80dba8d 2224
f3f96717 2225.. option:: clientname=str : [rbd,rados]
f80dba8d
MT
2226
2227 Specifies the username (without the 'client.' prefix) used to access the
2228 Ceph cluster. If the *clustername* is specified, the *clientname* shall be
2229 the full *type.id* string. If no type. prefix is given, fio will add
2230 'client.' by default.
2231
f3f96717
IF
2232.. option:: busy_poll=bool : [rbd,rados]
2233
2234 Poll store instead of waiting for completion. Usually this provides better
2235 throughput at cost of higher(up to 100%) CPU utilization.
2236
f80dba8d
MT
2237.. option:: skip_bad=bool : [mtd]
2238
2239 Skip operations against known bad blocks.
2240
2241.. option:: hdfsdirectory : [libhdfs]
2242
2243 libhdfs will create chunk in this HDFS directory.
2244
2245.. option:: chunk_size : [libhdfs]
2246
2b455dbf 2247 The size of the chunk to use for each file.
f80dba8d 2248
609ac152
SB
2249.. option:: verb=str : [rdma]
2250
2251 The RDMA verb to use on this side of the RDMA ioengine connection. Valid
2252 values are write, read, send and recv. These correspond to the equivalent
2253 RDMA verbs (e.g. write = rdma_write etc.). Note that this only needs to be
2254 specified on the client side of the connection. See the examples folder.
2255
2256.. option:: bindname=str : [rdma]
2257
2258 The name to use to bind the local RDMA-CM connection to a local RDMA device.
2259 This could be a hostname or an IPv4 or IPv6 address. On the server side this
2260 will be passed into the rdma_bind_addr() function and on the client site it
2261 will be used in the rdma_resolve_add() function. This can be useful when
2262 multiple paths exist between the client and the server or in certain loopback
2263 configurations.
f80dba8d 2264
93a13ba5 2265.. option:: stat_type=str : [filestat]
c446eff0 2266
93a13ba5
TK
2267 Specify stat system call type to measure lookup/getattr performance.
2268 Default is **stat** for :manpage:`stat(2)`.
c446eff0 2269
52b81b7c
KD
2270.. option:: readfua=bool : [sg]
2271
2272 With readfua option set to 1, read operations include
2273 the force unit access (fua) flag. Default is 0.
2274
2275.. option:: writefua=bool : [sg]
2276
2277 With writefua option set to 1, write operations include
2278 the force unit access (fua) flag. Default is 0.
2279
2c3a9150 2280.. option:: sg_write_mode=str : [sg]
3740cfc8 2281
2c3a9150
VF
2282 Specify the type of write commands to issue. This option can take three values:
2283
2284 **write**
2285 This is the default where write opcodes are issued as usual.
2286 **verify**
2287 Issue WRITE AND VERIFY commands. The BYTCHK bit is set to 0. This
2288 directs the device to carry out a medium verification with no data
2289 comparison. The writefua option is ignored with this selection.
2290 **same**
2291 Issue WRITE SAME commands. This transfers a single block to the device
2292 and writes this same block of data to a contiguous sequence of LBAs
2293 beginning at the specified offset. fio's block size parameter specifies
2294 the amount of data written with each command. However, the amount of data
2295 actually transferred to the device is equal to the device's block
2296 (sector) size. For a device with 512 byte sectors, blocksize=8k will
2297 write 16 sectors with each command. fio will still generate 8k of data
2298 for each command but only the first 512 bytes will be used and
2299 transferred to the device. The writefua option is ignored with this
2300 selection.
52b81b7c 2301
c2f6a13d
LMB
2302.. option:: http_host=str : [http]
2303
2304 Hostname to connect to. For S3, this could be the bucket hostname.
2305 Default is **localhost**
2306
2307.. option:: http_user=str : [http]
2308
2309 Username for HTTP authentication.
2310
2311.. option:: http_pass=str : [http]
2312
2313 Password for HTTP authentication.
2314
09fd2966 2315.. option:: https=str : [http]
c2f6a13d 2316
09fd2966
LMB
2317 Enable HTTPS instead of http. *on* enables HTTPS; *insecure*
2318 will enable HTTPS, but disable SSL peer verification (use with
2319 caution!). Default is **off**
c2f6a13d 2320
09fd2966 2321.. option:: http_mode=str : [http]
c2f6a13d 2322
09fd2966
LMB
2323 Which HTTP access mode to use: *webdav*, *swift*, or *s3*.
2324 Default is **webdav**
c2f6a13d
LMB
2325
2326.. option:: http_s3_region=str : [http]
2327
2328 The S3 region/zone string.
2329 Default is **us-east-1**
2330
2331.. option:: http_s3_key=str : [http]
2332
2333 The S3 secret key.
2334
2335.. option:: http_s3_keyid=str : [http]
2336
2337 The S3 key/access id.
2338
09fd2966
LMB
2339.. option:: http_swift_auth_token=str : [http]
2340
2341 The Swift auth token. See the example configuration file on how
2342 to retrieve this.
2343
c2f6a13d
LMB
2344.. option:: http_verbose=int : [http]
2345
2346 Enable verbose requests from libcurl. Useful for debugging. 1
2347 turns on verbose logging from libcurl, 2 additionally enables
2348 HTTP IO tracing. Default is **0**
2349
f2d6de5d
RJ
2350.. option:: uri=str : [nbd]
2351
2352 Specify the NBD URI of the server to test. The string
2353 is a standard NBD URI
2354 (see https://github.com/NetworkBlockDevice/nbd/tree/master/doc).
2355 Example URIs: nbd://localhost:10809
2356 nbd+unix:///?socket=/tmp/socket
2357 nbds://tlshost/exportname
2358
f80dba8d
MT
2359I/O depth
2360~~~~~~~~~
2361
2362.. option:: iodepth=int
2363
2364 Number of I/O units to keep in flight against the file. Note that
2365 increasing *iodepth* beyond 1 will not affect synchronous ioengines (except
c60ebc45 2366 for small degrees when :option:`verify_async` is in use). Even async
f80dba8d
MT
2367 engines may impose OS restrictions causing the desired depth not to be
2368 achieved. This may happen on Linux when using libaio and not setting
9207a0cb 2369 :option:`direct`\=1, since buffered I/O is not async on that OS. Keep an
f80dba8d
MT
2370 eye on the I/O depth distribution in the fio output to verify that the
2371 achieved depth is as expected. Default: 1.
2372
2373.. option:: iodepth_batch_submit=int, iodepth_batch=int
2374
2375 This defines how many pieces of I/O to submit at once. It defaults to 1
2376 which means that we submit each I/O as soon as it is available, but can be
2377 raised to submit bigger batches of I/O at the time. If it is set to 0 the
2378 :option:`iodepth` value will be used.
2379
2380.. option:: iodepth_batch_complete_min=int, iodepth_batch_complete=int
2381
2382 This defines how many pieces of I/O to retrieve at once. It defaults to 1
2383 which means that we'll ask for a minimum of 1 I/O in the retrieval process
2384 from the kernel. The I/O retrieval will go on until we hit the limit set by
2385 :option:`iodepth_low`. If this variable is set to 0, then fio will always
2386 check for completed events before queuing more I/O. This helps reduce I/O
2387 latency, at the cost of more retrieval system calls.
2388
2389.. option:: iodepth_batch_complete_max=int
2390
2391 This defines maximum pieces of I/O to retrieve at once. This variable should
9207a0cb 2392 be used along with :option:`iodepth_batch_complete_min`\=int variable,
f80dba8d 2393 specifying the range of min and max amount of I/O which should be
730bd7d9 2394 retrieved. By default it is equal to the :option:`iodepth_batch_complete_min`
f80dba8d
MT
2395 value.
2396
2397 Example #1::
2398
2399 iodepth_batch_complete_min=1
2400 iodepth_batch_complete_max=<iodepth>
2401
2402 which means that we will retrieve at least 1 I/O and up to the whole
2403 submitted queue depth. If none of I/O has been completed yet, we will wait.
2404
2405 Example #2::
2406
2407 iodepth_batch_complete_min=0
2408 iodepth_batch_complete_max=<iodepth>
2409
2410 which means that we can retrieve up to the whole submitted queue depth, but
2411 if none of I/O has been completed yet, we will NOT wait and immediately exit
2412 the system call. In this example we simply do polling.
2413
2414.. option:: iodepth_low=int
2415
2416 The low water mark indicating when to start filling the queue
2417 again. Defaults to the same as :option:`iodepth`, meaning that fio will
2418 attempt to keep the queue full at all times. If :option:`iodepth` is set to
c60ebc45 2419 e.g. 16 and *iodepth_low* is set to 4, then after fio has filled the queue of
f80dba8d
MT
2420 16 requests, it will let the depth drain down to 4 before starting to fill
2421 it again.
2422
997b5680
SW
2423.. option:: serialize_overlap=bool
2424
2425 Serialize in-flight I/Os that might otherwise cause or suffer from data races.
2426 When two or more I/Os are submitted simultaneously, there is no guarantee that
2427 the I/Os will be processed or completed in the submitted order. Further, if
2428 two or more of those I/Os are writes, any overlapping region between them can
2429 become indeterminate/undefined on certain storage. These issues can cause
2430 verification to fail erratically when at least one of the racing I/Os is
2431 changing data and the overlapping region has a non-zero size. Setting
2432 ``serialize_overlap`` tells fio to avoid provoking this behavior by explicitly
2433 serializing in-flight I/Os that have a non-zero overlap. Note that setting
ee21ebee 2434 this option can reduce both performance and the :option:`iodepth` achieved.
3d6a6f04
VF
2435
2436 This option only applies to I/Os issued for a single job except when it is
a02ec45a 2437 enabled along with :option:`io_submit_mode`\=offload. In offload mode, fio
3d6a6f04 2438 will check for overlap among all I/Os submitted by offload jobs with :option:`serialize_overlap`
307f2246 2439 enabled.
3d6a6f04
VF
2440
2441 Default: false.
997b5680 2442
f80dba8d
MT
2443.. option:: io_submit_mode=str
2444
2445 This option controls how fio submits the I/O to the I/O engine. The default
2446 is `inline`, which means that the fio job threads submit and reap I/O
2447 directly. If set to `offload`, the job threads will offload I/O submission
2448 to a dedicated pool of I/O threads. This requires some coordination and thus
2449 has a bit of extra overhead, especially for lower queue depth I/O where it
2450 can increase latencies. The benefit is that fio can manage submission rates
2451 independently of the device completion rates. This avoids skewed latency
730bd7d9 2452 reporting if I/O gets backed up on the device side (the coordinated omission
f80dba8d
MT
2453 problem).
2454
2455
2456I/O rate
2457~~~~~~~~
2458
a881438b 2459.. option:: thinktime=time
f80dba8d 2460
f75ede1d
SW
2461 Stall the job for the specified period of time after an I/O has completed before issuing the
2462 next. May be used to simulate processing being done by an application.
947e0fe0 2463 When the unit is omitted, the value is interpreted in microseconds. See
f80dba8d
MT
2464 :option:`thinktime_blocks` and :option:`thinktime_spin`.
2465
a881438b 2466.. option:: thinktime_spin=time
f80dba8d
MT
2467
2468 Only valid if :option:`thinktime` is set - pretend to spend CPU time doing
2469 something with the data received, before falling back to sleeping for the
f75ede1d 2470 rest of the period specified by :option:`thinktime`. When the unit is
947e0fe0 2471 omitted, the value is interpreted in microseconds.
f80dba8d
MT
2472
2473.. option:: thinktime_blocks=int
2474
2475 Only valid if :option:`thinktime` is set - control how many blocks to issue,
f50fbdda
TK
2476 before waiting :option:`thinktime` usecs. If not set, defaults to 1 which will make
2477 fio wait :option:`thinktime` usecs after every block. This effectively makes any
f80dba8d 2478 queue depth setting redundant, since no more than 1 I/O will be queued
f50fbdda 2479 before we have to complete it and do our :option:`thinktime`. In other words, this
f80dba8d 2480 setting effectively caps the queue depth if the latter is larger.
71bfa161 2481
f80dba8d 2482.. option:: rate=int[,int][,int]
71bfa161 2483
f80dba8d
MT
2484 Cap the bandwidth used by this job. The number is in bytes/sec, the normal
2485 suffix rules apply. Comma-separated values may be specified for reads,
2486 writes, and trims as described in :option:`blocksize`.
71bfa161 2487
b25b3464
SW
2488 For example, using `rate=1m,500k` would limit reads to 1MiB/sec and writes to
2489 500KiB/sec. Capping only reads or writes can be done with `rate=,500k` or
2490 `rate=500k,` where the former will only limit writes (to 500KiB/sec) and the
2491 latter will only limit reads.
2492
f80dba8d 2493.. option:: rate_min=int[,int][,int]
71bfa161 2494
f80dba8d
MT
2495 Tell fio to do whatever it can to maintain at least this bandwidth. Failing
2496 to meet this requirement will cause the job to exit. Comma-separated values
2497 may be specified for reads, writes, and trims as described in
2498 :option:`blocksize`.
71bfa161 2499
f80dba8d 2500.. option:: rate_iops=int[,int][,int]
71bfa161 2501
f80dba8d
MT
2502 Cap the bandwidth to this number of IOPS. Basically the same as
2503 :option:`rate`, just specified independently of bandwidth. If the job is
2504 given a block size range instead of a fixed value, the smallest block size
2505 is used as the metric. Comma-separated values may be specified for reads,
2506 writes, and trims as described in :option:`blocksize`.
71bfa161 2507
f80dba8d 2508.. option:: rate_iops_min=int[,int][,int]
71bfa161 2509
f80dba8d
MT
2510 If fio doesn't meet this rate of I/O, it will cause the job to exit.
2511 Comma-separated values may be specified for reads, writes, and trims as
2512 described in :option:`blocksize`.
71bfa161 2513
f80dba8d 2514.. option:: rate_process=str
66c098b8 2515
f80dba8d
MT
2516 This option controls how fio manages rated I/O submissions. The default is
2517 `linear`, which submits I/O in a linear fashion with fixed delays between
c60ebc45 2518 I/Os that gets adjusted based on I/O completion rates. If this is set to
f80dba8d
MT
2519 `poisson`, fio will submit I/O based on a more real world random request
2520 flow, known as the Poisson process
2521 (https://en.wikipedia.org/wiki/Poisson_point_process). The lambda will be
2522 10^6 / IOPS for the given workload.
71bfa161 2523
1a9bf814
JA
2524.. option:: rate_ignore_thinktime=bool
2525
2526 By default, fio will attempt to catch up to the specified rate setting,
2527 if any kind of thinktime setting was used. If this option is set, then
2528 fio will ignore the thinktime and continue doing IO at the specified
2529 rate, instead of entering a catch-up mode after thinktime is done.
2530
71bfa161 2531
f80dba8d
MT
2532I/O latency
2533~~~~~~~~~~~
71bfa161 2534
a881438b 2535.. option:: latency_target=time
71bfa161 2536
f80dba8d 2537 If set, fio will attempt to find the max performance point that the given
f75ede1d 2538 workload will run at while maintaining a latency below this target. When
947e0fe0 2539 the unit is omitted, the value is interpreted in microseconds. See
f75ede1d 2540 :option:`latency_window` and :option:`latency_percentile`.
71bfa161 2541
a881438b 2542.. option:: latency_window=time
71bfa161 2543
f80dba8d 2544 Used with :option:`latency_target` to specify the sample window that the job
f75ede1d 2545 is run at varying queue depths to test the performance. When the unit is
947e0fe0 2546 omitted, the value is interpreted in microseconds.
b4692828 2547
f80dba8d 2548.. option:: latency_percentile=float
71bfa161 2549
c60ebc45 2550 The percentage of I/Os that must fall within the criteria specified by
f80dba8d 2551 :option:`latency_target` and :option:`latency_window`. If not set, this
c60ebc45 2552 defaults to 100.0, meaning that all I/Os must be equal or below to the value
f80dba8d 2553 set by :option:`latency_target`.
71bfa161 2554
e1bcd541
SL
2555.. option:: latency_run=bool
2556
2557 Used with :option:`latency_target`. If false (default), fio will find
2558 the highest queue depth that meets :option:`latency_target` and exit. If
2559 true, fio will continue running and try to meet :option:`latency_target`
2560 by adjusting queue depth.
2561
a881438b 2562.. option:: max_latency=time
71bfa161 2563
f75ede1d 2564 If set, fio will exit the job with an ETIMEDOUT error if it exceeds this
947e0fe0 2565 maximum latency. When the unit is omitted, the value is interpreted in
f75ede1d 2566 microseconds.
71bfa161 2567
f80dba8d 2568.. option:: rate_cycle=int
71bfa161 2569
f80dba8d 2570 Average bandwidth for :option:`rate` and :option:`rate_min` over this number
a47b697c 2571 of milliseconds. Defaults to 1000.
71bfa161 2572
71bfa161 2573
f80dba8d
MT
2574I/O replay
2575~~~~~~~~~~
71bfa161 2576
f80dba8d 2577.. option:: write_iolog=str
c2b1e753 2578
f80dba8d
MT
2579 Write the issued I/O patterns to the specified file. See
2580 :option:`read_iolog`. Specify a separate file for each job, otherwise the
2581 iologs will be interspersed and the file may be corrupt.
c2b1e753 2582
f80dba8d 2583.. option:: read_iolog=str
71bfa161 2584
22413915 2585 Open an iolog with the specified filename and replay the I/O patterns it
f80dba8d
MT
2586 contains. This can be used to store a workload and replay it sometime
2587 later. The iolog given may also be a blktrace binary file, which allows fio
2588 to replay a workload captured by :command:`blktrace`. See
2589 :manpage:`blktrace(8)` for how to capture such logging data. For blktrace
2590 replay, the file needs to be turned into a blkparse binary data file first
2591 (``blkparse <device> -o /dev/null -d file_for_fio.bin``).
78439a18
JA
2592 You can specify a number of files by separating the names with a ':'
2593 character. See the :option:`filename` option for information on how to
3b803fe1 2594 escape ':' characters within the file names. These files will
78439a18 2595 be sequentially assigned to job clones created by :option:`numjobs`.
71bfa161 2596
77be374d
AK
2597.. option:: read_iolog_chunked=bool
2598
2599 Determines how iolog is read. If false(default) entire :option:`read_iolog`
2600 will be read at once. If selected true, input from iolog will be read
2601 gradually. Useful when iolog is very large, or it is generated.
2602
b9921d1a
DZ
2603.. option:: merge_blktrace_file=str
2604
2605 When specified, rather than replaying the logs passed to :option:`read_iolog`,
2606 the logs go through a merge phase which aggregates them into a single
2607 blktrace. The resulting file is then passed on as the :option:`read_iolog`
2608 parameter. The intention here is to make the order of events consistent.
2609 This limits the influence of the scheduler compared to replaying multiple
2610 blktraces via concurrent jobs.
2611
87a48ada
DZ
2612.. option:: merge_blktrace_scalars=float_list
2613
2614 This is a percentage based option that is index paired with the list of
2615 files passed to :option:`read_iolog`. When merging is performed, scale
2616 the time of each event by the corresponding amount. For example,
2617 ``--merge_blktrace_scalars="50:100"`` runs the first trace in halftime
2618 and the second trace in realtime. This knob is separately tunable from
2619 :option:`replay_time_scale` which scales the trace during runtime and
2620 does not change the output of the merge unlike this option.
2621
55bfd8c8
DZ
2622.. option:: merge_blktrace_iters=float_list
2623
2624 This is a whole number option that is index paired with the list of files
2625 passed to :option:`read_iolog`. When merging is performed, run each trace
2626 for the specified number of iterations. For example,
2627 ``--merge_blktrace_iters="2:1"`` runs the first trace for two iterations
2628 and the second trace for one iteration.
2629
589e88b7 2630.. option:: replay_no_stall=bool
71bfa161 2631
f80dba8d 2632 When replaying I/O with :option:`read_iolog` the default behavior is to
22413915 2633 attempt to respect the timestamps within the log and replay them with the
f80dba8d
MT
2634 appropriate delay between IOPS. By setting this variable fio will not
2635 respect the timestamps and attempt to replay them as fast as possible while
2636 still respecting ordering. The result is the same I/O pattern to a given
2637 device, but different timings.
71bfa161 2638
6dd7fa77
JA
2639.. option:: replay_time_scale=int
2640
2641 When replaying I/O with :option:`read_iolog`, fio will honor the
2642 original timing in the trace. With this option, it's possible to scale
2643 the time. It's a percentage option, if set to 50 it means run at 50%
2644 the original IO rate in the trace. If set to 200, run at twice the
2645 original IO rate. Defaults to 100.
2646
f80dba8d 2647.. option:: replay_redirect=str
b4692828 2648
f80dba8d
MT
2649 While replaying I/O patterns using :option:`read_iolog` the default behavior
2650 is to replay the IOPS onto the major/minor device that each IOP was recorded
2651 from. This is sometimes undesirable because on a different machine those
2652 major/minor numbers can map to a different device. Changing hardware on the
2653 same system can also result in a different major/minor mapping.
730bd7d9 2654 ``replay_redirect`` causes all I/Os to be replayed onto the single specified
f80dba8d 2655 device regardless of the device it was recorded
9207a0cb 2656 from. i.e. :option:`replay_redirect`\= :file:`/dev/sdc` would cause all I/O
f80dba8d
MT
2657 in the blktrace or iolog to be replayed onto :file:`/dev/sdc`. This means
2658 multiple devices will be replayed onto a single device, if the trace
2659 contains multiple devices. If you want multiple devices to be replayed
2660 concurrently to multiple redirected devices you must blkparse your trace
2661 into separate traces and replay them with independent fio invocations.
2662 Unfortunately this also breaks the strict time ordering between multiple
2663 device accesses.
71bfa161 2664
f80dba8d 2665.. option:: replay_align=int
74929ac2 2666
350a535d
DZ
2667 Force alignment of the byte offsets in a trace to this value. The value
2668 must be a power of 2.
3c54bc46 2669
f80dba8d 2670.. option:: replay_scale=int
3c54bc46 2671
350a535d
DZ
2672 Scale byte offsets down by this factor when replaying traces. Should most
2673 likely use :option:`replay_align` as well.
3c54bc46 2674
38f68906
JA
2675.. option:: replay_skip=str
2676
2677 Sometimes it's useful to skip certain IO types in a replay trace.
2678 This could be, for instance, eliminating the writes in the trace.
2679 Or not replaying the trims/discards, if you are redirecting to
2680 a device that doesn't support them. This option takes a comma
2681 separated list of read, write, trim, sync.
2682
3c54bc46 2683
f80dba8d
MT
2684Threads, processes and job synchronization
2685~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3c54bc46 2686
f80dba8d 2687.. option:: thread
3c54bc46 2688
730bd7d9
SW
2689 Fio defaults to creating jobs by using fork, however if this option is
2690 given, fio will create jobs by using POSIX Threads' function
2691 :manpage:`pthread_create(3)` to create threads instead.
71bfa161 2692
f80dba8d 2693.. option:: wait_for=str
74929ac2 2694
730bd7d9
SW
2695 If set, the current job won't be started until all workers of the specified
2696 waitee job are done.
74929ac2 2697
f80dba8d
MT
2698 ``wait_for`` operates on the job name basis, so there are a few
2699 limitations. First, the waitee must be defined prior to the waiter job
2700 (meaning no forward references). Second, if a job is being referenced as a
2701 waitee, it must have a unique name (no duplicate waitees).
74929ac2 2702
f80dba8d 2703.. option:: nice=int
892a6ffc 2704
f80dba8d 2705 Run the job with the given nice value. See man :manpage:`nice(2)`.
892a6ffc 2706
f80dba8d
MT
2707 On Windows, values less than -15 set the process class to "High"; -1 through
2708 -15 set "Above Normal"; 1 through 15 "Below Normal"; and above 15 "Idle"
2709 priority class.
74929ac2 2710
f80dba8d 2711.. option:: prio=int
71bfa161 2712
f80dba8d
MT
2713 Set the I/O priority value of this job. Linux limits us to a positive value
2714 between 0 and 7, with 0 being the highest. See man
2715 :manpage:`ionice(1)`. Refer to an appropriate manpage for other operating
b2a432bf
PC
2716 systems since meaning of priority may differ. For per-command priority
2717 setting, see I/O engine specific `cmdprio_percentage` and `hipri_percentage`
2718 options.
71bfa161 2719
f80dba8d 2720.. option:: prioclass=int
d59aa780 2721
b2a432bf
PC
2722 Set the I/O priority class. See man :manpage:`ionice(1)`. For per-command
2723 priority setting, see I/O engine specific `cmdprio_percentage` and
2724 `hipri_percentage` options.
d59aa780 2725
f80dba8d 2726.. option:: cpus_allowed=str
6d500c2e 2727
730bd7d9 2728 Controls the same options as :option:`cpumask`, but accepts a textual
b570e037
SW
2729 specification of the permitted CPUs instead and CPUs are indexed from 0. So
2730 to use CPUs 0 and 5 you would specify ``cpus_allowed=0,5``. This option also
2731 allows a range of CPUs to be specified -- say you wanted a binding to CPUs
2732 0, 5, and 8 to 15, you would set ``cpus_allowed=0,5,8-15``.
2733
2734 On Windows, when ``cpus_allowed`` is unset only CPUs from fio's current
2735 processor group will be used and affinity settings are inherited from the
2736 system. An fio build configured to target Windows 7 makes options that set
2737 CPUs processor group aware and values will set both the processor group
2738 and a CPU from within that group. For example, on a system where processor
2739 group 0 has 40 CPUs and processor group 1 has 32 CPUs, ``cpus_allowed``
2740 values between 0 and 39 will bind CPUs from processor group 0 and
2741 ``cpus_allowed`` values between 40 and 71 will bind CPUs from processor
2742 group 1. When using ``cpus_allowed_policy=shared`` all CPUs specified by a
2743 single ``cpus_allowed`` option must be from the same processor group. For
2744 Windows fio builds not built for Windows 7, CPUs will only be selected from
2745 (and be relative to) whatever processor group fio happens to be running in
2746 and CPUs from other processor groups cannot be used.
6d500c2e 2747
f80dba8d 2748.. option:: cpus_allowed_policy=str
6d500c2e 2749
f80dba8d 2750 Set the policy of how fio distributes the CPUs specified by
730bd7d9 2751 :option:`cpus_allowed` or :option:`cpumask`. Two policies are supported:
6d500c2e 2752
f80dba8d
MT
2753 **shared**
2754 All jobs will share the CPU set specified.
2755 **split**
2756 Each job will get a unique CPU from the CPU set.
6d500c2e 2757
22413915 2758 **shared** is the default behavior, if the option isn't specified. If
b21fc93f 2759 **split** is specified, then fio will assign one cpu per job. If not
f80dba8d
MT
2760 enough CPUs are given for the jobs listed, then fio will roundrobin the CPUs
2761 in the set.
6d500c2e 2762
b570e037
SW
2763.. option:: cpumask=int
2764
2765 Set the CPU affinity of this job. The parameter given is a bit mask of
2766 allowed CPUs the job may run on. So if you want the allowed CPUs to be 1
2767 and 5, you would pass the decimal value of (1 << 1 | 1 << 5), or 34. See man
2768 :manpage:`sched_setaffinity(2)`. This may not work on all supported
2769 operating systems or kernel versions. This option doesn't work well for a
2770 higher CPU count than what you can store in an integer mask, so it can only
2771 control cpus 1-32. For boxes with larger CPU counts, use
2772 :option:`cpus_allowed`.
2773
f80dba8d 2774.. option:: numa_cpu_nodes=str
6d500c2e 2775
f80dba8d
MT
2776 Set this job running on specified NUMA nodes' CPUs. The arguments allow
2777 comma delimited list of cpu numbers, A-B ranges, or `all`. Note, to enable
ac8ca2af 2778 NUMA options support, fio must be built on a system with libnuma-dev(el)
f80dba8d 2779 installed.
61b9861d 2780
f80dba8d 2781.. option:: numa_mem_policy=str
61b9861d 2782
f80dba8d
MT
2783 Set this job's memory policy and corresponding NUMA nodes. Format of the
2784 arguments::
5c94b008 2785
f80dba8d 2786 <mode>[:<nodelist>]
ce35b1ec 2787
804c0839 2788 ``mode`` is one of the following memory policies: ``default``, ``prefer``,
730bd7d9
SW
2789 ``bind``, ``interleave`` or ``local``. For ``default`` and ``local`` memory
2790 policies, no node needs to be specified. For ``prefer``, only one node is
2791 allowed. For ``bind`` and ``interleave`` the ``nodelist`` may be as
2792 follows: a comma delimited list of numbers, A-B ranges, or `all`.
71bfa161 2793
f80dba8d 2794.. option:: cgroup=str
390b1537 2795
f80dba8d
MT
2796 Add job to this control group. If it doesn't exist, it will be created. The
2797 system must have a mounted cgroup blkio mount point for this to work. If
2798 your system doesn't have it mounted, you can do so with::
5af1c6f3 2799
f80dba8d 2800 # mount -t cgroup -o blkio none /cgroup
5af1c6f3 2801
f80dba8d 2802.. option:: cgroup_weight=int
5af1c6f3 2803
f80dba8d
MT
2804 Set the weight of the cgroup to this value. See the documentation that comes
2805 with the kernel, allowed values are in the range of 100..1000.
a086c257 2806
f80dba8d 2807.. option:: cgroup_nodelete=bool
8c07860d 2808
f80dba8d
MT
2809 Normally fio will delete the cgroups it has created after the job
2810 completion. To override this behavior and to leave cgroups around after the
2811 job completion, set ``cgroup_nodelete=1``. This can be useful if one wants
2812 to inspect various cgroup files after job completion. Default: false.
8c07860d 2813
f80dba8d 2814.. option:: flow_id=int
8c07860d 2815
f80dba8d
MT
2816 The ID of the flow. If not specified, it defaults to being a global
2817 flow. See :option:`flow`.
1907dbc6 2818
f80dba8d 2819.. option:: flow=int
71bfa161 2820
f80dba8d
MT
2821 Weight in token-based flow control. If this value is used, then there is a
2822 'flow counter' which is used to regulate the proportion of activity between
2823 two or more jobs. Fio attempts to keep this flow counter near zero. The
2824 ``flow`` parameter stands for how much should be added or subtracted to the
2825 flow counter on each iteration of the main I/O loop. That is, if one job has
2826 ``flow=8`` and another job has ``flow=-1``, then there will be a roughly 1:8
2827 ratio in how much one runs vs the other.
71bfa161 2828
f80dba8d 2829.. option:: flow_watermark=int
a31041ea 2830
f80dba8d
MT
2831 The maximum value that the absolute value of the flow counter is allowed to
2832 reach before the job must wait for a lower value of the counter.
82407585 2833
f80dba8d 2834.. option:: flow_sleep=int
82407585 2835
f80dba8d
MT
2836 The period of time, in microseconds, to wait after the flow watermark has
2837 been exceeded before retrying operations.
82407585 2838
f80dba8d 2839.. option:: stonewall, wait_for_previous
82407585 2840
f80dba8d
MT
2841 Wait for preceding jobs in the job file to exit, before starting this
2842 one. Can be used to insert serialization points in the job file. A stone
2843 wall also implies starting a new reporting group, see
2844 :option:`group_reporting`.
2845
2846.. option:: exitall
2847
64402a8a
HW
2848 By default, fio will continue running all other jobs when one job finishes.
2849 Sometimes this is not the desired action. Setting ``exitall`` will instead
2850 make fio terminate all jobs in the same group, as soon as one job of that
2851 group finishes.
2852
2853.. option:: exit_what
2854
2855 By default, fio will continue running all other jobs when one job finishes.
2856 Sometimes this is not the desired action. Setting ``exit_all`` will
2857 instead make fio terminate all jobs in the same group. The option
2858 ``exit_what`` allows to control which jobs get terminated when ``exitall`` is
2859 enabled. The default is ``group`` and does not change the behaviour of
2860 ``exitall``. The setting ``all`` terminates all jobs. The setting ``stonewall``
2861 terminates all currently running jobs across all groups and continues execution
2862 with the next stonewalled group.
f80dba8d
MT
2863
2864.. option:: exec_prerun=str
2865
2866 Before running this job, issue the command specified through
2867 :manpage:`system(3)`. Output is redirected in a file called
2868 :file:`jobname.prerun.txt`.
2869
2870.. option:: exec_postrun=str
2871
2872 After the job completes, issue the command specified though
2873 :manpage:`system(3)`. Output is redirected in a file called
2874 :file:`jobname.postrun.txt`.
2875
2876.. option:: uid=int
2877
2878 Instead of running as the invoking user, set the user ID to this value
2879 before the thread/process does any work.
2880
2881.. option:: gid=int
2882
2883 Set group ID, see :option:`uid`.
2884
2885
2886Verification
2887~~~~~~~~~~~~
2888
2889.. option:: verify_only
2890
2891 Do not perform specified workload, only verify data still matches previous
2892 invocation of this workload. This option allows one to check data multiple
2893 times at a later date without overwriting it. This option makes sense only
2894 for workloads that write data, and does not support workloads with the
2895 :option:`time_based` option set.
2896
2897.. option:: do_verify=bool
2898
2899 Run the verify phase after a write phase. Only valid if :option:`verify` is
2900 set. Default: true.
2901
2902.. option:: verify=str
2903
2904 If writing to a file, fio can verify the file contents after each iteration
2905 of the job. Each verification method also implies verification of special
2906 header, which is written to the beginning of each block. This header also
2907 includes meta information, like offset of the block, block number, timestamp
2908 when block was written, etc. :option:`verify` can be combined with
2909 :option:`verify_pattern` option. The allowed values are:
2910
2911 **md5**
2912 Use an md5 sum of the data area and store it in the header of
2913 each block.
2914
2915 **crc64**
2916 Use an experimental crc64 sum of the data area and store it in the
2917 header of each block.
2918
2919 **crc32c**
a5896300
SW
2920 Use a crc32c sum of the data area and store it in the header of
2921 each block. This will automatically use hardware acceleration
2922 (e.g. SSE4.2 on an x86 or CRC crypto extensions on ARM64) but will
2923 fall back to software crc32c if none is found. Generally the
804c0839 2924 fastest checksum fio supports when hardware accelerated.
f80dba8d
MT
2925
2926 **crc32c-intel**
a5896300 2927 Synonym for crc32c.
f80dba8d
MT
2928
2929 **crc32**
2930 Use a crc32 sum of the data area and store it in the header of each
2931 block.
2932
2933 **crc16**
2934 Use a crc16 sum of the data area and store it in the header of each
2935 block.
2936
2937 **crc7**
2938 Use a crc7 sum of the data area and store it in the header of each
2939 block.
2940
2941 **xxhash**
2942 Use xxhash as the checksum function. Generally the fastest software
2943 checksum that fio supports.
2944
2945 **sha512**
2946 Use sha512 as the checksum function.
2947
2948 **sha256**
2949 Use sha256 as the checksum function.
2950
2951 **sha1**
2952 Use optimized sha1 as the checksum function.
82407585 2953
ae3a5acc
JA
2954 **sha3-224**
2955 Use optimized sha3-224 as the checksum function.
2956
2957 **sha3-256**
2958 Use optimized sha3-256 as the checksum function.
2959
2960 **sha3-384**
2961 Use optimized sha3-384 as the checksum function.
2962
2963 **sha3-512**
2964 Use optimized sha3-512 as the checksum function.
2965
f80dba8d
MT
2966 **meta**
2967 This option is deprecated, since now meta information is included in
2968 generic verification header and meta verification happens by
2969 default. For detailed information see the description of the
2970 :option:`verify` setting. This option is kept because of
2971 compatibility's sake with old configurations. Do not use it.
2972
2973 **pattern**
2974 Verify a strict pattern. Normally fio includes a header with some
2975 basic information and checksumming, but if this option is set, only
2976 the specific pattern set with :option:`verify_pattern` is verified.
2977
2978 **null**
2979 Only pretend to verify. Useful for testing internals with
9207a0cb 2980 :option:`ioengine`\=null, not for much else.
f80dba8d
MT
2981
2982 This option can be used for repeated burn-in tests of a system to make sure
2983 that the written data is also correctly read back. If the data direction
2984 given is a read or random read, fio will assume that it should verify a
2985 previously written file. If the data direction includes any form of write,
2986 the verify will be of the newly written data.
2987
47e6a6e5
SW
2988 To avoid false verification errors, do not use the norandommap option when
2989 verifying data with async I/O engines and I/O depths > 1. Or use the
2990 norandommap and the lfsr random generator together to avoid writing to the
2991 same offset with muliple outstanding I/Os.
2992
f80dba8d
MT
2993.. option:: verify_offset=int
2994
2995 Swap the verification header with data somewhere else in the block before
2996 writing. It is swapped back before verifying.
2997
2998.. option:: verify_interval=int
2999
3000 Write the verification header at a finer granularity than the
3001 :option:`blocksize`. It will be written for chunks the size of
3002 ``verify_interval``. :option:`blocksize` should divide this evenly.
3003
3004.. option:: verify_pattern=str
3005
3006 If set, fio will fill the I/O buffers with this pattern. Fio defaults to
3007 filling with totally random bytes, but sometimes it's interesting to fill
3008 with a known pattern for I/O verification purposes. Depending on the width
730bd7d9 3009 of the pattern, fio will fill 1/2/3/4 bytes of the buffer at the time (it can
f80dba8d
MT
3010 be either a decimal or a hex number). The ``verify_pattern`` if larger than
3011 a 32-bit quantity has to be a hex number that starts with either "0x" or
3012 "0X". Use with :option:`verify`. Also, ``verify_pattern`` supports %o
3013 format, which means that for each block offset will be written and then
3014 verified back, e.g.::
61b9861d
RP
3015
3016 verify_pattern=%o
3017
f80dba8d
MT
3018 Or use combination of everything::
3019
61b9861d 3020 verify_pattern=0xff%o"abcd"-12
e28218f3 3021
f80dba8d
MT
3022.. option:: verify_fatal=bool
3023
3024 Normally fio will keep checking the entire contents before quitting on a
3025 block verification failure. If this option is set, fio will exit the job on
3026 the first observed failure. Default: false.
3027
3028.. option:: verify_dump=bool
3029
3030 If set, dump the contents of both the original data block and the data block
3031 we read off disk to files. This allows later analysis to inspect just what
3032 kind of data corruption occurred. Off by default.
3033
3034.. option:: verify_async=int
3035
3036 Fio will normally verify I/O inline from the submitting thread. This option
3037 takes an integer describing how many async offload threads to create for I/O
3038 verification instead, causing fio to offload the duty of verifying I/O
3039 contents to one or more separate threads. If using this offload option, even
3040 sync I/O engines can benefit from using an :option:`iodepth` setting higher
3041 than 1, as it allows them to have I/O in flight while verifies are running.
d7e6ea1c 3042 Defaults to 0 async threads, i.e. verification is not asynchronous.
f80dba8d
MT
3043
3044.. option:: verify_async_cpus=str
3045
3046 Tell fio to set the given CPU affinity on the async I/O verification
3047 threads. See :option:`cpus_allowed` for the format used.
3048
3049.. option:: verify_backlog=int
3050
3051 Fio will normally verify the written contents of a job that utilizes verify
3052 once that job has completed. In other words, everything is written then
3053 everything is read back and verified. You may want to verify continually
3054 instead for a variety of reasons. Fio stores the meta data associated with
3055 an I/O block in memory, so for large verify workloads, quite a bit of memory
3056 would be used up holding this meta data. If this option is enabled, fio will
3057 write only N blocks before verifying these blocks.
3058
3059.. option:: verify_backlog_batch=int
3060
3061 Control how many blocks fio will verify if :option:`verify_backlog` is
3062 set. If not set, will default to the value of :option:`verify_backlog`
3063 (meaning the entire queue is read back and verified). If
3064 ``verify_backlog_batch`` is less than :option:`verify_backlog` then not all
3065 blocks will be verified, if ``verify_backlog_batch`` is larger than
3066 :option:`verify_backlog`, some blocks will be verified more than once.
3067
3068.. option:: verify_state_save=bool
3069
3070 When a job exits during the write phase of a verify workload, save its
3071 current state. This allows fio to replay up until that point, if the verify
3072 state is loaded for the verify read phase. The format of the filename is,
3073 roughly::
3074
f50fbdda 3075 <type>-<jobname>-<jobindex>-verify.state.
f80dba8d
MT
3076
3077 <type> is "local" for a local run, "sock" for a client/server socket
3078 connection, and "ip" (192.168.0.1, for instance) for a networked
d7e6ea1c 3079 client/server connection. Defaults to true.
f80dba8d
MT
3080
3081.. option:: verify_state_load=bool
3082
3083 If a verify termination trigger was used, fio stores the current write state
3084 of each thread. This can be used at verification time so that fio knows how
3085 far it should verify. Without this information, fio will run a full
a47b697c
SW
3086 verification pass, according to the settings in the job file used. Default
3087 false.
f80dba8d
MT
3088
3089.. option:: trim_percentage=int
3090
3091 Number of verify blocks to discard/trim.
3092
3093.. option:: trim_verify_zero=bool
3094
22413915 3095 Verify that trim/discarded blocks are returned as zeros.
f80dba8d
MT
3096
3097.. option:: trim_backlog=int
3098
5cfd1e9a 3099 Trim after this number of blocks are written.
f80dba8d
MT
3100
3101.. option:: trim_backlog_batch=int
3102
3103 Trim this number of I/O blocks.
3104
3105.. option:: experimental_verify=bool
3106
3107 Enable experimental verification.
3108
f80dba8d
MT
3109Steady state
3110~~~~~~~~~~~~
3111
3112.. option:: steadystate=str:float, ss=str:float
3113
3114 Define the criterion and limit for assessing steady state performance. The
3115 first parameter designates the criterion whereas the second parameter sets
3116 the threshold. When the criterion falls below the threshold for the
3117 specified duration, the job will stop. For example, `iops_slope:0.1%` will
3118 direct fio to terminate the job when the least squares regression slope
3119 falls below 0.1% of the mean IOPS. If :option:`group_reporting` is enabled
3120 this will apply to all jobs in the group. Below is the list of available
3121 steady state assessment criteria. All assessments are carried out using only
3122 data from the rolling collection window. Threshold limits can be expressed
3123 as a fixed value or as a percentage of the mean in the collection window.
3124
1cb049d9
VF
3125 When using this feature, most jobs should include the :option:`time_based`
3126 and :option:`runtime` options or the :option:`loops` option so that fio does not
3127 stop running after it has covered the full size of the specified file(s) or device(s).
3128
f80dba8d
MT
3129 **iops**
3130 Collect IOPS data. Stop the job if all individual IOPS measurements
3131 are within the specified limit of the mean IOPS (e.g., ``iops:2``
3132 means that all individual IOPS values must be within 2 of the mean,
3133 whereas ``iops:0.2%`` means that all individual IOPS values must be
3134 within 0.2% of the mean IOPS to terminate the job).
3135
3136 **iops_slope**
3137 Collect IOPS data and calculate the least squares regression
3138 slope. Stop the job if the slope falls below the specified limit.
3139
3140 **bw**
3141 Collect bandwidth data. Stop the job if all individual bandwidth
3142 measurements are within the specified limit of the mean bandwidth.
3143
3144 **bw_slope**
3145 Collect bandwidth data and calculate the least squares regression
3146 slope. Stop the job if the slope falls below the specified limit.
3147
3148.. option:: steadystate_duration=time, ss_dur=time
3149
3150 A rolling window of this duration will be used to judge whether steady state
3151 has been reached. Data will be collected once per second. The default is 0
f75ede1d 3152 which disables steady state detection. When the unit is omitted, the
947e0fe0 3153 value is interpreted in seconds.
f80dba8d
MT
3154
3155.. option:: steadystate_ramp_time=time, ss_ramp=time
3156
3157 Allow the job to run for the specified duration before beginning data
3158 collection for checking the steady state job termination criterion. The
947e0fe0 3159 default is 0. When the unit is omitted, the value is interpreted in seconds.
f80dba8d
MT
3160
3161
3162Measurements and reporting
3163~~~~~~~~~~~~~~~~~~~~~~~~~~
3164
3165.. option:: per_job_logs=bool
3166
3167 If set, this generates bw/clat/iops log with per file private filenames. If
3168 not set, jobs with identical names will share the log filename. Default:
3169 true.
3170
3171.. option:: group_reporting
3172
3173 It may sometimes be interesting to display statistics for groups of jobs as
3174 a whole instead of for each individual job. This is especially true if
3175 :option:`numjobs` is used; looking at individual thread/process output
3176 quickly becomes unwieldy. To see the final report per-group instead of
3177 per-job, use :option:`group_reporting`. Jobs in a file will be part of the
3178 same reporting group, unless if separated by a :option:`stonewall`, or by
3179 using :option:`new_group`.
3180
3181.. option:: new_group
3182
3183 Start a new reporting group. See: :option:`group_reporting`. If not given,
3184 all jobs in a file will be part of the same reporting group, unless
3185 separated by a :option:`stonewall`.
3186
589e88b7 3187.. option:: stats=bool
8243be59
JA
3188
3189 By default, fio collects and shows final output results for all jobs
3190 that run. If this option is set to 0, then fio will ignore it in
3191 the final stat output.
3192
f80dba8d
MT
3193.. option:: write_bw_log=str
3194
3195 If given, write a bandwidth log for this job. Can be used to store data of
074f0817 3196 the bandwidth of the jobs in their lifetime.
f80dba8d 3197
074f0817
SW
3198 If no str argument is given, the default filename of
3199 :file:`jobname_type.x.log` is used. Even when the argument is given, fio
3200 will still append the type of log. So if one specifies::
3201
3202 write_bw_log=foo
f80dba8d 3203
074f0817
SW
3204 The actual log name will be :file:`foo_bw.x.log` where `x` is the index
3205 of the job (`1..N`, where `N` is the number of jobs). If
3206 :option:`per_job_logs` is false, then the filename will not include the
3207 `.x` job index.
e3cedca7 3208
074f0817
SW
3209 The included :command:`fio_generate_plots` script uses :command:`gnuplot` to turn these
3210 text files into nice graphs. See `Log File Formats`_ for how data is
3211 structured within the file.
3212
3213.. option:: write_lat_log=str
e3cedca7 3214
074f0817 3215 Same as :option:`write_bw_log`, except this option creates I/O
77b7e675
SW
3216 submission (e.g., :file:`name_slat.x.log`), completion (e.g.,
3217 :file:`name_clat.x.log`), and total (e.g., :file:`name_lat.x.log`)
074f0817
SW
3218 latency files instead. See :option:`write_bw_log` for details about
3219 the filename format and `Log File Formats`_ for how data is structured
3220 within the files.
be4ecfdf 3221
f80dba8d 3222.. option:: write_hist_log=str
06842027 3223
074f0817 3224 Same as :option:`write_bw_log` but writes an I/O completion latency
77b7e675 3225 histogram file (e.g., :file:`name_hist.x.log`) instead. Note that this
074f0817
SW
3226 file will be empty unless :option:`log_hist_msec` has also been set.
3227 See :option:`write_bw_log` for details about the filename format and
3228 `Log File Formats`_ for how data is structured within the file.
06842027 3229
f80dba8d 3230.. option:: write_iops_log=str
06842027 3231
074f0817 3232 Same as :option:`write_bw_log`, but writes an IOPS file (e.g.
15417073
SW
3233 :file:`name_iops.x.log`) instead. Because fio defaults to individual
3234 I/O logging, the value entry in the IOPS log will be 1 unless windowed
3235 logging (see :option:`log_avg_msec`) has been enabled. See
3236 :option:`write_bw_log` for details about the filename format and `Log
3237 File Formats`_ for how data is structured within the file.
06842027 3238
f80dba8d 3239.. option:: log_avg_msec=int
06842027 3240
f80dba8d
MT
3241 By default, fio will log an entry in the iops, latency, or bw log for every
3242 I/O that completes. When writing to the disk log, that can quickly grow to a
3243 very large size. Setting this option makes fio average the each log entry
3244 over the specified period of time, reducing the resolution of the log. See
3245 :option:`log_max_value` as well. Defaults to 0, logging all entries.
6fc82095 3246 Also see `Log File Formats`_.
06842027 3247
f80dba8d 3248.. option:: log_hist_msec=int
06842027 3249
f80dba8d
MT
3250 Same as :option:`log_avg_msec`, but logs entries for completion latency
3251 histograms. Computing latency percentiles from averages of intervals using
c60ebc45 3252 :option:`log_avg_msec` is inaccurate. Setting this option makes fio log
f80dba8d
MT
3253 histogram entries over the specified period of time, reducing log sizes for
3254 high IOPS devices while retaining percentile accuracy. See
074f0817
SW
3255 :option:`log_hist_coarseness` and :option:`write_hist_log` as well.
3256 Defaults to 0, meaning histogram logging is disabled.
06842027 3257
f80dba8d 3258.. option:: log_hist_coarseness=int
06842027 3259
f80dba8d
MT
3260 Integer ranging from 0 to 6, defining the coarseness of the resolution of
3261 the histogram logs enabled with :option:`log_hist_msec`. For each increment
3262 in coarseness, fio outputs half as many bins. Defaults to 0, for which
074f0817
SW
3263 histogram logs contain 1216 latency bins. See :option:`write_hist_log`
3264 and `Log File Formats`_.
8b28bd41 3265
f80dba8d 3266.. option:: log_max_value=bool
66c098b8 3267
f80dba8d
MT
3268 If :option:`log_avg_msec` is set, fio logs the average over that window. If
3269 you instead want to log the maximum value, set this option to 1. Defaults to
3270 0, meaning that averaged values are logged.
a696fa2a 3271
589e88b7 3272.. option:: log_offset=bool
a696fa2a 3273
f80dba8d 3274 If this is set, the iolog options will include the byte offset for the I/O
5a83478f
SW
3275 entry as well as the other data values. Defaults to 0 meaning that
3276 offsets are not present in logs. Also see `Log File Formats`_.
71bfa161 3277
f80dba8d 3278.. option:: log_compression=int
7de87099 3279
f80dba8d
MT
3280 If this is set, fio will compress the I/O logs as it goes, to keep the
3281 memory footprint lower. When a log reaches the specified size, that chunk is
3282 removed and compressed in the background. Given that I/O logs are fairly
3283 highly compressible, this yields a nice memory savings for longer runs. The
3284 downside is that the compression will consume some background CPU cycles, so
3285 it may impact the run. This, however, is also true if the logging ends up
3286 consuming most of the system memory. So pick your poison. The I/O logs are
3287 saved normally at the end of a run, by decompressing the chunks and storing
3288 them in the specified log file. This feature depends on the availability of
3289 zlib.
e0b0d892 3290
f80dba8d 3291.. option:: log_compression_cpus=str
e0b0d892 3292
f80dba8d
MT
3293 Define the set of CPUs that are allowed to handle online log compression for
3294 the I/O jobs. This can provide better isolation between performance
0cf90a62
SW
3295 sensitive jobs, and background compression work. See
3296 :option:`cpus_allowed` for the format used.
9e684a49 3297
f80dba8d 3298.. option:: log_store_compressed=bool
9e684a49 3299
f80dba8d
MT
3300 If set, fio will store the log files in a compressed format. They can be
3301 decompressed with fio, using the :option:`--inflate-log` command line
3302 parameter. The files will be stored with a :file:`.fz` suffix.
9e684a49 3303
f80dba8d 3304.. option:: log_unix_epoch=bool
9e684a49 3305
f80dba8d
MT
3306 If set, fio will log Unix timestamps to the log files produced by enabling
3307 write_type_log for each log type, instead of the default zero-based
3308 timestamps.
3309
3310.. option:: block_error_percentiles=bool
3311
3312 If set, record errors in trim block-sized units from writes and trims and
3313 output a histogram of how many trims it took to get to errors, and what kind
3314 of error was encountered.
3315
3316.. option:: bwavgtime=int
3317
3318 Average the calculated bandwidth over the given time. Value is specified in
3319 milliseconds. If the job also does bandwidth logging through
3320 :option:`write_bw_log`, then the minimum of this option and
3321 :option:`log_avg_msec` will be used. Default: 500ms.
3322
3323.. option:: iopsavgtime=int
3324
3325 Average the calculated IOPS over the given time. Value is specified in
3326 milliseconds. If the job also does IOPS logging through
3327 :option:`write_iops_log`, then the minimum of this option and
3328 :option:`log_avg_msec` will be used. Default: 500ms.
3329
3330.. option:: disk_util=bool
3331
3332 Generate disk utilization statistics, if the platform supports it.
3333 Default: true.
3334
3335.. option:: disable_lat=bool
3336
3337 Disable measurements of total latency numbers. Useful only for cutting back
3338 the number of calls to :manpage:`gettimeofday(2)`, as that does impact
3339 performance at really high IOPS rates. Note that to really get rid of a
3340 large amount of these calls, this option must be used with
f75ede1d 3341 :option:`disable_slat` and :option:`disable_bw_measurement` as well.
f80dba8d
MT
3342
3343.. option:: disable_clat=bool
3344
3345 Disable measurements of completion latency numbers. See
3346 :option:`disable_lat`.
3347
3348.. option:: disable_slat=bool
3349
3350 Disable measurements of submission latency numbers. See
f50fbdda 3351 :option:`disable_lat`.
f80dba8d 3352
f75ede1d 3353.. option:: disable_bw_measurement=bool, disable_bw=bool
f80dba8d
MT
3354
3355 Disable measurements of throughput/bandwidth numbers. See
3356 :option:`disable_lat`.
3357
dd39b9ce
VF
3358.. option:: slat_percentiles=bool
3359
3360 Report submission latency percentiles. Submission latency is not recorded
3361 for synchronous ioengines.
3362
f80dba8d
MT
3363.. option:: clat_percentiles=bool
3364
dd39b9ce 3365 Report completion latency percentiles.
b599759b
JA
3366
3367.. option:: lat_percentiles=bool
3368
dd39b9ce
VF
3369 Report total latency percentiles. Total latency is the sum of submission
3370 latency and completion latency.
f80dba8d
MT
3371
3372.. option:: percentile_list=float_list
3373
dd39b9ce
VF
3374 Overwrite the default list of percentiles for latencies and the block error
3375 histogram. Each number is a floating point number in the range (0,100], and
3376 the maximum length of the list is 20. Use ``:`` to separate the numbers. For
c32ba107 3377 example, ``--percentile_list=99.5:99.9`` will cause fio to report the
dd39b9ce
VF
3378 latency durations below which 99.5% and 99.9% of the observed latencies fell,
3379 respectively.
f80dba8d 3380
e883cb35
JF
3381.. option:: significant_figures=int
3382
c32ba107
JA
3383 If using :option:`--output-format` of `normal`, set the significant
3384 figures to this value. Higher values will yield more precise IOPS and
3385 throughput units, while lower values will round. Requires a minimum
3386 value of 1 and a maximum value of 10. Defaults to 4.
e883cb35 3387
f80dba8d
MT
3388
3389Error handling
3390~~~~~~~~~~~~~~
3391
3392.. option:: exitall_on_error
3393
3394 When one job finishes in error, terminate the rest. The default is to wait
3395 for each job to finish.
3396
3397.. option:: continue_on_error=str
3398
3399 Normally fio will exit the job on the first observed failure. If this option
3400 is set, fio will continue the job when there is a 'non-fatal error' (EIO or
3401 EILSEQ) until the runtime is exceeded or the I/O size specified is
3402 completed. If this option is used, there are two more stats that are
3403 appended, the total error count and the first error. The error field given
3404 in the stats is the first error that was hit during the run.
3405
3406 The allowed values are:
3407
3408 **none**
3409 Exit on any I/O or verify errors.
3410
3411 **read**
3412 Continue on read errors, exit on all others.
3413
3414 **write**
3415 Continue on write errors, exit on all others.
3416
3417 **io**
3418 Continue on any I/O error, exit on all others.
3419
3420 **verify**
3421 Continue on verify errors, exit on all others.
3422
3423 **all**
3424 Continue on all errors.
3425
3426 **0**
3427 Backward-compatible alias for 'none'.
3428
3429 **1**
3430 Backward-compatible alias for 'all'.
3431
3432.. option:: ignore_error=str
3433
3434 Sometimes you want to ignore some errors during test in that case you can
a35ef7cb
TK
3435 specify error list for each error type, instead of only being able to
3436 ignore the default 'non-fatal error' using :option:`continue_on_error`.
f80dba8d
MT
3437 ``ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST`` errors for
3438 given error type is separated with ':'. Error may be symbol ('ENOSPC',
3439 'ENOMEM') or integer. Example::
3440
3441 ignore_error=EAGAIN,ENOSPC:122
3442
3443 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from
a35ef7cb
TK
3444 WRITE. This option works by overriding :option:`continue_on_error` with
3445 the list of errors for each error type if any.
f80dba8d
MT
3446
3447.. option:: error_dump=bool
3448
3449 If set dump every error even if it is non fatal, true by default. If
3450 disabled only fatal error will be dumped.
3451
f75ede1d
SW
3452Running predefined workloads
3453----------------------------
3454
3455Fio includes predefined profiles that mimic the I/O workloads generated by
3456other tools.
3457
3458.. option:: profile=str
3459
3460 The predefined workload to run. Current profiles are:
3461
3462 **tiobench**
3463 Threaded I/O bench (tiotest/tiobench) like workload.
3464
3465 **act**
3466 Aerospike Certification Tool (ACT) like workload.
3467
3468To view a profile's additional options use :option:`--cmdhelp` after specifying
3469the profile. For example::
3470
f50fbdda 3471 $ fio --profile=act --cmdhelp
f75ede1d
SW
3472
3473Act profile options
3474~~~~~~~~~~~~~~~~~~~
3475
3476.. option:: device-names=str
3477 :noindex:
3478
3479 Devices to use.
3480
3481.. option:: load=int
3482 :noindex:
3483
3484 ACT load multiplier. Default: 1.
3485
3486.. option:: test-duration=time
3487 :noindex:
3488
947e0fe0
SW
3489 How long the entire test takes to run. When the unit is omitted, the value
3490 is given in seconds. Default: 24h.
f75ede1d
SW
3491
3492.. option:: threads-per-queue=int
3493 :noindex:
3494
f50fbdda 3495 Number of read I/O threads per device. Default: 8.
f75ede1d
SW
3496
3497.. option:: read-req-num-512-blocks=int
3498 :noindex:
3499
3500 Number of 512B blocks to read at the time. Default: 3.
3501
3502.. option:: large-block-op-kbytes=int
3503 :noindex:
3504
3505 Size of large block ops in KiB (writes). Default: 131072.
3506
3507.. option:: prep
3508 :noindex:
3509
3510 Set to run ACT prep phase.
3511
3512Tiobench profile options
3513~~~~~~~~~~~~~~~~~~~~~~~~
3514
3515.. option:: size=str
3516 :noindex:
3517
f50fbdda 3518 Size in MiB.
f75ede1d
SW
3519
3520.. option:: block=int
3521 :noindex:
3522
3523 Block size in bytes. Default: 4096.
3524
3525.. option:: numruns=int
3526 :noindex:
3527
3528 Number of runs.
3529
3530.. option:: dir=str
3531 :noindex:
3532
3533 Test directory.
3534
3535.. option:: threads=int
3536 :noindex:
3537
3538 Number of threads.
f80dba8d
MT
3539
3540Interpreting the output
3541-----------------------
3542
36214730
SW
3543..
3544 Example output was based on the following:
3545 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --time_based \
3546 --rate=1256k --bs=14K --name=quick --runtime=1s --name=mixed \
3547 --runtime=2m --rw=rw
3548
f80dba8d
MT
3549Fio spits out a lot of output. While running, fio will display the status of the
3550jobs created. An example of that would be::
3551
9d25d068 3552 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 3553
36214730
SW
3554The characters inside the first set of square brackets denote the current status of
3555each thread. The first character is the first job defined in the job file, and so
3556forth. The possible values (in typical life cycle order) are:
f80dba8d
MT
3557
3558+------+-----+-----------------------------------------------------------+
3559| Idle | Run | |
3560+======+=====+===========================================================+
3561| P | | Thread setup, but not started. |
3562+------+-----+-----------------------------------------------------------+
3563| C | | Thread created. |
3564+------+-----+-----------------------------------------------------------+
3565| I | | Thread initialized, waiting or generating necessary data. |
3566+------+-----+-----------------------------------------------------------+
3567| | p | Thread running pre-reading file(s). |
3568+------+-----+-----------------------------------------------------------+
36214730
SW
3569| | / | Thread is in ramp period. |
3570+------+-----+-----------------------------------------------------------+
f80dba8d
MT
3571| | R | Running, doing sequential reads. |
3572+------+-----+-----------------------------------------------------------+
3573| | r | Running, doing random reads. |
3574+------+-----+-----------------------------------------------------------+
3575| | W | Running, doing sequential writes. |
3576+------+-----+-----------------------------------------------------------+
3577| | w | Running, doing random writes. |
3578+------+-----+-----------------------------------------------------------+
3579| | M | Running, doing mixed sequential reads/writes. |
3580+------+-----+-----------------------------------------------------------+
3581| | m | Running, doing mixed random reads/writes. |
3582+------+-----+-----------------------------------------------------------+
36214730
SW
3583| | D | Running, doing sequential trims. |
3584+------+-----+-----------------------------------------------------------+
3585| | d | Running, doing random trims. |
3586+------+-----+-----------------------------------------------------------+
3587| | F | Running, currently waiting for :manpage:`fsync(2)`. |
f80dba8d
MT
3588+------+-----+-----------------------------------------------------------+
3589| | V | Running, doing verification of written data. |
3590+------+-----+-----------------------------------------------------------+
36214730
SW
3591| f | | Thread finishing. |
3592+------+-----+-----------------------------------------------------------+
f80dba8d
MT
3593| E | | Thread exited, not reaped by main thread yet. |
3594+------+-----+-----------------------------------------------------------+
36214730 3595| _ | | Thread reaped. |
f80dba8d
MT
3596+------+-----+-----------------------------------------------------------+
3597| X | | Thread reaped, exited with an error. |
3598+------+-----+-----------------------------------------------------------+
3599| K | | Thread reaped, exited due to signal. |
3600+------+-----+-----------------------------------------------------------+
3601
36214730
SW
3602..
3603 Example output was based on the following:
3604 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --runtime=58m \
3605 --time_based --rate=2512k --bs=256K --numjobs=10 \
3606 --name=readers --rw=read --name=writers --rw=write
3607
f80dba8d 3608Fio will condense the thread string as not to take up more space on the command
36214730 3609line than needed. For instance, if you have 10 readers and 10 writers running,
f80dba8d
MT
3610the output would look like this::
3611
9d25d068 3612 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 3613
36214730
SW
3614Note that the status string is displayed in order, so it's possible to tell which of
3615the jobs are currently doing what. In the example above this means that jobs 1--10
3616are readers and 11--20 are writers.
f80dba8d
MT
3617
3618The other values are fairly self explanatory -- number of threads currently
36214730
SW
3619running and doing I/O, the number of currently open files (f=), the estimated
3620completion percentage, the rate of I/O since last check (read speed listed first,
f50fbdda
TK
3621then write speed and optionally trim speed) in terms of bandwidth and IOPS,
3622and time to completion for the current running group. It's impossible to estimate
3623runtime of the following groups (if any).
36214730
SW
3624
3625..
3626 Example output was based on the following:
3627 TZ=UTC fio --iodepth=16 --ioengine=posixaio --filename=/tmp/fiofile \
3628 --direct=1 --size=100M --time_based --runtime=50s --rate_iops=89 \
3629 --bs=7K --name=Client1 --rw=write
3630
3631When fio is done (or interrupted by :kbd:`Ctrl-C`), it will show the data for
3632each thread, group of threads, and disks in that order. For each overall thread (or
3633group) the output looks like::
3634
3635 Client1: (groupid=0, jobs=1): err= 0: pid=16109: Sat Jun 24 12:07:54 2017
3636 write: IOPS=88, BW=623KiB/s (638kB/s)(30.4MiB/50032msec)
3637 slat (nsec): min=500, max=145500, avg=8318.00, stdev=4781.50
3638 clat (usec): min=170, max=78367, avg=4019.02, stdev=8293.31
3639 lat (usec): min=174, max=78375, avg=4027.34, stdev=8291.79
3640 clat percentiles (usec):
3641 | 1.00th=[ 302], 5.00th=[ 326], 10.00th=[ 343], 20.00th=[ 363],
3642 | 30.00th=[ 392], 40.00th=[ 404], 50.00th=[ 416], 60.00th=[ 445],
3643 | 70.00th=[ 816], 80.00th=[ 6718], 90.00th=[12911], 95.00th=[21627],
3644 | 99.00th=[43779], 99.50th=[51643], 99.90th=[68682], 99.95th=[72877],
3645 | 99.99th=[78119]
3646 bw ( KiB/s): min= 532, max= 686, per=0.10%, avg=622.87, stdev=24.82, samples= 100
3647 iops : min= 76, max= 98, avg=88.98, stdev= 3.54, samples= 100
29092211
VF
3648 lat (usec) : 250=0.04%, 500=64.11%, 750=4.81%, 1000=2.79%
3649 lat (msec) : 2=4.16%, 4=1.84%, 10=4.90%, 20=11.33%, 50=5.37%
3650 lat (msec) : 100=0.65%
36214730
SW
3651 cpu : usr=0.27%, sys=0.18%, ctx=12072, majf=0, minf=21
3652 IO depths : 1=85.0%, 2=13.1%, 4=1.8%, 8=0.1%, 16=0.0%, 32=0.0%, >=64=0.0%
3653 submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3654 complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3655 issued rwt: total=0,4450,0, short=0,0,0, dropped=0,0,0
3656 latency : target=0, window=0, percentile=100.00%, depth=8
3657
3658The job name (or first job's name when using :option:`group_reporting`) is printed,
3659along with the group id, count of jobs being aggregated, last error id seen (which
3660is 0 when there are no errors), pid/tid of that thread and the time the job/group
3661completed. Below are the I/O statistics for each data direction performed (showing
3662writes in the example above). In the order listed, they denote:
3663
3664**read/write/trim**
3665 The string before the colon shows the I/O direction the statistics
3666 are for. **IOPS** is the average I/Os performed per second. **BW**
3667 is the average bandwidth rate shown as: value in power of 2 format
3668 (value in power of 10 format). The last two values show: (**total
3669 I/O performed** in power of 2 format / **runtime** of that thread).
f80dba8d
MT
3670
3671**slat**
36214730
SW
3672 Submission latency (**min** being the minimum, **max** being the
3673 maximum, **avg** being the average, **stdev** being the standard
3674 deviation). This is the time it took to submit the I/O. For
3675 sync I/O this row is not displayed as the slat is really the
3676 completion latency (since queue/complete is one operation there).
3677 This value can be in nanoseconds, microseconds or milliseconds ---
3678 fio will choose the most appropriate base and print that (in the
3679 example above nanoseconds was the best scale). Note: in :option:`--minimal` mode
0d237712 3680 latencies are always expressed in microseconds.
f80dba8d
MT
3681
3682**clat**
3683 Completion latency. Same names as slat, this denotes the time from
3684 submission to completion of the I/O pieces. For sync I/O, clat will
3685 usually be equal (or very close) to 0, as the time from submit to
3686 complete is basically just CPU time (I/O has already been done, see slat
3687 explanation).
3688
29092211
VF
3689**lat**
3690 Total latency. Same names as slat and clat, this denotes the time from
3691 when fio created the I/O unit to completion of the I/O operation.
3692
f80dba8d 3693**bw**
36214730
SW
3694 Bandwidth statistics based on samples. Same names as the xlat stats,
3695 but also includes the number of samples taken (**samples**) and an
3696 approximate percentage of total aggregate bandwidth this thread
3697 received in its group (**per**). This last value is only really
3698 useful if the threads in this group are on the same disk, since they
3699 are then competing for disk access.
3700
3701**iops**
3702 IOPS statistics based on samples. Same names as bw.
f80dba8d 3703
29092211
VF
3704**lat (nsec/usec/msec)**
3705 The distribution of I/O completion latencies. This is the time from when
3706 I/O leaves fio and when it gets completed. Unlike the separate
3707 read/write/trim sections above, the data here and in the remaining
3708 sections apply to all I/Os for the reporting group. 250=0.04% means that
3709 0.04% of the I/Os completed in under 250us. 500=64.11% means that 64.11%
3710 of the I/Os required 250 to 499us for completion.
3711
f80dba8d
MT
3712**cpu**
3713 CPU usage. User and system time, along with the number of context
3714 switches this thread went through, usage of system and user time, and
3715 finally the number of major and minor page faults. The CPU utilization
3716 numbers are averages for the jobs in that reporting group, while the
23a8e176 3717 context and fault counters are summed.
f80dba8d
MT
3718
3719**IO depths**
a2140525
SW
3720 The distribution of I/O depths over the job lifetime. The numbers are
3721 divided into powers of 2 and each entry covers depths from that value
3722 up to those that are lower than the next entry -- e.g., 16= covers
3723 depths from 16 to 31. Note that the range covered by a depth
3724 distribution entry can be different to the range covered by the
3725 equivalent submit/complete distribution entry.
f80dba8d
MT
3726
3727**IO submit**
3728 How many pieces of I/O were submitting in a single submit call. Each
c60ebc45 3729 entry denotes that amount and below, until the previous entry -- e.g.,
a2140525
SW
3730 16=100% means that we submitted anywhere between 9 to 16 I/Os per submit
3731 call. Note that the range covered by a submit distribution entry can
3732 be different to the range covered by the equivalent depth distribution
3733 entry.
f80dba8d
MT
3734
3735**IO complete**
3736 Like the above submit number, but for completions instead.
3737
36214730
SW
3738**IO issued rwt**
3739 The number of read/write/trim requests issued, and how many of them were
3740 short or dropped.
f80dba8d 3741
29092211 3742**IO latency**
ee21ebee 3743 These values are for :option:`latency_target` and related options. When
29092211
VF
3744 these options are engaged, this section describes the I/O depth required
3745 to meet the specified latency target.
71bfa161 3746
36214730
SW
3747..
3748 Example output was based on the following:
3749 TZ=UTC fio --ioengine=null --iodepth=2 --size=100M --numjobs=2 \
3750 --rate_process=poisson --io_limit=32M --name=read --bs=128k \
3751 --rate=11M --name=write --rw=write --bs=2k --rate=700k
3752
71bfa161 3753After each client has been listed, the group statistics are printed. They
f80dba8d 3754will look like this::
71bfa161 3755
f80dba8d 3756 Run status group 0 (all jobs):
36214730
SW
3757 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
3758 WRITE: bw=1231KiB/s (1261kB/s), 616KiB/s-621KiB/s (630kB/s-636kB/s), io=64.0MiB (67.1MB), run=52747-53223msec
71bfa161 3759
36214730 3760For each data direction it prints:
71bfa161 3761
36214730
SW
3762**bw**
3763 Aggregate bandwidth of threads in this group followed by the
3764 minimum and maximum bandwidth of all the threads in this group.
3765 Values outside of brackets are power-of-2 format and those
3766 within are the equivalent value in a power-of-10 format.
f80dba8d 3767**io**
36214730
SW
3768 Aggregate I/O performed of all threads in this group. The
3769 format is the same as bw.
3770**run**
3771 The smallest and longest runtimes of the threads in this group.
71bfa161 3772
f50fbdda 3773And finally, the disk statistics are printed. This is Linux specific. They will look like this::
71bfa161 3774
f80dba8d
MT
3775 Disk stats (read/write):
3776 sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00%
71bfa161
JA
3777
3778Each value is printed for both reads and writes, with reads first. The
3779numbers denote:
3780
f80dba8d 3781**ios**
c60ebc45 3782 Number of I/Os performed by all groups.
f80dba8d 3783**merge**
007c7be9 3784 Number of merges performed by the I/O scheduler.
f80dba8d
MT
3785**ticks**
3786 Number of ticks we kept the disk busy.
36214730 3787**in_queue**
f80dba8d
MT
3788 Total time spent in the disk queue.
3789**util**
3790 The disk utilization. A value of 100% means we kept the disk
71bfa161
JA
3791 busy constantly, 50% would be a disk idling half of the time.
3792
f80dba8d
MT
3793It is also possible to get fio to dump the current output while it is running,
3794without terminating the job. To do that, send fio the **USR1** signal. You can
3795also get regularly timed dumps by using the :option:`--status-interval`
3796parameter, or by creating a file in :file:`/tmp` named
3797:file:`fio-dump-status`. If fio sees this file, it will unlink it and dump the
3798current output status.
8423bd11 3799
71bfa161 3800
f80dba8d
MT
3801Terse output
3802------------