4 The first step in getting fio to simulate a desired I/O workload, is writing a
5 job file describing that specific setup. A job file may contain any number of
6 threads and/or files -- the typical contents of the job file is a *global*
7 section defining shared parameters, and one or more job sections describing the
8 jobs involved. When run, fio parses this file and sets everything up as
9 described. If we break down a job from top to bottom, it contains the following
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?
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.
26 How much data are we going to be reading/writing.
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
36 If the I/O engine is async, how large a queuing depth do we want to
42 How many files are we spreading the workload over.
44 `Threads, processes and job synchronization`_
46 How many threads or processes should we spread this workload over.
48 The above are the basic parameters defined for a workload, in addition there's a
49 multitude of parameters that modify other aspects of how this job behaves.
55 .. option:: --debug=type
57 Enable verbose tracing `type` of various fio actions. May be ``all`` for all types
58 or individual types separated by a comma (e.g. ``--debug=file,mem`` will
59 enable file and memory debugging). Currently, additional logging is
63 Dump info related to processes.
65 Dump info related to file actions.
67 Dump info related to I/O queuing.
69 Dump info related to memory allocations.
71 Dump info related to blktrace setup.
73 Dump info related to I/O verification.
75 Enable all debug options.
77 Dump info related to random offset generation.
79 Dump info related to option matching and parsing.
81 Dump info related to disk utilization updates.
83 Dump info only related to job number x.
85 Dump info only related to mutex up/down ops.
87 Dump info related to profile extensions.
89 Dump info related to internal time keeping.
91 Dump info related to networking connections.
93 Dump info related to I/O rate switching.
95 Dump info related to log compress/decompress.
97 Dump info related to steadystate detection.
99 Dump info related to the helper thread.
101 Dump info related to support for zoned block devices.
103 Show available debug options.
105 .. option:: --parse-only
107 Parse options only, don't start any I/O.
109 .. option:: --merge-blktrace-only
111 Merge blktraces only, don't start any I/O.
113 .. option:: --output=filename
115 Write output to file `filename`.
117 .. option:: --output-format=format
119 Set the reporting `format` to `normal`, `terse`, `json`, or `json+`. Multiple
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
124 .. option:: --bandwidth-log
126 Generate aggregate bandwidth logs.
128 .. option:: --minimal
130 Print statistics in a terse, semicolon-delimited format.
132 .. option:: --append-terse
134 Print statistics in selected mode AND terse, semicolon-delimited format.
135 **Deprecated**, use :option:`--output-format` instead to select multiple
138 .. option:: --terse-version=version
140 Set terse `version` output format (default 3, or 2 or 4 or 5).
142 .. option:: --version
144 Print version information and exit.
148 Print a summary of the command line options and exit.
150 .. option:: --cpuclock-test
152 Perform test and validation of internal CPU clock.
154 .. option:: --crctest=[test]
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.
160 .. option:: --cmdhelp=command
162 Print help information for `command`. May be ``all`` for all commands.
164 .. option:: --enghelp=[ioengine[,command]]
166 List all commands defined by `ioengine`, or print help for `command`
167 defined by `ioengine`. If no `ioengine` is given, list all
170 .. option:: --showcmd
172 Convert given job files to a set of command-line options.
174 .. option:: --readonly
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.
183 .. option:: --eta=when
185 Specifies when real-time ETA estimate should be printed. `when` may be
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.
190 .. option:: --eta-interval=time
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
197 .. option:: --eta-newline=time
199 Force a new line for every `time` period passed. When the unit is omitted,
200 the value is interpreted in seconds.
202 .. option:: --status-interval=time
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,
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.
212 .. option:: --section=name
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
223 .. option:: --alloc-size=kb
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.
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.
231 NOTE: While running :file:`.fio_smalloc.*` backing store files are visible
234 .. option:: --warnings-fatal
236 All fio parser warnings are fatal, causing fio to exit with an
239 .. option:: --max-jobs=nr
241 Set the maximum number of threads/processes to support to `nr`.
242 NOTE: On Linux, it may be necessary to increase the shared-memory
243 limit (:file:`/proc/sys/kernel/shmmax`) if fio runs into errors while
246 .. option:: --server=args
248 Start a backend server, with `args` specifying what to listen to.
249 See `Client/Server`_ section.
251 .. option:: --daemonize=pidfile
253 Background a fio server, writing the pid to the given `pidfile` file.
255 .. option:: --client=hostname
257 Instead of running the jobs locally, send and run them on the given `hostname`
258 or set of `hostname`\s. See `Client/Server`_ section.
260 .. option:: --remote-config=file
262 Tell fio server to load this local `file`.
264 .. option:: --idle-prof=option
266 Report CPU idleness. `option` is one of the following:
269 Run unit work calibration only and exit.
272 Show aggregate system idleness and unit work.
275 As **system** but also show per CPU idleness.
277 .. option:: --inflate-log=log
279 Inflate and output compressed `log`.
281 .. option:: --trigger-file=file
283 Execute trigger command when `file` exists.
285 .. option:: --trigger-timeout=time
287 Execute trigger at this `time`.
289 .. option:: --trigger=command
291 Set this `command` as local trigger.
293 .. option:: --trigger-remote=command
295 Set this `command` as remote trigger.
297 .. option:: --aux-path=path
299 Use the directory specified by `path` for generated state files instead
300 of the current working directory.
302 Any parameters following the options will be assumed to be job files, unless
303 they match a job file parameter. Multiple job files can be listed and each job
304 file will be regarded as a separate group. Fio will :option:`stonewall`
305 execution between each group.
311 As previously described, fio accepts one or more job files describing what it is
312 supposed to do. The job file format is the classic ini file, where the names
313 enclosed in [] brackets define the job name. You are free to use any ASCII name
314 you want, except *global* which has special meaning. Following the job name is
315 a sequence of zero or more parameters, one per line, that define the behavior of
316 the job. If the first character in a line is a ';' or a '#', the entire line is
317 discarded as a comment.
319 A *global* section sets defaults for the jobs described in that file. A job may
320 override 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
324 The :option:`--cmdhelp` option also lists all options. If used with a `command`
325 argument, :option:`--cmdhelp` will detail the given `command`.
327 See the `examples/` directory for inspiration on how to write job files. Note
328 the copyright and license requirements currently apply to `examples/` files.
330 So let's look at a really simple job file that defines two processes, each
331 randomly reading from a 128MiB file:
335 ; -- start job file --
346 As you can see, the job file sections themselves are empty as all the described
347 parameters 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
349 would look as follows::
351 $ fio --name=global --rw=randread --size=128m --name=job1 --name=job2
354 Let's look at an example that has a number of processes writing randomly to
359 ; -- start job file --
370 Here we have no *global* section, as we only have one job defined anyway. We
371 want to use async I/O here, with a depth of 4 for each file. We also increased
372 the buffer size used to 32KiB and define numjobs to 4 to fork 4 identical
373 jobs. The result is 4 processes each randomly writing to their own 64MiB
374 file. Instead of using the above job file, you could have given the parameters
375 on the command line. For this case, you would specify::
377 $ fio --name=random-writers --ioengine=libaio --iodepth=4 --rw=randwrite --bs=32k --direct=0 --size=64m --numjobs=4
379 When fio is utilized as a basis of any reasonably large test suite, it might be
380 desirable to share a set of standardized settings across multiple job files.
381 Instead 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
385 ; -- start job file including.fio --
389 include glob-include.fio
396 include test-include.fio
397 ; -- end job file including.fio --
401 ; -- start job file glob-include.fio --
404 ; -- end job file glob-include.fio --
408 ; -- start job file test-include.fio --
411 ; -- end job file test-include.fio --
413 Settings pulled into a section apply to that section only (except *global*
414 section). Include directives may be nested in that any included file may contain
415 further include directive(s). Include files may not contain [] sections.
418 Environment variables
419 ~~~~~~~~~~~~~~~~~~~~~
421 Fio also supports environment variable expansion in job files. Any sub-string of
422 the form ``${VARNAME}`` as part of an option value (in other words, on the right
423 of 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
425 empty string, the empty string will be substituted.
427 As an example, let's look at a sample fio invocation and job file::
429 $ SIZE=64m NUMJOBS=4 fio jobfile.fio
433 ; -- start job file --
440 This will expand to the following equivalent job file at runtime:
444 ; -- start job file --
451 Fio ships with a few example job files, you can also look there for inspiration.
456 Additionally, fio has a set of reserved keywords that will be replaced
457 internally with the appropriate value. Those keywords are:
461 The architecture page size of the running system.
465 Megabytes of total memory in the system.
469 Number of online available CPUs.
471 These can be used on the command line or in the job file, and will be
472 automatically substituted with the current system values when the job is
473 run. Simple math is also supported on these keywords, so you can perform actions
478 and get that properly expanded to 8 times the size of memory in the machine.
484 This section describes in details each parameter associated with a job. Some
485 parameters take an option of a given type, such as an integer or a
486 string. Anywhere a numeric value is required, an arithmetic expression may be
487 used, provided it is surrounded by parentheses. Supported operators are:
496 For time values in expressions, units are microseconds by default. This is
497 different than for time values not in expressions (not enclosed in
498 parentheses). The following types are used:
505 String: A sequence of alphanumeric characters.
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.
516 Integer. A whole number value, which may contain an integer prefix
517 and an integer suffix:
519 [*integer prefix*] **number** [*integer suffix*]
521 The optional *integer prefix* specifies the number's base. The default
522 is decimal. *0x* specifies hexadecimal.
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
526 default unit is bytes. For quantities of time, the default unit is seconds
527 unless otherwise specified.
529 With :option:`kb_base`\=1000, fio follows international standards for unit
530 prefixes. To specify power-of-10 decimal values defined in the
531 International System of Units (SI):
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
539 To specify power-of-2 binary values defined in IEC 80000-13:
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
547 For Zone Block Device Mode:
550 With :option:`kb_base`\=1024 (the default), the unit prefixes are opposite
551 from those specified in the SI and IEC 80000-13 standards to provide
552 compatibility with old scripts. For example, 4k means 4096.
554 For quantities of data, an optional unit of 'B' may be included
555 (e.g., 'kB' is the same as 'k').
557 The *integer suffix* is not case sensitive (e.g., m/mi mean mebi/mega,
558 not milli). 'b' and 'B' both mean byte, not bit.
560 Examples with :option:`kb_base`\=1000:
562 * *4 KiB*: 4096, 4096b, 4096B, 4ki, 4kib, 4kiB, 4Ki, 4KiB
563 * *1 MiB*: 1048576, 1mi, 1024ki
564 * *1 MB*: 1000000, 1m, 1000k
565 * *1 TiB*: 1099511627776, 1ti, 1024gi, 1048576mi
566 * *1 TB*: 1000000000, 1t, 1000m, 1000000k
568 Examples with :option:`kb_base`\=1024 (default):
570 * *4 KiB*: 4096, 4096b, 4096B, 4k, 4kb, 4kB, 4K, 4KB
571 * *1 MiB*: 1048576, 1m, 1024k
572 * *1 MB*: 1000000, 1mi, 1000ki
573 * *1 TiB*: 1099511627776, 1t, 1024g, 1048576m
574 * *1 TB*: 1000000000, 1ti, 1000mi, 1000000ki
576 To specify times (units are not case sensitive):
580 * *M* -- means minutes
581 * *s* -- or sec means seconds (default)
582 * *ms* -- or *msec* means milliseconds
583 * *us* -- or *usec* means microseconds
585 If the option accepts an upper and lower range, use a colon ':' or
586 minus '-' to separate such values. See :ref:`irange <irange>`.
587 If the lower value specified happens to be larger than the upper value
588 the two values are swapped.
593 Boolean. Usually parsed as an integer, however only defined for
594 true and false (1 and 0).
599 Integer range with suffix. Allows value range to be given, such as
600 1024-4096. A colon may also be used as the separator, e.g. 1k:4k. If the
601 option allows two sets of ranges, they can be specified with a ',' or '/'
602 delimiter: 1k-4k/8k-32k. Also see :ref:`int <int>`.
605 A list of floating point numbers, separated by a ':' character.
607 With the above in mind, here follows the complete list of fio job parameters.
613 .. option:: kb_base=int
615 Select the interpretation of unit prefixes in input parameters.
618 Inputs comply with IEC 80000-13 and the International
619 System of Units (SI). Use:
621 - power-of-2 values with IEC prefixes (e.g., KiB)
622 - power-of-10 values with SI prefixes (e.g., kB)
625 Compatibility mode (default). To avoid breaking old scripts:
627 - power-of-2 values with SI prefixes
628 - power-of-10 values with IEC prefixes
630 See :option:`bs` for more details on input parameters.
632 Outputs always use correct prefixes. Most outputs include both
635 bw=2383.3kB/s (2327.4KiB/s)
637 If only one value is reported, then kb_base selects the one to use:
639 **1000** -- SI prefixes
641 **1024** -- IEC prefixes
643 .. option:: unit_base=int
645 Base unit for reporting. Allowed values are:
648 Use auto-detection (default).
660 ASCII name of the job. This may be used to override the name printed by fio
661 for this job. Otherwise the job name is used. On the command line this
662 parameter has the special purpose of also signaling the start of a new job.
664 .. option:: description=str
666 Text description of the job. Doesn't do anything except dump this text
667 description when this job is run. It's not parsed.
669 .. option:: loops=int
671 Run the specified number of iterations of this job. Used to repeat the same
672 workload a given number of times. Defaults to 1.
674 .. option:: numjobs=int
676 Create the specified number of clones of this job. Each clone of job
677 is spawned as an independent thread or process. May be used to setup a
678 larger number of threads/processes doing the same thing. Each thread is
679 reported separately; to see statistics for all clones as a whole, use
680 :option:`group_reporting` in conjunction with :option:`new_group`.
681 See :option:`--max-jobs`. Default: 1.
684 Time related parameters
685 ~~~~~~~~~~~~~~~~~~~~~~~
687 .. option:: runtime=time
689 Tell fio to terminate processing after the specified period of time. It
690 can be quite hard to determine for how long a specified job will run, so
691 this parameter is handy to cap the total runtime to a given time. When
692 the unit is omitted, the value is interpreted in seconds.
694 .. option:: time_based
696 If set, fio will run for the duration of the :option:`runtime` specified
697 even if the file(s) are completely read or written. It will simply loop over
698 the same workload as many times as the :option:`runtime` allows.
700 .. option:: startdelay=irange(time)
702 Delay the start of job for the specified amount of time. Can be a single
703 value or a range. When given as a range, each thread will choose a value
704 randomly from within the range. Value is in seconds if a unit is omitted.
706 .. option:: ramp_time=time
708 If set, fio will run the specified workload for this amount of time before
709 logging any performance numbers. Useful for letting performance settle
710 before logging results, thus minimizing the runtime required for stable
711 results. Note that the ``ramp_time`` is considered lead in time for a job,
712 thus it will increase the total runtime if a special timeout or
713 :option:`runtime` is specified. When the unit is omitted, the value is
716 .. option:: clocksource=str
718 Use the given clocksource as the base of timing. The supported options are:
721 :manpage:`gettimeofday(2)`
724 :manpage:`clock_gettime(2)`
727 Internal CPU clock source
729 cpu is the preferred clocksource if it is reliable, as it is very fast (and
730 fio is heavy on time calls). Fio will automatically use this clocksource if
731 it's supported and considered reliable on the system it is running on,
732 unless another clocksource is specifically set. For x86/x86-64 CPUs, this
733 means supporting TSC Invariant.
735 .. option:: gtod_reduce=bool
737 Enable all of the :manpage:`gettimeofday(2)` reducing options
738 (:option:`disable_clat`, :option:`disable_slat`, :option:`disable_bw_measurement`) plus
739 reduce precision of the timeout somewhat to really shrink the
740 :manpage:`gettimeofday(2)` call count. With this option enabled, we only do
741 about 0.4% of the :manpage:`gettimeofday(2)` calls we would have done if all
742 time keeping was enabled.
744 .. option:: gtod_cpu=int
746 Sometimes it's cheaper to dedicate a single thread of execution to just
747 getting the current time. Fio (and databases, for instance) are very
748 intensive on :manpage:`gettimeofday(2)` calls. With this option, you can set
749 one CPU aside for doing nothing but logging current time to a shared memory
750 location. Then the other threads/processes that run I/O workloads need only
751 copy that segment, instead of entering the kernel with a
752 :manpage:`gettimeofday(2)` call. The CPU set aside for doing these time
753 calls will be excluded from other uses. Fio will manually clear it from the
754 CPU mask of other jobs.
760 .. option:: directory=str
762 Prefix filenames with this directory. Used to place files in a different
763 location than :file:`./`. You can specify a number of directories by
764 separating the names with a ':' character. These directories will be
765 assigned equally distributed to job clones created by :option:`numjobs` as
766 long as they are using generated filenames. If specific `filename(s)` are
767 set fio will use the first listed directory, and thereby matching the
768 `filename` semantic (which generates a file for each clone if not
769 specified, but lets all clones use the same file if set).
771 See the :option:`filename` option for information on how to escape "``:``"
772 characters within the directory path itself.
774 Note: To control the directory fio will use for internal state files
775 use :option:`--aux-path`.
777 .. option:: filename=str
779 Fio normally makes up a `filename` based on the job name, thread number, and
780 file number (see :option:`filename_format`). If you want to share files
781 between threads in a job or several
782 jobs with fixed file paths, specify a `filename` for each of them to override
783 the default. If the ioengine is file based, you can specify a number of files
784 by separating the names with a ':' colon. So if you wanted a job to open
785 :file:`/dev/sda` and :file:`/dev/sdb` as the two working files, you would use
786 ``filename=/dev/sda:/dev/sdb``. This also means that whenever this option is
787 specified, :option:`nrfiles` is ignored. The size of regular files specified
788 by this option will be :option:`size` divided by number of files unless an
789 explicit size is specified by :option:`filesize`.
791 Each colon in the wanted path must be escaped with a ``\``
792 character. For instance, if the path is :file:`/dev/dsk/foo@3,0:c` then you
793 would use ``filename=/dev/dsk/foo@3,0\:c`` and if the path is
794 :file:`F:\\filename` then you would use ``filename=F\:\filename``.
796 On Windows, disk devices are accessed as :file:`\\\\.\\PhysicalDrive0` for
797 the first device, :file:`\\\\.\\PhysicalDrive1` for the second etc.
798 Note: Windows and FreeBSD prevent write access to areas
799 of the disk containing in-use data (e.g. filesystems).
801 The filename "`-`" is a reserved name, meaning *stdin* or *stdout*. Which
802 of the two depends on the read/write direction set.
804 .. option:: filename_format=str
806 If sharing multiple files between jobs, it is usually necessary to have fio
807 generate the exact names that you want. By default, fio will name a file
808 based on the default file format specification of
809 :file:`jobname.jobnumber.filenumber`. With this option, that can be
810 customized. Fio will recognize and replace the following keywords in this
814 The name of the worker thread or process.
816 IP of the fio process when using client/server mode.
818 The incremental number of the worker thread or process.
820 The incremental number of the file for that worker thread or
823 To have dependent jobs share a set of files, this option can be set to have
824 fio generate filenames that are shared between the two. For instance, if
825 :file:`testfiles.$filenum` is specified, file number 4 for any job will be
826 named :file:`testfiles.4`. The default of :file:`$jobname.$jobnum.$filenum`
827 will be used if no other format specifier is given.
829 If you specify a path then the directories will be created up to the
830 main directory for the file. So for example if you specify
831 ``filename_format=a/b/c/$jobnum`` then the directories a/b/c will be
832 created before the file setup part of the job. If you specify
833 :option:`directory` then the path will be relative that directory,
834 otherwise it is treated as the absolute path.
836 .. option:: unique_filename=bool
838 To avoid collisions between networked clients, fio defaults to prefixing any
839 generated filenames (with a directory specified) with the source of the
840 client connecting. To disable this behavior, set this option to 0.
842 .. option:: opendir=str
844 Recursively open any files below directory `str`.
846 .. option:: lockfile=str
848 Fio defaults to not locking any files before it does I/O to them. If a file
849 or file descriptor is shared, fio can serialize I/O to that file to make the
850 end result consistent. This is usual for emulating real workloads that share
851 files. The lock modes are:
854 No locking. The default.
856 Only one thread or process may do I/O at a time, excluding all
859 Read-write locking on the file. Many readers may
860 access the file at the same time, but writes get exclusive access.
862 .. option:: nrfiles=int
864 Number of files to use for this job. Defaults to 1. The size of files
865 will be :option:`size` divided by this unless explicit size is specified by
866 :option:`filesize`. Files are created for each thread separately, and each
867 file will have a file number within its name by default, as explained in
868 :option:`filename` section.
871 .. option:: openfiles=int
873 Number of files to keep open at the same time. Defaults to the same as
874 :option:`nrfiles`, can be set smaller to limit the number simultaneous
877 .. option:: file_service_type=str
879 Defines how fio decides which file from a job to service next. The following
883 Choose a file at random.
886 Round robin over opened files. This is the default.
889 Finish one file before moving on to the next. Multiple files can
890 still be open depending on :option:`openfiles`.
893 Use a *Zipf* distribution to decide what file to access.
896 Use a *Pareto* distribution to decide what file to access.
899 Use a *Gaussian* (normal) distribution to decide what file to
905 For *random*, *roundrobin*, and *sequential*, a postfix can be appended to
906 tell fio how many I/Os to issue before switching to a new file. For example,
907 specifying ``file_service_type=random:8`` would cause fio to issue
908 8 I/Os before selecting a new file at random. For the non-uniform
909 distributions, a floating point postfix can be given to influence how the
910 distribution is skewed. See :option:`random_distribution` for a description
911 of how that would work.
913 .. option:: ioscheduler=str
915 Attempt to switch the device hosting the file to the specified I/O scheduler
918 .. option:: create_serialize=bool
920 If true, serialize the file creation for the jobs. This may be handy to
921 avoid interleaving of data files, which may greatly depend on the filesystem
922 used and even the number of processors in the system. Default: true.
924 .. option:: create_fsync=bool
926 :manpage:`fsync(2)` the data file after creation. This is the default.
928 .. option:: create_on_open=bool
930 If true, don't pre-create files but allow the job's open() to create a file
931 when it's time to do I/O. Default: false -- pre-create all necessary files
934 .. option:: create_only=bool
936 If true, fio will only run the setup phase of the job. If files need to be
937 laid out or updated on disk, only that will be done -- the actual job contents
938 are not executed. Default: false.
940 .. option:: allow_file_create=bool
942 If true, fio is permitted to create files as part of its workload. If this
943 option is false, then fio will error out if
944 the files it needs to use don't already exist. Default: true.
946 .. option:: allow_mounted_write=bool
948 If this isn't set, fio will abort jobs that are destructive (e.g. that write)
949 to what appears to be a mounted device or partition. This should help catch
950 creating inadvertently destructive tests, not realizing that the test will
951 destroy data on the mounted file system. Note that some platforms don't allow
952 writing against a mounted device regardless of this option. Default: false.
954 .. option:: pre_read=bool
956 If this is given, files will be pre-read into memory before starting the
957 given I/O operation. This will also clear the :option:`invalidate` flag,
958 since it is pointless to pre-read and then drop the cache. This will only
959 work for I/O engines that are seek-able, since they allow you to read the
960 same data multiple times. Thus it will not work on non-seekable I/O engines
961 (e.g. network, splice). Default: false.
963 .. option:: unlink=bool
965 Unlink the job files when done. Not the default, as repeated runs of that
966 job would then waste time recreating the file set again and again. Default:
969 .. option:: unlink_each_loop=bool
971 Unlink job files after each iteration or loop. Default: false.
973 .. option:: zonemode=str
978 The :option:`zonerange`, :option:`zonesize`,
979 :option `zonecapacity` and option:`zoneskip`
980 parameters are ignored.
982 I/O happens in a single zone until
983 :option:`zonesize` bytes have been transferred.
984 After that number of bytes has been
985 transferred processing of the next zone
986 starts. :option `zonecapacity` is ignored.
988 Zoned block device mode. I/O happens
989 sequentially in each zone, even if random I/O
990 has been selected. Random I/O happens across
991 all zones instead of being restricted to a
992 single zone. The :option:`zoneskip` parameter
993 is ignored. :option:`zonerange` and
994 :option:`zonesize` must be identical.
995 Trim is handled using a zone reset operation.
996 Trim only considers non-empty sequential write
997 required and sequential write preferred zones.
999 .. option:: zonerange=int
1001 Size of a single zone. See also :option:`zonesize` and
1004 .. option:: zonesize=int
1006 For :option:`zonemode` =strided, this is the number of bytes to
1007 transfer before skipping :option:`zoneskip` bytes. If this parameter
1008 is smaller than :option:`zonerange` then only a fraction of each zone
1009 with :option:`zonerange` bytes will be accessed. If this parameter is
1010 larger than :option:`zonerange` then each zone will be accessed
1011 multiple times before skipping to the next zone.
1013 For :option:`zonemode` =zbd, this is the size of a single zone. The
1014 :option:`zonerange` parameter is ignored in this mode.
1017 .. option:: zonecapacity=int
1019 For :option:`zonemode` =zbd, this defines the capacity of a single zone,
1020 which is the accessible area starting from the zone start address.
1021 This parameter only applies when using :option:`zonemode` =zbd in
1022 combination with regular block devices. If not specified it defaults to
1023 the zone size. If the target device is a zoned block device, the zone
1024 capacity is obtained from the device information and this option is
1027 .. option:: zoneskip=int
1029 For :option:`zonemode` =strided, the number of bytes to skip after
1030 :option:`zonesize` bytes of data have been transferred. This parameter
1031 must be zero for :option:`zonemode` =zbd.
1033 .. option:: read_beyond_wp=bool
1035 This parameter applies to :option:`zonemode` =zbd only.
1037 Zoned block devices are block devices that consist of multiple zones.
1038 Each zone has a type, e.g. conventional or sequential. A conventional
1039 zone can be written at any offset that is a multiple of the block
1040 size. Sequential zones must be written sequentially. The position at
1041 which a write must occur is called the write pointer. A zoned block
1042 device can be either drive managed, host managed or host aware. For
1043 host managed devices the host must ensure that writes happen
1044 sequentially. Fio recognizes host managed devices and serializes
1045 writes to sequential zones for these devices.
1047 If a read occurs in a sequential zone beyond the write pointer then
1048 the zoned block device will complete the read without reading any data
1049 from the storage medium. Since such reads lead to unrealistically high
1050 bandwidth and IOPS numbers fio only reads beyond the write pointer if
1051 explicitly told to do so. Default: false.
1053 .. option:: max_open_zones=int
1055 A zone of a zoned block device is in the open state when it is partially
1056 written (i.e. not all sectors of the zone have been written). Zoned
1057 block devices may have a limit on the total number of zones that can
1058 be simultaneously in the open state, that is, the number of zones that
1059 can be written to simultaneously. The :option:`max_open_zones` parameter
1060 limits the number of zones to which write commands are issued by all fio
1061 jobs, that is, limits the number of zones that will be in the open
1062 state. This parameter is relevant only if the :option:`zonemode` =zbd is
1063 used. The default value is always equal to maximum number of open zones
1064 of the target zoned block device and a value higher than this limit
1065 cannot be specified by users unless the option
1066 :option:`ignore_zone_limits` is specified. When
1067 :option:`ignore_zone_limits` is specified or the target device has no
1068 limit on the number of zones that can be in an open state,
1069 :option:`max_open_zones` can specify 0 to disable any limit on the
1070 number of zones that can be simultaneously written to by all jobs.
1072 .. option:: job_max_open_zones=int
1074 In the same manner as :option:`max_open_zones`, limit the number of open
1075 zones per fio job, that is, the number of zones that a single job can
1076 simultaneously write to. A value of zero indicates no limit.
1079 .. option:: ignore_zone_limits=bool
1081 If this option is used, fio will ignore the maximum number of open
1082 zones limit of the zoned block device in use, thus allowing the
1083 option :option:`max_open_zones` value to be larger than the device
1084 reported limit. Default: false.
1086 .. option:: zone_reset_threshold=float
1088 A number between zero and one that indicates the ratio of logical
1089 blocks with data to the total number of logical blocks in the test
1090 above which zones should be reset periodically.
1092 .. option:: zone_reset_frequency=float
1094 A number between zero and one that indicates how often a zone reset
1095 should be issued if the zone reset threshold has been exceeded. A zone
1096 reset is submitted after each (1 / zone_reset_frequency) write
1097 requests. This and the previous parameter can be used to simulate
1098 garbage collection activity.
1104 .. option:: direct=bool
1106 If value is true, use non-buffered I/O. This is usually O_DIRECT. Note that
1107 OpenBSD and ZFS on Solaris don't support direct I/O. On Windows the synchronous
1108 ioengines don't support direct I/O. Default: false.
1110 .. option:: atomic=bool
1112 If value is true, attempt to use atomic direct I/O. Atomic writes are
1113 guaranteed to be stable once acknowledged by the operating system. Only
1114 Linux supports O_ATOMIC right now.
1116 .. option:: buffered=bool
1118 If value is true, use buffered I/O. This is the opposite of the
1119 :option:`direct` option. Defaults to true.
1121 .. option:: readwrite=str, rw=str
1123 Type of I/O pattern. Accepted values are:
1130 Sequential trims (Linux block devices and SCSI
1131 character devices only).
1137 Random trims (Linux block devices and SCSI
1138 character devices only).
1140 Sequential mixed reads and writes.
1142 Random mixed reads and writes.
1144 Sequential trim+write sequences. Blocks will be trimmed first,
1145 then the same blocks will be written to. So if ``io_size=64K``
1146 is specified, Fio will trim a total of 64K bytes and also
1147 write 64K bytes on the same trimmed blocks. This behaviour
1148 will be consistent with ``number_ios`` or other Fio options
1149 limiting the total bytes or number of I/O's.
1151 Like trimwrite, but uses random offsets rather
1152 than sequential writes.
1154 Fio defaults to read if the option is not specified. For the mixed I/O
1155 types, the default is to split them 50/50. For certain types of I/O the
1156 result may still be skewed a bit, since the speed may be different.
1158 It is possible to specify the number of I/Os to do before getting a new
1159 offset by appending ``:<nr>`` to the end of the string given. For a
1160 random read, it would look like ``rw=randread:8`` for passing in an offset
1161 modifier with a value of 8. If the suffix is used with a sequential I/O
1162 pattern, then the *<nr>* value specified will be **added** to the generated
1163 offset for each I/O turning sequential I/O into sequential I/O with holes.
1164 For instance, using ``rw=write:4k`` will skip 4k for every write. Also see
1165 the :option:`rw_sequencer` option.
1167 .. option:: rw_sequencer=str
1169 If an offset modifier is given by appending a number to the ``rw=<str>``
1170 line, then this option controls how that number modifies the I/O offset
1171 being generated. Accepted values are:
1174 Generate sequential offset.
1176 Generate the same offset.
1178 ``sequential`` is only useful for random I/O, where fio would normally
1179 generate a new random offset for every I/O. If you append e.g. 8 to
1180 randread, i.e. ``rw=randread:8`` you would get a new random offset for
1181 every 8 I/Os. The result would be a sequence of 8 sequential offsets
1182 with a random starting point. However this behavior may change if a
1183 sequential I/O reaches end of the file. As sequential I/O is already
1184 sequential, setting ``sequential`` for that would not result in any
1185 difference. ``identical`` behaves in a similar fashion, except it sends
1186 the same offset 8 number of times before generating a new offset.
1191 rw_sequencer=sequential
1194 The generated sequence of offsets will look like this:
1195 4k, 8k, 12k, 16k, 20k, 24k, 28k, 32k, 92k, 96k, 100k, 104k, 108k,
1196 112k, 116k, 120k, 48k, 52k ...
1201 rw_sequencer=identical
1204 The generated sequence of offsets will look like this:
1205 4k, 4k, 4k, 4k, 4k, 4k, 4k, 4k, 92k, 92k, 92k, 92k, 92k, 92k, 92k, 92k,
1208 .. option:: unified_rw_reporting=str
1210 Fio normally reports statistics on a per data direction basis, meaning that
1211 reads, writes, and trims are accounted and reported separately. This option
1212 determines whether fio reports the results normally, summed together, or as
1214 Accepted values are:
1217 Normal statistics reporting.
1220 Statistics are summed per data direction and reported together.
1223 Statistics are reported normally, followed by the mixed statistics.
1226 Backward-compatible alias for **none**.
1229 Backward-compatible alias for **mixed**.
1234 .. option:: randrepeat=bool
1236 Seed the random number generator used for random I/O patterns in a
1237 predictable way so the pattern is repeatable across runs. Default: true.
1239 .. option:: allrandrepeat=bool
1241 Seed all random number generators in a predictable way so results are
1242 repeatable across runs. Default: false.
1244 .. option:: randseed=int
1246 Seed the random number generators based on this seed value, to be able to
1247 control what sequence of output is being generated. If not set, the random
1248 sequence depends on the :option:`randrepeat` setting.
1250 .. option:: fallocate=str
1252 Whether pre-allocation is performed when laying down files.
1253 Accepted values are:
1256 Do not pre-allocate space.
1259 Use a platform's native pre-allocation call but fall back to
1260 **none** behavior if it fails/is not implemented.
1263 Pre-allocate via :manpage:`posix_fallocate(3)`.
1266 Pre-allocate via :manpage:`fallocate(2)` with
1267 FALLOC_FL_KEEP_SIZE set.
1270 Extend file to final size via :manpage:`ftruncate(2)`
1271 instead of allocating.
1274 Backward-compatible alias for **none**.
1277 Backward-compatible alias for **posix**.
1279 May not be available on all supported platforms. **keep** is only available
1280 on Linux. If using ZFS on Solaris this cannot be set to **posix**
1281 because ZFS doesn't support pre-allocation. Default: **native** if any
1282 pre-allocation methods except **truncate** are available, **none** if not.
1284 Note that using **truncate** on Windows will interact surprisingly
1285 with non-sequential write patterns. When writing to a file that has
1286 been extended by setting the end-of-file information, Windows will
1287 backfill the unwritten portion of the file up to that offset with
1288 zeroes before issuing the new write. This means that a single small
1289 write to the end of an extended file will stall until the entire
1290 file has been filled with zeroes.
1292 .. option:: fadvise_hint=str
1294 Use :manpage:`posix_fadvise(2)` or :manpage:`posix_fadvise(2)` to
1295 advise the kernel on what I/O patterns are likely to be issued.
1296 Accepted values are:
1299 Backwards-compatible hint for "no hint".
1302 Backwards compatible hint for "advise with fio workload type". This
1303 uses **FADV_RANDOM** for a random workload, and **FADV_SEQUENTIAL**
1304 for a sequential workload.
1307 Advise using **FADV_SEQUENTIAL**.
1310 Advise using **FADV_RANDOM**.
1312 .. option:: write_hint=str
1314 Use :manpage:`fcntl(2)` to advise the kernel what life time to expect
1315 from a write. Only supported on Linux, as of version 4.13. Accepted
1319 No particular life time associated with this file.
1322 Data written to this file has a short life time.
1325 Data written to this file has a medium life time.
1328 Data written to this file has a long life time.
1331 Data written to this file has a very long life time.
1333 The values are all relative to each other, and no absolute meaning
1334 should be associated with them.
1336 .. option:: offset=int
1338 Start I/O at the provided offset in the file, given as either a fixed size in
1339 bytes, zones or a percentage. If a percentage is given, the generated offset will be
1340 aligned to the minimum ``blocksize`` or to the value of ``offset_align`` if
1341 provided. Data before the given offset will not be touched. This
1342 effectively caps the file size at `real_size - offset`. Can be combined with
1343 :option:`size` to constrain the start and end range of the I/O workload.
1344 A percentage can be specified by a number between 1 and 100 followed by '%',
1345 for example, ``offset=20%`` to specify 20%. In ZBD mode, value can be set as
1346 number of zones using 'z'.
1348 .. option:: offset_align=int
1350 If set to non-zero value, the byte offset generated by a percentage ``offset``
1351 is aligned upwards to this value. Defaults to 0 meaning that a percentage
1352 offset is aligned to the minimum block size.
1354 .. option:: offset_increment=int
1356 If this is provided, then the real offset becomes `offset + offset_increment
1357 * thread_number`, where the thread number is a counter that starts at 0 and
1358 is incremented for each sub-job (i.e. when :option:`numjobs` option is
1359 specified). This option is useful if there are several jobs which are
1360 intended to operate on a file in parallel disjoint segments, with even
1361 spacing between the starting points. Percentages can be used for this option.
1362 If a percentage is given, the generated offset will be aligned to the minimum
1363 ``blocksize`` or to the value of ``offset_align`` if provided. In ZBD mode, value can
1364 also be set as number of zones using 'z'.
1366 .. option:: number_ios=int
1368 Fio will normally perform I/Os until it has exhausted the size of the region
1369 set by :option:`size`, or if it exhaust the allocated time (or hits an error
1370 condition). With this setting, the range/size can be set independently of
1371 the number of I/Os to perform. When fio reaches this number, it will exit
1372 normally and report status. Note that this does not extend the amount of I/O
1373 that will be done, it will only stop fio if this condition is met before
1374 other end-of-job criteria.
1376 .. option:: fsync=int
1378 If writing to a file, issue an :manpage:`fsync(2)` (or its equivalent) of
1379 the dirty data for every number of blocks given. For example, if you give 32
1380 as a parameter, fio will sync the file after every 32 writes issued. If fio is
1381 using non-buffered I/O, we may not sync the file. The exception is the sg
1382 I/O engine, which synchronizes the disk cache anyway. Defaults to 0, which
1383 means fio does not periodically issue and wait for a sync to complete. Also
1384 see :option:`end_fsync` and :option:`fsync_on_close`.
1386 .. option:: fdatasync=int
1388 Like :option:`fsync` but uses :manpage:`fdatasync(2)` to only sync data and
1389 not metadata blocks. In Windows, DragonFlyBSD or OSX there is no
1390 :manpage:`fdatasync(2)` so this falls back to using :manpage:`fsync(2)`.
1391 Defaults to 0, which means fio does not periodically issue and wait for a
1392 data-only sync to complete.
1394 .. option:: write_barrier=int
1396 Make every `N-th` write a barrier write.
1398 .. option:: sync_file_range=str:int
1400 Use :manpage:`sync_file_range(2)` for every `int` number of write
1401 operations. Fio will track range of writes that have happened since the last
1402 :manpage:`sync_file_range(2)` call. `str` can currently be one or more of:
1405 SYNC_FILE_RANGE_WAIT_BEFORE
1407 SYNC_FILE_RANGE_WRITE
1409 SYNC_FILE_RANGE_WAIT_AFTER
1411 So if you do ``sync_file_range=wait_before,write:8``, fio would use
1412 ``SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE`` for every 8
1413 writes. Also see the :manpage:`sync_file_range(2)` man page. This option is
1416 .. option:: overwrite=bool
1418 If true, writes to a file will always overwrite existing data. If the file
1419 doesn't already exist, it will be created before the write phase begins. If
1420 the file exists and is large enough for the specified write phase, nothing
1421 will be done. Default: false.
1423 .. option:: end_fsync=bool
1425 If true, :manpage:`fsync(2)` file contents when a write stage has completed.
1428 .. option:: fsync_on_close=bool
1430 If true, fio will :manpage:`fsync(2)` a dirty file on close. This differs
1431 from :option:`end_fsync` in that it will happen on every file close, not
1432 just at the end of the job. Default: false.
1434 .. option:: rwmixread=int
1436 Percentage of a mixed workload that should be reads. Default: 50.
1438 .. option:: rwmixwrite=int
1440 Percentage of a mixed workload that should be writes. If both
1441 :option:`rwmixread` and :option:`rwmixwrite` is given and the values do not
1442 add up to 100%, the latter of the two will be used to override the
1443 first. This may interfere with a given rate setting, if fio is asked to
1444 limit reads or writes to a certain rate. If that is the case, then the
1445 distribution may be skewed. Default: 50.
1447 .. option:: random_distribution=str:float[:float][,str:float][,str:float]
1449 By default, fio will use a completely uniform random distribution when asked
1450 to perform random I/O. Sometimes it is useful to skew the distribution in
1451 specific ways, ensuring that some parts of the data is more hot than others.
1452 fio includes the following distribution models:
1455 Uniform random distribution
1464 Normal (Gaussian) distribution
1467 Zoned random distribution
1470 Zone absolute random distribution
1472 When using a **zipf** or **pareto** distribution, an input value is also
1473 needed to define the access pattern. For **zipf**, this is the `Zipf
1474 theta`. For **pareto**, it's the `Pareto power`. Fio includes a test
1475 program, :command:`fio-genzipf`, that can be used visualize what the given input
1476 values will yield in terms of hit rates. If you wanted to use **zipf** with
1477 a `theta` of 1.2, you would use ``random_distribution=zipf:1.2`` as the
1478 option. If a non-uniform model is used, fio will disable use of the random
1479 map. For the **normal** distribution, a normal (Gaussian) deviation is
1480 supplied as a value between 0 and 100.
1482 The second, optional float is allowed for **pareto**, **zipf** and **normal** distributions.
1483 It allows one to set base of distribution in non-default place, giving more control
1484 over most probable outcome. This value is in range [0-1] which maps linearly to
1485 range of possible random values.
1486 Defaults are: random for **pareto** and **zipf**, and 0.5 for **normal**.
1487 If you wanted to use **zipf** with a `theta` of 1.2 centered on 1/4 of allowed value range,
1488 you would use ``random_distribution=zipf:1.2:0.25``.
1490 For a **zoned** distribution, fio supports specifying percentages of I/O
1491 access that should fall within what range of the file or device. For
1492 example, given a criteria of:
1494 * 60% of accesses should be to the first 10%
1495 * 30% of accesses should be to the next 20%
1496 * 8% of accesses should be to the next 30%
1497 * 2% of accesses should be to the next 40%
1499 we can define that through zoning of the random accesses. For the above
1500 example, the user would do::
1502 random_distribution=zoned:60/10:30/20:8/30:2/40
1504 A **zoned_abs** distribution works exactly like the **zoned**, except
1505 that it takes absolute sizes. For example, let's say you wanted to
1506 define access according to the following criteria:
1508 * 60% of accesses should be to the first 20G
1509 * 30% of accesses should be to the next 100G
1510 * 10% of accesses should be to the next 500G
1512 we can define an absolute zoning distribution with:
1514 random_distribution=zoned_abs=60/20G:30/100G:10/500g
1516 For both **zoned** and **zoned_abs**, fio supports defining up to
1519 Similarly to how :option:`bssplit` works for setting ranges and
1520 percentages of block sizes. Like :option:`bssplit`, it's possible to
1521 specify separate zones for reads, writes, and trims. If just one set
1522 is given, it'll apply to all of them. This goes for both **zoned**
1523 **zoned_abs** distributions.
1525 .. option:: percentage_random=int[,int][,int]
1527 For a random workload, set how big a percentage should be random. This
1528 defaults to 100%, in which case the workload is fully random. It can be set
1529 from anywhere from 0 to 100. Setting it to 0 would make the workload fully
1530 sequential. Any setting in between will result in a random mix of sequential
1531 and random I/O, at the given percentages. Comma-separated values may be
1532 specified for reads, writes, and trims as described in :option:`blocksize`.
1534 .. option:: norandommap
1536 Normally fio will cover every block of the file when doing random I/O. If
1537 this option is given, fio will just get a new random offset without looking
1538 at past I/O history. This means that some blocks may not be read or written,
1539 and that some blocks may be read/written more than once. If this option is
1540 used with :option:`verify` and multiple blocksizes (via :option:`bsrange`),
1541 only intact blocks are verified, i.e., partially-overwritten blocks are
1542 ignored. With an async I/O engine and an I/O depth > 1, it is possible for
1543 the same block to be overwritten, which can cause verification errors. Either
1544 do not use norandommap in this case, or also use the lfsr random generator.
1546 .. option:: softrandommap=bool
1548 See :option:`norandommap`. If fio runs with the random block map enabled and
1549 it fails to allocate the map, if this option is set it will continue without
1550 a random block map. As coverage will not be as complete as with random maps,
1551 this option is disabled by default.
1553 .. option:: random_generator=str
1555 Fio supports the following engines for generating I/O offsets for random I/O:
1558 Strong 2^88 cycle random number generator.
1560 Linear feedback shift register generator.
1562 Strong 64-bit 2^258 cycle random number generator.
1564 **tausworthe** is a strong random number generator, but it requires tracking
1565 on the side if we want to ensure that blocks are only read or written
1566 once. **lfsr** guarantees that we never generate the same offset twice, and
1567 it's also less computationally expensive. It's not a true random generator,
1568 however, though for I/O purposes it's typically good enough. **lfsr** only
1569 works with single block sizes, not with workloads that use multiple block
1570 sizes. If used with such a workload, fio may read or write some blocks
1571 multiple times. The default value is **tausworthe**, unless the required
1572 space exceeds 2^32 blocks. If it does, then **tausworthe64** is
1573 selected automatically.
1579 .. option:: blocksize=int[,int][,int], bs=int[,int][,int]
1581 The block size in bytes used for I/O units. Default: 4096. A single value
1582 applies to reads, writes, and trims. Comma-separated values may be
1583 specified for reads, writes, and trims. A value not terminated in a comma
1584 applies to subsequent types.
1589 means 256k for reads, writes and trims.
1592 means 8k for reads, 32k for writes and trims.
1595 means 8k for reads, 32k for writes, and default for trims.
1598 means default for reads, 8k for writes and trims.
1601 means default for reads, 8k for writes, and default for trims.
1603 .. option:: blocksize_range=irange[,irange][,irange], bsrange=irange[,irange][,irange]
1605 A range of block sizes in bytes for I/O units. The issued I/O unit will
1606 always be a multiple of the minimum size, unless
1607 :option:`blocksize_unaligned` is set.
1609 Comma-separated ranges may be specified for reads, writes, and trims as
1610 described in :option:`blocksize`.
1612 Example: ``bsrange=1k-4k,2k-8k``.
1614 .. option:: bssplit=str[,str][,str]
1616 Sometimes you want even finer grained control of the block sizes
1617 issued, not just an even split between them. This option allows you to
1618 weight various block sizes, so that you are able to define a specific
1619 amount of block sizes issued. The format for this option is::
1621 bssplit=blocksize/percentage:blocksize/percentage
1623 for as many block sizes as needed. So if you want to define a workload
1624 that has 50% 64k blocks, 10% 4k blocks, and 40% 32k blocks, you would
1627 bssplit=4k/10:64k/50:32k/40
1629 Ordering does not matter. If the percentage is left blank, fio will
1630 fill in the remaining values evenly. So a bssplit option like this one::
1632 bssplit=4k/50:1k/:32k/
1634 would have 50% 4k ios, and 25% 1k and 32k ios. The percentages always
1635 add up to 100, if bssplit is given a range that adds up to more, it
1638 Comma-separated values may be specified for reads, writes, and trims as
1639 described in :option:`blocksize`.
1641 If you want a workload that has 50% 2k reads and 50% 4k reads, while
1642 having 90% 4k writes and 10% 8k writes, you would specify::
1644 bssplit=2k/50:4k/50,4k/90:8k/10
1646 Fio supports defining up to 64 different weights for each data
1649 .. option:: blocksize_unaligned, bs_unaligned
1651 If set, fio will issue I/O units with any size within
1652 :option:`blocksize_range`, not just multiples of the minimum size. This
1653 typically won't work with direct I/O, as that normally requires sector
1656 .. option:: bs_is_seq_rand=bool
1658 If this option is set, fio will use the normal read,write blocksize settings
1659 as sequential,random blocksize settings instead. Any random read or write
1660 will use the WRITE blocksize settings, and any sequential read or write will
1661 use the READ blocksize settings.
1663 .. option:: blockalign=int[,int][,int], ba=int[,int][,int]
1665 Boundary to which fio will align random I/O units. Default:
1666 :option:`blocksize`. Minimum alignment is typically 512b for using direct
1667 I/O, though it usually depends on the hardware block size. This option is
1668 mutually exclusive with using a random map for files, so it will turn off
1669 that option. Comma-separated values may be specified for reads, writes, and
1670 trims as described in :option:`blocksize`.
1676 .. option:: zero_buffers
1678 Initialize buffers with all zeros. Default: fill buffers with random data.
1680 .. option:: refill_buffers
1682 If this option is given, fio will refill the I/O buffers on every
1683 submit. Only makes sense if :option:`zero_buffers` isn't specified,
1684 naturally. Defaults to being unset i.e., the buffer is only filled at
1685 init time and the data in it is reused when possible but if any of
1686 :option:`verify`, :option:`buffer_compress_percentage` or
1687 :option:`dedupe_percentage` are enabled then `refill_buffers` is also
1688 automatically enabled.
1690 .. option:: scramble_buffers=bool
1692 If :option:`refill_buffers` is too costly and the target is using data
1693 deduplication, then setting this option will slightly modify the I/O buffer
1694 contents to defeat normal de-dupe attempts. This is not enough to defeat
1695 more clever block compression attempts, but it will stop naive dedupe of
1696 blocks. Default: true.
1698 .. option:: buffer_compress_percentage=int
1700 If this is set, then fio will attempt to provide I/O buffer content
1701 (on WRITEs) that compresses to the specified level. Fio does this by
1702 providing a mix of random data followed by fixed pattern data. The
1703 fixed pattern is either zeros, or the pattern specified by
1704 :option:`buffer_pattern`. If the `buffer_pattern` option is used, it
1705 might skew the compression ratio slightly. Setting
1706 `buffer_compress_percentage` to a value other than 100 will also
1707 enable :option:`refill_buffers` in order to reduce the likelihood that
1708 adjacent blocks are so similar that they over compress when seen
1709 together. See :option:`buffer_compress_chunk` for how to set a finer or
1710 coarser granularity for the random/fixed data region. Defaults to unset
1711 i.e., buffer data will not adhere to any compression level.
1713 .. option:: buffer_compress_chunk=int
1715 This setting allows fio to manage how big the random/fixed data region
1716 is when using :option:`buffer_compress_percentage`. When
1717 `buffer_compress_chunk` is set to some non-zero value smaller than the
1718 block size, fio can repeat the random/fixed region throughout the I/O
1719 buffer at the specified interval (which particularly useful when
1720 bigger block sizes are used for a job). When set to 0, fio will use a
1721 chunk size that matches the block size resulting in a single
1722 random/fixed region within the I/O buffer. Defaults to 512. When the
1723 unit is omitted, the value is interpreted in bytes.
1725 .. option:: buffer_pattern=str
1727 If set, fio will fill the I/O buffers with this pattern or with the contents
1728 of a file. If not set, the contents of I/O buffers are defined by the other
1729 options related to buffer contents. The setting can be any pattern of bytes,
1730 and can be prefixed with 0x for hex values. It may also be a string, where
1731 the string must then be wrapped with ``""``. Or it may also be a filename,
1732 where the filename must be wrapped with ``''`` in which case the file is
1733 opened and read. Note that not all the file contents will be read if that
1734 would cause the buffers to overflow. So, for example::
1736 buffer_pattern='filename'
1740 buffer_pattern="abcd"
1748 buffer_pattern=0xdeadface
1750 Also you can combine everything together in any order::
1752 buffer_pattern=0xdeadface"abcd"-12'filename'
1754 .. option:: dedupe_percentage=int
1756 If set, fio will generate this percentage of identical buffers when
1757 writing. These buffers will be naturally dedupable. The contents of the
1758 buffers depend on what other buffer compression settings have been set. It's
1759 possible to have the individual buffers either fully compressible, or not at
1760 all -- this option only controls the distribution of unique buffers. Setting
1761 this option will also enable :option:`refill_buffers` to prevent every buffer
1764 .. option:: dedupe_mode=str
1766 If ``dedupe_percentage=<int>`` is given, then this option controls how fio
1767 generates the dedupe buffers.
1770 Generate dedupe buffers by repeating previous writes
1772 Generate dedupe buffers from working set
1774 ``repeat`` is the default option for fio. Dedupe buffers are generated
1775 by repeating previous unique write.
1777 ``working_set`` is a more realistic workload.
1778 With ``working_set``, ``dedupe_working_set_percentage=<int>`` should be provided.
1779 Given that, fio will use the initial unique write buffers as its working set.
1780 Upon deciding to dedupe, fio will randomly choose a buffer from the working set.
1781 Note that by using ``working_set`` the dedupe percentage will converge
1782 to the desired over time while ``repeat`` maintains the desired percentage
1785 .. option:: dedupe_working_set_percentage=int
1787 If ``dedupe_mode=<str>`` is set to ``working_set``, then this controls
1788 the percentage of size of the file or device used as the buffers
1789 fio will choose to generate the dedupe buffers from
1791 Note that size needs to be explicitly provided and only 1 file per
1794 .. option:: dedupe_global=bool
1796 This controls whether the deduplication buffers will be shared amongst
1797 all jobs that have this option set. The buffers are spread evenly between
1800 .. option:: invalidate=bool
1802 Invalidate the buffer/page cache parts of the files to be used prior to
1803 starting I/O if the platform and file type support it. Defaults to true.
1804 This will be ignored if :option:`pre_read` is also specified for the
1807 .. option:: sync=str
1809 Whether, and what type, of synchronous I/O to use for writes. The allowed
1813 Do not use synchronous IO, the default.
1819 Use synchronous file IO. For the majority of I/O engines,
1820 this means using O_SYNC.
1826 Use synchronous data IO. For the majority of I/O engines,
1827 this means using O_DSYNC.
1830 .. option:: iomem=str, mem=str
1832 Fio can use various types of memory as the I/O unit buffer. The allowed
1836 Use memory from :manpage:`malloc(3)` as the buffers. Default memory
1840 Use shared memory as the buffers. Allocated through
1841 :manpage:`shmget(2)`.
1844 Same as shm, but use huge pages as backing.
1847 Use :manpage:`mmap(2)` to allocate buffers. May either be anonymous memory, or can
1848 be file backed if a filename is given after the option. The format
1849 is `mem=mmap:/path/to/file`.
1852 Use a memory mapped huge file as the buffer backing. Append filename
1853 after mmaphuge, ala `mem=mmaphuge:/hugetlbfs/file`.
1856 Same as mmap, but use a MMAP_SHARED mapping.
1859 Use GPU memory as the buffers for GPUDirect RDMA benchmark.
1860 The :option:`ioengine` must be `rdma`.
1862 The area allocated is a function of the maximum allowed bs size for the job,
1863 multiplied by the I/O depth given. Note that for **shmhuge** and
1864 **mmaphuge** to work, the system must have free huge pages allocated. This
1865 can normally be checked and set by reading/writing
1866 :file:`/proc/sys/vm/nr_hugepages` on a Linux system. Fio assumes a huge page
1867 is 2 or 4MiB in size depending on the platform. So to calculate the
1868 number of huge pages you need for a given job file, add up the I/O
1869 depth of all jobs (normally one unless :option:`iodepth` is used) and
1870 multiply by the maximum bs set. Then divide that number by the huge
1871 page size. You can see the size of the huge pages in
1872 :file:`/proc/meminfo`. If no huge pages are allocated by having a
1873 non-zero number in `nr_hugepages`, using **mmaphuge** or **shmhuge**
1874 will fail. Also see :option:`hugepage-size`.
1876 **mmaphuge** also needs to have hugetlbfs mounted and the file location
1877 should point there. So if it's mounted in :file:`/huge`, you would use
1878 `mem=mmaphuge:/huge/somefile`.
1880 .. option:: iomem_align=int, mem_align=int
1882 This indicates the memory alignment of the I/O memory buffers. Note that
1883 the given alignment is applied to the first I/O unit buffer, if using
1884 :option:`iodepth` the alignment of the following buffers are given by the
1885 :option:`bs` used. In other words, if using a :option:`bs` that is a
1886 multiple of the page sized in the system, all buffers will be aligned to
1887 this value. If using a :option:`bs` that is not page aligned, the alignment
1888 of subsequent I/O memory buffers is the sum of the :option:`iomem_align` and
1891 .. option:: hugepage-size=int
1893 Defines the size of a huge page. Must at least be equal to the system
1894 setting, see :file:`/proc/meminfo` and
1895 :file:`/sys/kernel/mm/hugepages/`. Defaults to 2 or 4MiB depending on
1896 the platform. Should probably always be a multiple of megabytes, so
1897 using ``hugepage-size=Xm`` is the preferred way to set this to avoid
1898 setting a non-pow-2 bad value.
1900 .. option:: lockmem=int
1902 Pin the specified amount of memory with :manpage:`mlock(2)`. Can be used to
1903 simulate a smaller amount of memory. The amount specified is per worker.
1909 .. option:: size=int
1911 The total size of file I/O for each thread of this job. Fio will run until
1912 this many bytes has been transferred, unless runtime is altered by other means
1913 such as (1) :option:`runtime`, (2) :option:`io_size` (3) :option:`number_ios`,
1914 (4) gaps/holes while doing I/O's such as ``rw=read:16K``, or (5) sequential
1915 I/O reaching end of the file which is possible when :option:`percentage_random`
1917 Fio will divide this size between the available files determined by options
1918 such as :option:`nrfiles`, :option:`filename`, unless :option:`filesize` is
1919 specified by the job. If the result of division happens to be 0, the size is
1920 set to the physical size of the given files or devices if they exist.
1921 If this option is not specified, fio will use the full size of the given
1922 files or devices. If the files do not exist, size must be given. It is also
1923 possible to give size as a percentage between 1 and 100. If ``size=20%`` is
1924 given, fio will use 20% of the full size of the given files or devices.
1925 In ZBD mode, value can also be set as number of zones using 'z'.
1926 Can be combined with :option:`offset` to constrain the start and end range
1927 that I/O will be done within.
1929 .. option:: io_size=int, io_limit=int
1931 Normally fio operates within the region set by :option:`size`, which means
1932 that the :option:`size` option sets both the region and size of I/O to be
1933 performed. Sometimes that is not what you want. With this option, it is
1934 possible to define just the amount of I/O that fio should do. For instance,
1935 if :option:`size` is set to 20GiB and :option:`io_size` is set to 5GiB, fio
1936 will perform I/O within the first 20GiB but exit when 5GiB have been
1937 done. The opposite is also possible -- if :option:`size` is set to 20GiB,
1938 and :option:`io_size` is set to 40GiB, then fio will do 40GiB of I/O within
1939 the 0..20GiB region.
1941 .. option:: filesize=irange(int)
1943 Individual file sizes. May be a range, in which case fio will select sizes for
1944 files at random within the given range. If not given, each created file is the
1945 same size. This option overrides :option:`size` in terms of file size, i.e. if
1946 :option:`filesize` is specified then :option:`size` becomes merely the default
1947 for :option:`io_size` and has no effect at all if :option:`io_size` is set
1950 .. option:: file_append=bool
1952 Perform I/O after the end of the file. Normally fio will operate within the
1953 size of a file. If this option is set, then fio will append to the file
1954 instead. This has identical behavior to setting :option:`offset` to the size
1955 of a file. This option is ignored on non-regular files.
1957 .. option:: fill_device=bool, fill_fs=bool
1959 Sets size to something really large and waits for ENOSPC (no space left on
1960 device) or EDQUOT (disk quota exceeded)
1961 as the terminating condition. Only makes sense with sequential
1962 write. For a read workload, the mount point will be filled first then I/O
1963 started on the result. This option doesn't make sense if operating on a raw
1964 device node, since the size of that is already known by the file system.
1965 Additionally, writing beyond end-of-device will not return ENOSPC there.
1971 .. option:: ioengine=str
1973 Defines how the job issues I/O to the file. The following types are defined:
1976 Basic :manpage:`read(2)` or :manpage:`write(2)`
1977 I/O. :manpage:`lseek(2)` is used to position the I/O location.
1978 See :option:`fsync` and :option:`fdatasync` for syncing write I/Os.
1981 Basic :manpage:`pread(2)` or :manpage:`pwrite(2)` I/O. Default on
1982 all supported operating systems except for Windows.
1985 Basic :manpage:`readv(2)` or :manpage:`writev(2)` I/O. Will emulate
1986 queuing by coalescing adjacent I/Os into a single submission.
1989 Basic :manpage:`preadv(2)` or :manpage:`pwritev(2)` I/O.
1992 Basic :manpage:`preadv2(2)` or :manpage:`pwritev2(2)` I/O.
1995 Fast Linux native asynchronous I/O. Supports async IO
1996 for both direct and buffered IO.
1997 This engine defines engine specific options.
2000 Fast Linux native asynchronous I/O for pass through commands.
2001 This engine defines engine specific options.
2004 Linux native asynchronous I/O. Note that Linux may only support
2005 queued behavior with non-buffered I/O (set ``direct=1`` or
2007 This engine defines engine specific options.
2010 POSIX asynchronous I/O using :manpage:`aio_read(3)` and
2011 :manpage:`aio_write(3)`.
2014 Solaris native asynchronous I/O.
2017 Windows native asynchronous I/O. Default on Windows.
2020 File is memory mapped with :manpage:`mmap(2)` and data copied
2021 to/from using :manpage:`memcpy(3)`.
2024 :manpage:`splice(2)` is used to transfer the data and
2025 :manpage:`vmsplice(2)` to transfer data from user space to the
2029 SCSI generic sg v3 I/O. May either be synchronous using the SG_IO
2030 ioctl, or if the target is an sg character device we use
2031 :manpage:`read(2)` and :manpage:`write(2)` for asynchronous
2032 I/O. Requires :option:`filename` option to specify either block or
2033 character devices. This engine supports trim operations.
2034 The sg engine includes engine specific options.
2037 Read, write, trim and ZBC/ZAC operations to a zoned
2038 block device using libzbc library. The target can be
2039 either an SG character device or a block device file.
2042 Doesn't transfer any data, just pretends to. This is mainly used to
2043 exercise fio itself and for debugging/testing purposes.
2046 Transfer over the network to given ``host:port``. Depending on the
2047 :option:`protocol` used, the :option:`hostname`, :option:`port`,
2048 :option:`listen` and :option:`filename` options are used to specify
2049 what sort of connection to make, while the :option:`protocol` option
2050 determines which protocol will be used. This engine defines engine
2054 Like **net**, but uses :manpage:`splice(2)` and
2055 :manpage:`vmsplice(2)` to map data and send/receive.
2056 This engine defines engine specific options.
2059 Doesn't transfer any data, but burns CPU cycles according to the
2060 :option:`cpuload`, :option:`cpuchunks` and :option:`cpumode` options.
2061 Setting :option:`cpuload`\=85 will cause that job to do nothing but burn 85%
2062 of the CPU. In case of SMP machines, use :option:`numjobs`\=<nr_of_cpu>
2063 to get desired CPU usage, as the cpuload only loads a
2064 single CPU at the desired rate. A job never finishes unless there is
2065 at least one non-cpuio job.
2066 Setting :option:`cpumode`\=qsort replace the default noop instructions loop
2067 by a qsort algorithm to consume more energy.
2070 The RDMA I/O engine supports both RDMA memory semantics
2071 (RDMA_WRITE/RDMA_READ) and channel semantics (Send/Recv) for the
2072 InfiniBand, RoCE and iWARP protocols. This engine defines engine
2076 I/O engine that does regular fallocate to simulate data transfer as
2080 does fallocate(,mode = FALLOC_FL_KEEP_SIZE,).
2083 does fallocate(,mode = 0).
2086 does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE).
2089 I/O engine that sends :manpage:`ftruncate(2)` operations in response
2090 to write (DDIR_WRITE) events. Each ftruncate issued sets the file's
2091 size to the current block offset. :option:`blocksize` is ignored.
2094 I/O engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate
2095 defragment activity in request to DDIR_WRITE event.
2098 I/O engine supporting direct access to Ceph Reliable Autonomic
2099 Distributed Object Store (RADOS) via librados. This ioengine
2100 defines engine specific options.
2103 I/O engine supporting direct access to Ceph Rados Block Devices
2104 (RBD) via librbd without the need to use the kernel rbd driver. This
2105 ioengine defines engine specific options.
2108 I/O engine supporting GET/PUT requests over HTTP(S) with libcurl to
2109 a WebDAV or S3 endpoint. This ioengine defines engine specific options.
2111 This engine only supports direct IO of iodepth=1; you need to scale this
2112 via numjobs. blocksize defines the size of the objects to be created.
2114 TRIM is translated to object deletion.
2117 Using GlusterFS libgfapi sync interface to direct access to
2118 GlusterFS volumes without having to go through FUSE. This ioengine
2119 defines engine specific options.
2122 Using GlusterFS libgfapi async interface to direct access to
2123 GlusterFS volumes without having to go through FUSE. This ioengine
2124 defines engine specific options.
2127 Read and write through Hadoop (HDFS). The :option:`filename` option
2128 is used to specify host,port of the hdfs name-node to connect. This
2129 engine interprets offsets a little differently. In HDFS, files once
2130 created cannot be modified so random writes are not possible. To
2131 imitate this the libhdfs engine expects a bunch of small files to be
2132 created over HDFS and will randomly pick a file from them
2133 based on the offset generated by fio backend (see the example
2134 job file to create such files, use ``rw=write`` option). Please
2135 note, it may be necessary to set environment variables to work
2136 with HDFS/libhdfs properly. Each job uses its own connection to
2140 Read, write and erase an MTD character device (e.g.,
2141 :file:`/dev/mtd0`). Discards are treated as erases. Depending on the
2142 underlying device type, the I/O may have to go in a certain pattern,
2143 e.g., on NAND, writing sequentially to erase blocks and discarding
2144 before overwriting. The `trimwrite` mode works well for this
2148 Read and write using filesystem DAX to a file on a filesystem
2149 mounted with DAX on a persistent memory device through the PMDK
2153 Read and write using device DAX to a persistent memory device (e.g.,
2154 /dev/dax0.0) through the PMDK libpmem library.
2157 Prefix to specify loading an external I/O engine object file. Append
2158 the engine filename, e.g. ``ioengine=external:/tmp/foo.o`` to load
2159 ioengine :file:`foo.o` in :file:`/tmp`. The path can be either
2160 absolute or relative. See :file:`engines/skeleton_external.c` for
2161 details of writing an external I/O engine.
2164 Simply create the files and do no I/O to them. You still need to
2165 set `filesize` so that all the accounting still occurs, but no
2166 actual I/O will be done other than creating the file.
2169 Simply do stat() and do no I/O to the file. You need to set 'filesize'
2170 and 'nrfiles', so that files will be created.
2171 This engine is to measure file lookup and meta data access.
2174 Simply delete the files by unlink() and do no I/O to them. You need to set 'filesize'
2175 and 'nrfiles', so that the files will be created.
2176 This engine is to measure file delete.
2179 Read and write using mmap I/O to a file on a filesystem
2180 mounted with DAX on a persistent memory device through the PMDK
2184 Synchronous read and write using DDN's Infinite Memory Engine (IME).
2185 This engine is very basic and issues calls to IME whenever an IO is
2189 Synchronous read and write using DDN's Infinite Memory Engine (IME).
2190 This engine uses iovecs and will try to stack as much IOs as possible
2191 (if the IOs are "contiguous" and the IO depth is not exceeded)
2192 before issuing a call to IME.
2195 Asynchronous read and write using DDN's Infinite Memory Engine (IME).
2196 This engine will try to stack as much IOs as possible by creating
2197 requests for IME. FIO will then decide when to commit these requests.
2200 Read and write iscsi lun with libiscsi.
2203 Read and write a Network Block Device (NBD).
2206 I/O engine supporting libcufile synchronous access to nvidia-fs and a
2207 GPUDirect Storage-supported filesystem. This engine performs
2208 I/O without transferring buffers between user-space and the kernel,
2209 unless :option:`verify` is set or :option:`cuda_io` is `posix`.
2210 :option:`iomem` must not be `cudamalloc`. This ioengine defines
2211 engine specific options.
2214 I/O engine supporting asynchronous read and write operations to the
2215 DAOS File System (DFS) via libdfs.
2218 I/O engine supporting asynchronous read and write operations to
2219 NFS filesystems from userspace via libnfs. This is useful for
2220 achieving higher concurrency and thus throughput than is possible
2224 Execute 3rd party tools. Could be used to perform monitoring during jobs runtime.
2227 I/O engine using the xNVMe C API, for NVMe devices. The xnvme engine provides
2228 flexibility to access GNU/Linux Kernel NVMe driver via libaio, IOCTLs, io_uring,
2229 the SPDK NVMe driver, or your own custom NVMe driver. The xnvme engine includes
2230 engine specific options. (See https://xnvme.io).
2233 Use the libblkio library
2234 (https://gitlab.com/libblkio/libblkio). The specific
2235 *driver* to use must be set using
2236 :option:`libblkio_driver`. If
2237 :option:`mem`/:option:`iomem` is not specified, memory
2238 allocation is delegated to libblkio (and so is
2239 guaranteed to work with the selected *driver*). One
2240 libblkio instance is used per process, so all jobs
2241 setting option :option:`thread` will share a single
2242 instance (with one queue per thread) and must specify
2243 compatible options. Note that some drivers don't allow
2244 several instances to access the same device or file
2245 simultaneously, but allow it for threads.
2247 I/O engine specific parameters
2248 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2250 In addition, there are some parameters which are only valid when a specific
2251 :option:`ioengine` is in use. These are used identically to normal parameters,
2252 with the caveat that when used on the command line, they must come after the
2253 :option:`ioengine` that defines them is selected.
2255 .. option:: cmdprio_percentage=int[,int] : [io_uring] [libaio]
2257 Set the percentage of I/O that will be issued with the highest priority.
2258 Default: 0. A single value applies to reads and writes. Comma-separated
2259 values may be specified for reads and writes. For this option to be
2260 effective, NCQ priority must be supported and enabled, and the :option:`direct`
2261 option must be set. fio must also be run as the root user. Unlike
2262 slat/clat/lat stats, which can be tracked and reported independently, per
2263 priority stats only track and report a single type of latency. By default,
2264 completion latency (clat) will be reported, if :option:`lat_percentiles` is
2265 set, total latency (lat) will be reported.
2267 .. option:: cmdprio_class=int[,int] : [io_uring] [libaio]
2269 Set the I/O priority class to use for I/Os that must be issued with
2270 a priority when :option:`cmdprio_percentage` or
2271 :option:`cmdprio_bssplit` is set. If not specified when
2272 :option:`cmdprio_percentage` or :option:`cmdprio_bssplit` is set,
2273 this defaults to the highest priority class. A single value applies
2274 to reads and writes. Comma-separated values may be specified for
2275 reads and writes. See :manpage:`ionice(1)`. See also the
2276 :option:`prioclass` option.
2278 .. option:: cmdprio=int[,int] : [io_uring] [libaio]
2280 Set the I/O priority value to use for I/Os that must be issued with
2281 a priority when :option:`cmdprio_percentage` or
2282 :option:`cmdprio_bssplit` is set. If not specified when
2283 :option:`cmdprio_percentage` or :option:`cmdprio_bssplit` is set,
2285 Linux limits us to a positive value between 0 and 7, with 0 being the
2286 highest. A single value applies to reads and writes. Comma-separated
2287 values may be specified for reads and writes. See :manpage:`ionice(1)`.
2288 Refer to an appropriate manpage for other operating systems since
2289 meaning of priority may differ. See also the :option:`prio` option.
2291 .. option:: cmdprio_bssplit=str[,str] : [io_uring] [libaio]
2293 To get a finer control over I/O priority, this option allows
2294 specifying the percentage of IOs that must have a priority set
2295 depending on the block size of the IO. This option is useful only
2296 when used together with the :option:`bssplit` option, that is,
2297 multiple different block sizes are used for reads and writes.
2299 The first accepted format for this option is the same as the format of
2300 the :option:`bssplit` option:
2302 cmdprio_bssplit=blocksize/percentage:blocksize/percentage
2304 In this case, each entry will use the priority class and priority
2305 level defined by the options :option:`cmdprio_class` and
2306 :option:`cmdprio` respectively.
2308 The second accepted format for this option is:
2310 cmdprio_bssplit=blocksize/percentage/class/level:blocksize/percentage/class/level
2312 In this case, the priority class and priority level is defined inside
2313 each entry. In comparison with the first accepted format, the second
2314 accepted format does not restrict all entries to have the same priority
2315 class and priority level.
2317 For both formats, only the read and write data directions are supported,
2318 values for trim IOs are ignored. This option is mutually exclusive with
2319 the :option:`cmdprio_percentage` option.
2321 .. option:: fixedbufs : [io_uring] [io_uring_cmd]
2323 If fio is asked to do direct IO, then Linux will map pages for each
2324 IO call, and release them when IO is done. If this option is set, the
2325 pages are pre-mapped before IO is started. This eliminates the need to
2326 map and release for each IO. This is more efficient, and reduces the
2329 .. option:: nonvectored=int : [io_uring] [io_uring_cmd]
2331 With this option, fio will use non-vectored read/write commands, where
2332 address must contain the address directly. Default is -1.
2334 .. option:: force_async=int : [io_uring] [io_uring_cmd]
2336 Normal operation for io_uring is to try and issue an sqe as
2337 non-blocking first, and if that fails, execute it in an async manner.
2338 With this option set to N, then every N request fio will ask sqe to
2339 be issued in an async manner. Default is 0.
2341 .. option:: registerfiles : [io_uring] [io_uring_cmd]
2343 With this option, fio registers the set of files being used with the
2344 kernel. This avoids the overhead of managing file counts in the kernel,
2345 making the submission and completion part more lightweight. Required
2346 for the below :option:`sqthread_poll` option.
2348 .. option:: sqthread_poll : [io_uring] [io_uring_cmd] [xnvme]
2350 Normally fio will submit IO by issuing a system call to notify the
2351 kernel of available items in the SQ ring. If this option is set, the
2352 act of submitting IO will be done by a polling thread in the kernel.
2353 This frees up cycles for fio, at the cost of using more CPU in the
2354 system. As submission is just the time it takes to fill in the sqe
2355 entries and any syscall required to wake up the idle kernel thread,
2356 fio will not report submission latencies.
2358 .. option:: sqthread_poll_cpu=int : [io_uring] [io_uring_cmd]
2360 When :option:`sqthread_poll` is set, this option provides a way to
2361 define which CPU should be used for the polling thread.
2363 .. option:: cmd_type=str : [io_uring_cmd]
2365 Specifies the type of uring passthrough command to be used. Supported
2366 value is nvme. Default is nvme.
2370 [io_uring] [io_uring_cmd] [xnvme]
2372 If this option is set, fio will attempt to use polled IO completions.
2373 Normal IO completions generate interrupts to signal the completion of
2374 IO, polled completions do not. Hence they are require active reaping
2375 by the application. The benefits are more efficient IO for high IOPS
2376 scenarios, and lower latencies for low queue depth IO.
2380 Use poll queues. This is incompatible with
2381 :option:`libblkio_wait_mode=eventfd <libblkio_wait_mode>` and
2382 :option:`libblkio_force_enable_completion_eventfd`.
2386 Set RWF_HIPRI on I/O, indicating to the kernel that it's of higher priority
2391 If this option is set, fio will attempt to use polled IO completions.
2392 This will have a similar effect as (io_uring)hipri. Only SCSI READ and
2393 WRITE commands will have the SGV4_FLAG_HIPRI set (not UNMAP (trim) nor
2394 VERIFY). Older versions of the Linux sg driver that do not support
2395 hipri will simply ignore this flag and do normal IO. The Linux SCSI
2396 Low Level Driver (LLD) that "owns" the device also needs to support
2397 hipri (also known as iopoll and mq_poll). The MegaRAID driver is an
2398 example of a SCSI LLD. Default: clear (0) which does normal
2399 (interrupted based) IO.
2401 .. option:: userspace_reap : [libaio]
2403 Normally, with the libaio engine in use, fio will use the
2404 :manpage:`io_getevents(2)` system call to reap newly returned events. With
2405 this flag turned on, the AIO ring will be read directly from user-space to
2406 reap events. The reaping mode is only enabled when polling for a minimum of
2407 0 events (e.g. when :option:`iodepth_batch_complete` `=0`).
2409 .. option:: hipri_percentage : [pvsync2]
2411 When hipri is set this determines the probability of a pvsync2 I/O being high
2412 priority. The default is 100%.
2414 .. option:: nowait=bool : [pvsync2] [libaio] [io_uring] [io_uring_cmd]
2416 By default if a request cannot be executed immediately (e.g. resource starvation,
2417 waiting on locks) it is queued and the initiating process will be blocked until
2418 the required resource becomes free.
2420 This option sets the RWF_NOWAIT flag (supported from the 4.14 Linux kernel) and
2421 the call will return instantly with EAGAIN or a partial result rather than waiting.
2423 It is useful to also use ignore_error=EAGAIN when using this option.
2425 Note: glibc 2.27, 2.28 have a bug in syscall wrappers preadv2, pwritev2.
2426 They return EOPNOTSUP instead of EAGAIN.
2428 For cached I/O, using this option usually means a request operates only with
2429 cached data. Currently the RWF_NOWAIT flag does not supported for cached write.
2431 For direct I/O, requests will only succeed if cache invalidation isn't required,
2432 file blocks are fully allocated and the disk request could be issued immediately.
2434 .. option:: cpuload=int : [cpuio]
2436 Attempt to use the specified percentage of CPU cycles. This is a mandatory
2437 option when using cpuio I/O engine.
2439 .. option:: cpuchunks=int : [cpuio]
2441 Split the load into cycles of the given time. In microseconds.
2443 .. option:: cpumode=str : [cpuio]
2445 Specify how to stress the CPU. It can take these two values:
2448 This is the default where the CPU executes noop instructions.
2450 Replace the default noop instructions loop with a qsort algorithm to
2451 consume more energy.
2453 .. option:: exit_on_io_done=bool : [cpuio]
2455 Detect when I/O threads are done, then exit.
2457 .. option:: namenode=str : [libhdfs]
2459 The hostname or IP address of a HDFS cluster namenode to contact.
2461 .. option:: port=int
2465 The listening port of the HFDS cluster namenode.
2469 The TCP or UDP port to bind to or connect to. If this is used with
2470 :option:`numjobs` to spawn multiple instances of the same job type, then
2471 this will be the starting port number since fio will use a range of
2476 The port to use for RDMA-CM communication. This should be the same value
2477 on the client and the server side.
2479 .. option:: hostname=str : [netsplice] [net] [rdma]
2481 The hostname or IP address to use for TCP, UDP or RDMA-CM based I/O. If the job
2482 is a TCP listener or UDP reader, the hostname is not used and must be omitted
2483 unless it is a valid UDP multicast address.
2485 .. option:: serverip=str : [librpma_*]
2487 The IP address to be used for RDMA-CM based I/O.
2489 .. option:: direct_write_to_pmem=bool : [librpma_*]
2491 Set to 1 only when Direct Write to PMem from the remote host is possible.
2492 Otherwise, set to 0.
2494 .. option:: busy_wait_polling=bool : [librpma_*_server]
2496 Set to 0 to wait for completion instead of busy-wait polling completion.
2499 .. option:: interface=str : [netsplice] [net]
2501 The IP address of the network interface used to send or receive UDP
2504 .. option:: ttl=int : [netsplice] [net]
2506 Time-to-live value for outgoing UDP multicast packets. Default: 1.
2508 .. option:: nodelay=bool : [netsplice] [net]
2510 Set TCP_NODELAY on TCP connections.
2512 .. option:: protocol=str, proto=str : [netsplice] [net]
2514 The network protocol to use. Accepted values are:
2517 Transmission control protocol.
2519 Transmission control protocol V6.
2521 User datagram protocol.
2523 User datagram protocol V6.
2527 When the protocol is TCP or UDP, the port must also be given, as well as the
2528 hostname if the job is a TCP listener or UDP reader. For unix sockets, the
2529 normal :option:`filename` option should be used and the port is invalid.
2531 .. option:: listen : [netsplice] [net]
2533 For TCP network connections, tell fio to listen for incoming connections
2534 rather than initiating an outgoing connection. The :option:`hostname` must
2535 be omitted if this option is used.
2537 .. option:: pingpong : [netsplice] [net]
2539 Normally a network writer will just continue writing data, and a network
2540 reader will just consume packages. If ``pingpong=1`` is set, a writer will
2541 send its normal payload to the reader, then wait for the reader to send the
2542 same payload back. This allows fio to measure network latencies. The
2543 submission and completion latencies then measure local time spent sending or
2544 receiving, and the completion latency measures how long it took for the
2545 other end to receive and send back. For UDP multicast traffic
2546 ``pingpong=1`` should only be set for a single reader when multiple readers
2547 are listening to the same address.
2549 .. option:: window_size : [netsplice] [net]
2551 Set the desired socket buffer size for the connection.
2553 .. option:: mss : [netsplice] [net]
2555 Set the TCP maximum segment size (TCP_MAXSEG).
2557 .. option:: donorname=str : [e4defrag]
2559 File will be used as a block donor (swap extents between files).
2561 .. option:: inplace=int : [e4defrag]
2563 Configure donor file blocks allocation strategy:
2566 Default. Preallocate donor's file on init.
2568 Allocate space immediately inside defragment event, and free right
2571 .. option:: clustername=str : [rbd,rados]
2573 Specifies the name of the Ceph cluster.
2575 .. option:: rbdname=str : [rbd]
2577 Specifies the name of the RBD.
2579 .. option:: clientname=str : [rbd,rados]
2581 Specifies the username (without the 'client.' prefix) used to access the
2582 Ceph cluster. If the *clustername* is specified, the *clientname* shall be
2583 the full *type.id* string. If no type. prefix is given, fio will add
2584 'client.' by default.
2586 .. option:: conf=str : [rados]
2588 Specifies the configuration path of ceph cluster, so conf file does not
2589 have to be /etc/ceph/ceph.conf.
2591 .. option:: busy_poll=bool : [rbd,rados]
2593 Poll store instead of waiting for completion. Usually this provides better
2594 throughput at cost of higher(up to 100%) CPU utilization.
2596 .. option:: touch_objects=bool : [rados]
2598 During initialization, touch (create if do not exist) all objects (files).
2599 Touching all objects affects ceph caches and likely impacts test results.
2602 .. option:: pool=str :
2606 Specifies the name of the Ceph pool containing RBD or RADOS data.
2610 Specify the label or UUID of the DAOS pool to connect to.
2612 .. option:: cont=str : [dfs]
2614 Specify the label or UUID of the DAOS container to open.
2616 .. option:: chunk_size=int
2620 Specify a different chunk size (in bytes) for the dfs file.
2621 Use DAOS container's chunk size by default.
2625 The size of the chunk to use for each file.
2627 .. option:: object_class=str : [dfs]
2629 Specify a different object class for the dfs file.
2630 Use DAOS container's object class by default.
2632 .. option:: skip_bad=bool : [mtd]
2634 Skip operations against known bad blocks.
2636 .. option:: hdfsdirectory : [libhdfs]
2638 libhdfs will create chunk in this HDFS directory.
2640 .. option:: verb=str : [rdma]
2642 The RDMA verb to use on this side of the RDMA ioengine connection. Valid
2643 values are write, read, send and recv. These correspond to the equivalent
2644 RDMA verbs (e.g. write = rdma_write etc.). Note that this only needs to be
2645 specified on the client side of the connection. See the examples folder.
2647 .. option:: bindname=str : [rdma]
2649 The name to use to bind the local RDMA-CM connection to a local RDMA device.
2650 This could be a hostname or an IPv4 or IPv6 address. On the server side this
2651 will be passed into the rdma_bind_addr() function and on the client site it
2652 will be used in the rdma_resolve_add() function. This can be useful when
2653 multiple paths exist between the client and the server or in certain loopback
2656 .. option:: stat_type=str : [filestat]
2658 Specify stat system call type to measure lookup/getattr performance.
2659 Default is **stat** for :manpage:`stat(2)`.
2661 .. option:: readfua=bool : [sg]
2663 With readfua option set to 1, read operations include
2664 the force unit access (fua) flag. Default is 0.
2666 .. option:: writefua=bool : [sg]
2668 With writefua option set to 1, write operations include
2669 the force unit access (fua) flag. Default is 0.
2671 .. option:: sg_write_mode=str : [sg]
2673 Specify the type of write commands to issue. This option can take three values:
2676 This is the default where write opcodes are issued as usual.
2677 **write_and_verify**
2678 Issue WRITE AND VERIFY commands. The BYTCHK bit is set to 0. This
2679 directs the device to carry out a medium verification with no data
2680 comparison. The writefua option is ignored with this selection.
2682 This option is deprecated. Use write_and_verify instead.
2684 Issue WRITE SAME commands. This transfers a single block to the device
2685 and writes this same block of data to a contiguous sequence of LBAs
2686 beginning at the specified offset. fio's block size parameter specifies
2687 the amount of data written with each command. However, the amount of data
2688 actually transferred to the device is equal to the device's block
2689 (sector) size. For a device with 512 byte sectors, blocksize=8k will
2690 write 16 sectors with each command. fio will still generate 8k of data
2691 for each command but only the first 512 bytes will be used and
2692 transferred to the device. The writefua option is ignored with this
2695 This option is deprecated. Use write_same instead.
2697 Issue WRITE SAME(16) commands as above but with the No Data Output
2698 Buffer (NDOB) bit set. No data will be transferred to the device with
2699 this bit set. Data written will be a pre-determined pattern such as
2702 Issue WRITE STREAM(16) commands. Use the **stream_id** option to specify
2703 the stream identifier.
2704 **verify_bytchk_00**
2705 Issue VERIFY commands with BYTCHK set to 00. This directs the
2706 device to carry out a medium verification with no data comparison.
2707 **verify_bytchk_01**
2708 Issue VERIFY commands with BYTCHK set to 01. This directs the device to
2709 compare the data on the device with the data transferred to the device.
2710 **verify_bytchk_11**
2711 Issue VERIFY commands with BYTCHK set to 11. This transfers a
2712 single block to the device and compares the contents of this block with the
2713 data on the device beginning at the specified offset. fio's block size
2714 parameter specifies the total amount of data compared with this command.
2715 However, only one block (sector) worth of data is transferred to the device.
2716 This is similar to the WRITE SAME command except that data is compared instead
2719 .. option:: stream_id=int : [sg]
2721 Set the stream identifier for WRITE STREAM commands. If this is set to 0 (which is not
2722 a valid stream identifier) fio will open a stream and then close it when done. Default
2725 .. option:: http_host=str : [http]
2727 Hostname to connect to. For S3, this could be the bucket hostname.
2728 Default is **localhost**
2730 .. option:: http_user=str : [http]
2732 Username for HTTP authentication.
2734 .. option:: http_pass=str : [http]
2736 Password for HTTP authentication.
2738 .. option:: https=str : [http]
2740 Enable HTTPS instead of http. *on* enables HTTPS; *insecure*
2741 will enable HTTPS, but disable SSL peer verification (use with
2742 caution!). Default is **off**
2744 .. option:: http_mode=str : [http]
2746 Which HTTP access mode to use: *webdav*, *swift*, or *s3*.
2747 Default is **webdav**
2749 .. option:: http_s3_region=str : [http]
2751 The S3 region/zone string.
2752 Default is **us-east-1**
2754 .. option:: http_s3_key=str : [http]
2758 .. option:: http_s3_keyid=str : [http]
2760 The S3 key/access id.
2762 .. option:: http_s3_sse_customer_key=str : [http]
2764 The encryption customer key in SSE server side.
2766 .. option:: http_s3_sse_customer_algorithm=str : [http]
2768 The encryption customer algorithm in SSE server side.
2769 Default is **AES256**
2771 .. option:: http_s3_storage_class=str : [http]
2773 Which storage class to access. User-customizable settings.
2774 Default is **STANDARD**
2776 .. option:: http_swift_auth_token=str : [http]
2778 The Swift auth token. See the example configuration file on how
2781 .. option:: http_verbose=int : [http]
2783 Enable verbose requests from libcurl. Useful for debugging. 1
2784 turns on verbose logging from libcurl, 2 additionally enables
2785 HTTP IO tracing. Default is **0**
2787 .. option:: uri=str : [nbd]
2789 Specify the NBD URI of the server to test. The string
2790 is a standard NBD URI
2791 (see https://github.com/NetworkBlockDevice/nbd/tree/master/doc).
2792 Example URIs: nbd://localhost:10809
2793 nbd+unix:///?socket=/tmp/socket
2794 nbds://tlshost/exportname
2796 .. option:: gpu_dev_ids=str : [libcufile]
2798 Specify the GPU IDs to use with CUDA. This is a colon-separated list of
2799 int. GPUs are assigned to workers roundrobin. Default is 0.
2801 .. option:: cuda_io=str : [libcufile]
2803 Specify the type of I/O to use with CUDA. Default is **cufile**.
2806 Use libcufile and nvidia-fs. This option performs I/O directly
2807 between a GPUDirect Storage filesystem and GPU buffers,
2808 avoiding use of a bounce buffer. If :option:`verify` is set,
2809 cudaMemcpy is used to copy verificaton data between RAM and GPU.
2810 Verification data is copied from RAM to GPU before a write
2811 and from GPU to RAM after a read. :option:`direct` must be 1.
2813 Use POSIX to perform I/O with a RAM buffer, and use cudaMemcpy
2814 to transfer data between RAM and the GPUs. Data is copied from
2815 GPU to RAM before a write and copied from RAM to GPU after a
2816 read. :option:`verify` does not affect use of cudaMemcpy.
2818 .. option:: nfs_url=str : [nfs]
2820 URL in libnfs format, eg nfs://<server|ipv4|ipv6>/path[?arg=val[&arg=val]*]
2821 Refer to the libnfs README for more details.
2823 .. option:: program=str : [exec]
2825 Specify the program to execute.
2827 .. option:: arguments=str : [exec]
2829 Specify arguments to pass to program.
2830 Some special variables can be expanded to pass fio's job details to the program.
2833 Replaced by the duration of the job in seconds.
2835 Replaced by the name of the job.
2837 .. option:: grace_time=int : [exec]
2839 Specify the time between the SIGTERM and SIGKILL signals. Default is 1 second.
2841 .. option:: std_redirect=bool : [exec]
2843 If set, stdout and stderr streams are redirected to files named from the job name. Default is true.
2845 .. option:: xnvme_async=str : [xnvme]
2847 Select the xnvme async command interface. This can take these values.
2850 This is default and use to emulate asynchronous I/O by using a
2851 single thread to create a queue pair on top of a synchronous
2852 I/O interface using the NVMe driver IOCTL.
2854 Emulate an asynchronous I/O interface with a pool of userspace
2855 threads on top of a synchronous I/O interface using the NVMe
2856 driver IOCTL. By default four threads are used.
2858 Linux native asynchronous I/O interface which supports both
2859 direct and buffered I/O.
2861 Fast Linux native asynchronous I/O interface for NVMe pass
2862 through commands. This only works with NVMe character device
2865 Use Linux aio for Asynchronous I/O.
2867 Use the posix asynchronous I/O interface to perform one or
2868 more I/O operations asynchronously.
2870 Use the user-space VFIO-based backend, implemented using
2871 libvfn instead of SPDK.
2873 Do not transfer any data; just pretend to. This is mainly used
2874 for introspective performance evaluation.
2876 .. option:: xnvme_sync=str : [xnvme]
2878 Select the xnvme synchronous command interface. This can take these values.
2881 This is default and uses Linux NVMe Driver ioctl() for
2884 This supports regular as well as vectored pread() and pwrite()
2887 This is the same as psync except that it also supports zone
2888 management commands using Linux block layer IOCTLs.
2890 .. option:: xnvme_admin=str : [xnvme]
2892 Select the xnvme admin command interface. This can take these values.
2895 This is default and uses linux NVMe Driver ioctl() for admin
2898 Use Linux Block Layer ioctl() and sysfs for admin commands.
2900 .. option:: xnvme_dev_nsid=int : [xnvme]
2902 xnvme namespace identifier for userspace NVMe driver, SPDK or vfio.
2904 .. option:: xnvme_dev_subnqn=str : [xnvme]
2906 Sets the subsystem NQN for fabrics. This is for xNVMe to utilize a
2907 fabrics target with multiple systems.
2909 .. option:: xnvme_mem=str : [xnvme]
2911 Select the xnvme memory backend. This can take these values.
2914 This is the default posix memory backend for linux NVMe driver.
2916 Use hugepages, instead of existing posix memory backend. The
2917 memory backend uses hugetlbfs. This require users to allocate
2918 hugepages, mount hugetlbfs and set an enviornment variable for
2921 Uses SPDK's memory allocator.
2923 Uses libvfn's memory allocator. This also specifies the use
2924 of libvfn backend instead of SPDK.
2926 .. option:: xnvme_iovec=int : [xnvme]
2928 If this option is set. xnvme will use vectored read/write commands.
2930 .. option:: libblkio_driver=str : [libblkio]
2932 The libblkio *driver* to use. Different drivers access devices through
2933 different underlying interfaces. Available drivers depend on the
2934 libblkio version in use and are listed at
2935 https://libblkio.gitlab.io/libblkio/blkio.html#drivers
2937 .. option:: libblkio_path=str : [libblkio]
2939 Sets the value of the driver-specific "path" property before connecting
2940 the libblkio instance, which identifies the target device or file on
2941 which to perform I/O. Its exact semantics are driver-dependent and not
2942 all drivers may support it; see
2943 https://libblkio.gitlab.io/libblkio/blkio.html#drivers
2945 .. option:: libblkio_pre_connect_props=str : [libblkio]
2947 A colon-separated list of additional libblkio properties to be set after
2948 creating but before connecting the libblkio instance. Each property must
2949 have the format ``<name>=<value>``. Colons can be escaped as ``\:``.
2950 These are set after the engine sets any other properties, so those can
2951 be overriden. Available properties depend on the libblkio version in use
2953 https://libblkio.gitlab.io/libblkio/blkio.html#properties
2955 .. option:: libblkio_num_entries=int : [libblkio]
2957 Sets the value of the driver-specific "num-entries" property before
2958 starting the libblkio instance. Its exact semantics are driver-dependent
2959 and not all drivers may support it; see
2960 https://libblkio.gitlab.io/libblkio/blkio.html#drivers
2962 .. option:: libblkio_queue_size=int : [libblkio]
2964 Sets the value of the driver-specific "queue-size" property before
2965 starting the libblkio instance. Its exact semantics are driver-dependent
2966 and not all drivers may support it; see
2967 https://libblkio.gitlab.io/libblkio/blkio.html#drivers
2969 .. option:: libblkio_pre_start_props=str : [libblkio]
2971 A colon-separated list of additional libblkio properties to be set after
2972 connecting but before starting the libblkio instance. Each property must
2973 have the format ``<name>=<value>``. Colons can be escaped as ``\:``.
2974 These are set after the engine sets any other properties, so those can
2975 be overriden. Available properties depend on the libblkio version in use
2977 https://libblkio.gitlab.io/libblkio/blkio.html#properties
2979 .. option:: libblkio_vectored : [libblkio]
2981 Submit vectored read and write requests.
2983 .. option:: libblkio_write_zeroes_on_trim : [libblkio]
2985 Submit trims as "write zeroes" requests instead of discard requests.
2987 .. option:: libblkio_wait_mode=str : [libblkio]
2989 How to wait for completions:
2992 Use a blocking call to ``blkioq_do_io()``.
2994 Use a blocking call to ``read()`` on the completion eventfd.
2996 Use a busy loop with a non-blocking call to ``blkioq_do_io()``.
2998 .. option:: libblkio_force_enable_completion_eventfd : [libblkio]
3000 Enable the queue's completion eventfd even when unused. This may impact
3001 performance. The default is to enable it only if
3002 :option:`libblkio_wait_mode=eventfd <libblkio_wait_mode>`.
3007 .. option:: iodepth=int
3009 Number of I/O units to keep in flight against the file. Note that
3010 increasing *iodepth* beyond 1 will not affect synchronous ioengines (except
3011 for small degrees when :option:`verify_async` is in use). Even async
3012 engines may impose OS restrictions causing the desired depth not to be
3013 achieved. This may happen on Linux when using libaio and not setting
3014 :option:`direct`\=1, since buffered I/O is not async on that OS. Keep an
3015 eye on the I/O depth distribution in the fio output to verify that the
3016 achieved depth is as expected. Default: 1.
3018 .. option:: iodepth_batch_submit=int, iodepth_batch=int
3020 This defines how many pieces of I/O to submit at once. It defaults to 1
3021 which means that we submit each I/O as soon as it is available, but can be
3022 raised to submit bigger batches of I/O at the time. If it is set to 0 the
3023 :option:`iodepth` value will be used.
3025 .. option:: iodepth_batch_complete_min=int, iodepth_batch_complete=int
3027 This defines how many pieces of I/O to retrieve at once. It defaults to 1
3028 which means that we'll ask for a minimum of 1 I/O in the retrieval process
3029 from the kernel. The I/O retrieval will go on until we hit the limit set by
3030 :option:`iodepth_low`. If this variable is set to 0, then fio will always
3031 check for completed events before queuing more I/O. This helps reduce I/O
3032 latency, at the cost of more retrieval system calls.
3034 .. option:: iodepth_batch_complete_max=int
3036 This defines maximum pieces of I/O to retrieve at once. This variable should
3037 be used along with :option:`iodepth_batch_complete_min`\=int variable,
3038 specifying the range of min and max amount of I/O which should be
3039 retrieved. By default it is equal to the :option:`iodepth_batch_complete_min`
3044 iodepth_batch_complete_min=1
3045 iodepth_batch_complete_max=<iodepth>
3047 which means that we will retrieve at least 1 I/O and up to the whole
3048 submitted queue depth. If none of I/O has been completed yet, we will wait.
3052 iodepth_batch_complete_min=0
3053 iodepth_batch_complete_max=<iodepth>
3055 which means that we can retrieve up to the whole submitted queue depth, but
3056 if none of I/O has been completed yet, we will NOT wait and immediately exit
3057 the system call. In this example we simply do polling.
3059 .. option:: iodepth_low=int
3061 The low water mark indicating when to start filling the queue
3062 again. Defaults to the same as :option:`iodepth`, meaning that fio will
3063 attempt to keep the queue full at all times. If :option:`iodepth` is set to
3064 e.g. 16 and *iodepth_low* is set to 4, then after fio has filled the queue of
3065 16 requests, it will let the depth drain down to 4 before starting to fill
3068 .. option:: serialize_overlap=bool
3070 Serialize in-flight I/Os that might otherwise cause or suffer from data races.
3071 When two or more I/Os are submitted simultaneously, there is no guarantee that
3072 the I/Os will be processed or completed in the submitted order. Further, if
3073 two or more of those I/Os are writes, any overlapping region between them can
3074 become indeterminate/undefined on certain storage. These issues can cause
3075 verification to fail erratically when at least one of the racing I/Os is
3076 changing data and the overlapping region has a non-zero size. Setting
3077 ``serialize_overlap`` tells fio to avoid provoking this behavior by explicitly
3078 serializing in-flight I/Os that have a non-zero overlap. Note that setting
3079 this option can reduce both performance and the :option:`iodepth` achieved.
3081 This option only applies to I/Os issued for a single job except when it is
3082 enabled along with :option:`io_submit_mode`\=offload. In offload mode, fio
3083 will check for overlap among all I/Os submitted by offload jobs with :option:`serialize_overlap`
3088 .. option:: io_submit_mode=str
3090 This option controls how fio submits the I/O to the I/O engine. The default
3091 is `inline`, which means that the fio job threads submit and reap I/O
3092 directly. If set to `offload`, the job threads will offload I/O submission
3093 to a dedicated pool of I/O threads. This requires some coordination and thus
3094 has a bit of extra overhead, especially for lower queue depth I/O where it
3095 can increase latencies. The benefit is that fio can manage submission rates
3096 independently of the device completion rates. This avoids skewed latency
3097 reporting if I/O gets backed up on the device side (the coordinated omission
3098 problem). Note that this option cannot reliably be used with async IO
3105 .. option:: thinktime=time
3107 Stall the job for the specified period of time after an I/O has completed before issuing the
3108 next. May be used to simulate processing being done by an application.
3109 When the unit is omitted, the value is interpreted in microseconds. See
3110 :option:`thinktime_blocks`, :option:`thinktime_iotime` and :option:`thinktime_spin`.
3112 .. option:: thinktime_spin=time
3114 Only valid if :option:`thinktime` is set - pretend to spend CPU time doing
3115 something with the data received, before falling back to sleeping for the
3116 rest of the period specified by :option:`thinktime`. When the unit is
3117 omitted, the value is interpreted in microseconds.
3119 .. option:: thinktime_blocks=int
3121 Only valid if :option:`thinktime` is set - control how many blocks to issue,
3122 before waiting :option:`thinktime` usecs. If not set, defaults to 1 which will make
3123 fio wait :option:`thinktime` usecs after every block. This effectively makes any
3124 queue depth setting redundant, since no more than 1 I/O will be queued
3125 before we have to complete it and do our :option:`thinktime`. In other words, this
3126 setting effectively caps the queue depth if the latter is larger.
3128 .. option:: thinktime_blocks_type=str
3130 Only valid if :option:`thinktime` is set - control how :option:`thinktime_blocks`
3131 triggers. The default is `complete`, which triggers thinktime when fio completes
3132 :option:`thinktime_blocks` blocks. If this is set to `issue`, then the trigger happens
3135 .. option:: thinktime_iotime=time
3137 Only valid if :option:`thinktime` is set - control :option:`thinktime`
3138 interval by time. The :option:`thinktime` stall is repeated after IOs
3139 are executed for :option:`thinktime_iotime`. For example,
3140 ``--thinktime_iotime=9s --thinktime=1s`` repeat 10-second cycle with IOs
3141 for 9 seconds and stall for 1 second. When the unit is omitted,
3142 :option:`thinktime_iotime` is interpreted as a number of seconds. If
3143 this option is used together with :option:`thinktime_blocks`, the
3144 :option:`thinktime` stall is repeated after :option:`thinktime_iotime`
3145 or after :option:`thinktime_blocks` IOs, whichever happens first.
3147 .. option:: rate=int[,int][,int]
3149 Cap the bandwidth used by this job. The number is in bytes/sec, the normal
3150 suffix rules apply. Comma-separated values may be specified for reads,
3151 writes, and trims as described in :option:`blocksize`.
3153 For example, using `rate=1m,500k` would limit reads to 1MiB/sec and writes to
3154 500KiB/sec. Capping only reads or writes can be done with `rate=,500k` or
3155 `rate=500k,` where the former will only limit writes (to 500KiB/sec) and the
3156 latter will only limit reads.
3158 .. option:: rate_min=int[,int][,int]
3160 Tell fio to do whatever it can to maintain at least this bandwidth. Failing
3161 to meet this requirement will cause the job to exit. Comma-separated values
3162 may be specified for reads, writes, and trims as described in
3163 :option:`blocksize`.
3165 .. option:: rate_iops=int[,int][,int]
3167 Cap the bandwidth to this number of IOPS. Basically the same as
3168 :option:`rate`, just specified independently of bandwidth. If the job is
3169 given a block size range instead of a fixed value, the smallest block size
3170 is used as the metric. Comma-separated values may be specified for reads,
3171 writes, and trims as described in :option:`blocksize`.
3173 .. option:: rate_iops_min=int[,int][,int]
3175 If fio doesn't meet this rate of I/O, it will cause the job to exit.
3176 Comma-separated values may be specified for reads, writes, and trims as
3177 described in :option:`blocksize`.
3179 .. option:: rate_process=str
3181 This option controls how fio manages rated I/O submissions. The default is
3182 `linear`, which submits I/O in a linear fashion with fixed delays between
3183 I/Os that gets adjusted based on I/O completion rates. If this is set to
3184 `poisson`, fio will submit I/O based on a more real world random request
3185 flow, known as the Poisson process
3186 (https://en.wikipedia.org/wiki/Poisson_point_process). The lambda will be
3187 10^6 / IOPS for the given workload.
3189 .. option:: rate_ignore_thinktime=bool
3191 By default, fio will attempt to catch up to the specified rate setting,
3192 if any kind of thinktime setting was used. If this option is set, then
3193 fio will ignore the thinktime and continue doing IO at the specified
3194 rate, instead of entering a catch-up mode after thinktime is done.
3200 .. option:: latency_target=time
3202 If set, fio will attempt to find the max performance point that the given
3203 workload will run at while maintaining a latency below this target. When
3204 the unit is omitted, the value is interpreted in microseconds. See
3205 :option:`latency_window` and :option:`latency_percentile`.
3207 .. option:: latency_window=time
3209 Used with :option:`latency_target` to specify the sample window that the job
3210 is run at varying queue depths to test the performance. When the unit is
3211 omitted, the value is interpreted in microseconds.
3213 .. option:: latency_percentile=float
3215 The percentage of I/Os that must fall within the criteria specified by
3216 :option:`latency_target` and :option:`latency_window`. If not set, this
3217 defaults to 100.0, meaning that all I/Os must be equal or below to the value
3218 set by :option:`latency_target`.
3220 .. option:: latency_run=bool
3222 Used with :option:`latency_target`. If false (default), fio will find
3223 the highest queue depth that meets :option:`latency_target` and exit. If
3224 true, fio will continue running and try to meet :option:`latency_target`
3225 by adjusting queue depth.
3227 .. option:: max_latency=time[,time][,time]
3229 If set, fio will exit the job with an ETIMEDOUT error if it exceeds this
3230 maximum latency. When the unit is omitted, the value is interpreted in
3231 microseconds. Comma-separated values may be specified for reads, writes,
3232 and trims as described in :option:`blocksize`.
3234 .. option:: rate_cycle=int
3236 Average bandwidth for :option:`rate` and :option:`rate_min` over this number
3237 of milliseconds. Defaults to 1000.
3243 .. option:: write_iolog=str
3245 Write the issued I/O patterns to the specified file. See
3246 :option:`read_iolog`. Specify a separate file for each job, otherwise the
3247 iologs will be interspersed and the file may be corrupt. This file will
3248 be opened in append mode.
3250 .. option:: read_iolog=str
3252 Open an iolog with the specified filename and replay the I/O patterns it
3253 contains. This can be used to store a workload and replay it sometime
3254 later. The iolog given may also be a blktrace binary file, which allows fio
3255 to replay a workload captured by :command:`blktrace`. See
3256 :manpage:`blktrace(8)` for how to capture such logging data. For blktrace
3257 replay, the file needs to be turned into a blkparse binary data file first
3258 (``blkparse <device> -o /dev/null -d file_for_fio.bin``).
3259 You can specify a number of files by separating the names with a ':'
3260 character. See the :option:`filename` option for information on how to
3261 escape ':' characters within the file names. These files will
3262 be sequentially assigned to job clones created by :option:`numjobs`.
3263 '-' is a reserved name, meaning read from stdin, notably if
3264 :option:`filename` is set to '-' which means stdin as well, then
3265 this flag can't be set to '-'.
3267 .. option:: read_iolog_chunked=bool
3269 Determines how iolog is read. If false(default) entire :option:`read_iolog`
3270 will be read at once. If selected true, input from iolog will be read
3271 gradually. Useful when iolog is very large, or it is generated.
3273 .. option:: merge_blktrace_file=str
3275 When specified, rather than replaying the logs passed to :option:`read_iolog`,
3276 the logs go through a merge phase which aggregates them into a single
3277 blktrace. The resulting file is then passed on as the :option:`read_iolog`
3278 parameter. The intention here is to make the order of events consistent.
3279 This limits the influence of the scheduler compared to replaying multiple
3280 blktraces via concurrent jobs.
3282 .. option:: merge_blktrace_scalars=float_list
3284 This is a percentage based option that is index paired with the list of
3285 files passed to :option:`read_iolog`. When merging is performed, scale
3286 the time of each event by the corresponding amount. For example,
3287 ``--merge_blktrace_scalars="50:100"`` runs the first trace in halftime
3288 and the second trace in realtime. This knob is separately tunable from
3289 :option:`replay_time_scale` which scales the trace during runtime and
3290 does not change the output of the merge unlike this option.
3292 .. option:: merge_blktrace_iters=float_list
3294 This is a whole number option that is index paired with the list of files
3295 passed to :option:`read_iolog`. When merging is performed, run each trace
3296 for the specified number of iterations. For example,
3297 ``--merge_blktrace_iters="2:1"`` runs the first trace for two iterations
3298 and the second trace for one iteration.
3300 .. option:: replay_no_stall=bool
3302 When replaying I/O with :option:`read_iolog` the default behavior is to
3303 attempt to respect the timestamps within the log and replay them with the
3304 appropriate delay between IOPS. By setting this variable fio will not
3305 respect the timestamps and attempt to replay them as fast as possible while
3306 still respecting ordering. The result is the same I/O pattern to a given
3307 device, but different timings.
3309 .. option:: replay_time_scale=int
3311 When replaying I/O with :option:`read_iolog`, fio will honor the
3312 original timing in the trace. With this option, it's possible to scale
3313 the time. It's a percentage option, if set to 50 it means run at 50%
3314 the original IO rate in the trace. If set to 200, run at twice the
3315 original IO rate. Defaults to 100.
3317 .. option:: replay_redirect=str
3319 While replaying I/O patterns using :option:`read_iolog` the default behavior
3320 is to replay the IOPS onto the major/minor device that each IOP was recorded
3321 from. This is sometimes undesirable because on a different machine those
3322 major/minor numbers can map to a different device. Changing hardware on the
3323 same system can also result in a different major/minor mapping.
3324 ``replay_redirect`` causes all I/Os to be replayed onto the single specified
3325 device regardless of the device it was recorded
3326 from. i.e. :option:`replay_redirect`\= :file:`/dev/sdc` would cause all I/O
3327 in the blktrace or iolog to be replayed onto :file:`/dev/sdc`. This means
3328 multiple devices will be replayed onto a single device, if the trace
3329 contains multiple devices. If you want multiple devices to be replayed
3330 concurrently to multiple redirected devices you must blkparse your trace
3331 into separate traces and replay them with independent fio invocations.
3332 Unfortunately this also breaks the strict time ordering between multiple
3335 .. option:: replay_align=int
3337 Force alignment of the byte offsets in a trace to this value. The value
3338 must be a power of 2.
3340 .. option:: replay_scale=int
3342 Scale byte offsets down by this factor when replaying traces. Should most
3343 likely use :option:`replay_align` as well.
3345 .. option:: replay_skip=str
3347 Sometimes it's useful to skip certain IO types in a replay trace.
3348 This could be, for instance, eliminating the writes in the trace.
3349 Or not replaying the trims/discards, if you are redirecting to
3350 a device that doesn't support them. This option takes a comma
3351 separated list of read, write, trim, sync.
3354 Threads, processes and job synchronization
3355 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3359 Fio defaults to creating jobs by using fork, however if this option is
3360 given, fio will create jobs by using POSIX Threads' function
3361 :manpage:`pthread_create(3)` to create threads instead.
3363 .. option:: wait_for=str
3365 If set, the current job won't be started until all workers of the specified
3366 waitee job are done.
3368 ``wait_for`` operates on the job name basis, so there are a few
3369 limitations. First, the waitee must be defined prior to the waiter job
3370 (meaning no forward references). Second, if a job is being referenced as a
3371 waitee, it must have a unique name (no duplicate waitees).
3373 .. option:: nice=int
3375 Run the job with the given nice value. See man :manpage:`nice(2)`.
3377 On Windows, values less than -15 set the process class to "High"; -1 through
3378 -15 set "Above Normal"; 1 through 15 "Below Normal"; and above 15 "Idle"
3381 .. option:: prio=int
3383 Set the I/O priority value of this job. Linux limits us to a positive value
3384 between 0 and 7, with 0 being the highest. See man
3385 :manpage:`ionice(1)`. Refer to an appropriate manpage for other operating
3386 systems since meaning of priority may differ. For per-command priority
3387 setting, see I/O engine specific :option:`cmdprio_percentage` and
3388 :option:`cmdprio` options.
3390 .. option:: prioclass=int
3392 Set the I/O priority class. See man :manpage:`ionice(1)`. For per-command
3393 priority setting, see I/O engine specific :option:`cmdprio_percentage`
3394 and :option:`cmdprio_class` options.
3396 .. option:: cpus_allowed=str
3398 Controls the same options as :option:`cpumask`, but accepts a textual
3399 specification of the permitted CPUs instead and CPUs are indexed from 0. So
3400 to use CPUs 0 and 5 you would specify ``cpus_allowed=0,5``. This option also
3401 allows a range of CPUs to be specified -- say you wanted a binding to CPUs
3402 0, 5, and 8 to 15, you would set ``cpus_allowed=0,5,8-15``.
3404 On Windows, when ``cpus_allowed`` is unset only CPUs from fio's current
3405 processor group will be used and affinity settings are inherited from the
3406 system. An fio build configured to target Windows 7 makes options that set
3407 CPUs processor group aware and values will set both the processor group
3408 and a CPU from within that group. For example, on a system where processor
3409 group 0 has 40 CPUs and processor group 1 has 32 CPUs, ``cpus_allowed``
3410 values between 0 and 39 will bind CPUs from processor group 0 and
3411 ``cpus_allowed`` values between 40 and 71 will bind CPUs from processor
3412 group 1. When using ``cpus_allowed_policy=shared`` all CPUs specified by a
3413 single ``cpus_allowed`` option must be from the same processor group. For
3414 Windows fio builds not built for Windows 7, CPUs will only be selected from
3415 (and be relative to) whatever processor group fio happens to be running in
3416 and CPUs from other processor groups cannot be used.
3418 .. option:: cpus_allowed_policy=str
3420 Set the policy of how fio distributes the CPUs specified by
3421 :option:`cpus_allowed` or :option:`cpumask`. Two policies are supported:
3424 All jobs will share the CPU set specified.
3426 Each job will get a unique CPU from the CPU set.
3428 **shared** is the default behavior, if the option isn't specified. If
3429 **split** is specified, then fio will assign one cpu per job. If not
3430 enough CPUs are given for the jobs listed, then fio will roundrobin the CPUs
3433 .. option:: cpumask=int
3435 Set the CPU affinity of this job. The parameter given is a bit mask of
3436 allowed CPUs the job may run on. So if you want the allowed CPUs to be 1
3437 and 5, you would pass the decimal value of (1 << 1 | 1 << 5), or 34. See man
3438 :manpage:`sched_setaffinity(2)`. This may not work on all supported
3439 operating systems or kernel versions. This option doesn't work well for a
3440 higher CPU count than what you can store in an integer mask, so it can only
3441 control cpus 1-32. For boxes with larger CPU counts, use
3442 :option:`cpus_allowed`.
3444 .. option:: numa_cpu_nodes=str
3446 Set this job running on specified NUMA nodes' CPUs. The arguments allow
3447 comma delimited list of cpu numbers, A-B ranges, or `all`. Note, to enable
3448 NUMA options support, fio must be built on a system with libnuma-dev(el)
3451 .. option:: numa_mem_policy=str
3453 Set this job's memory policy and corresponding NUMA nodes. Format of the
3458 ``mode`` is one of the following memory policies: ``default``, ``prefer``,
3459 ``bind``, ``interleave`` or ``local``. For ``default`` and ``local`` memory
3460 policies, no node needs to be specified. For ``prefer``, only one node is
3461 allowed. For ``bind`` and ``interleave`` the ``nodelist`` may be as
3462 follows: a comma delimited list of numbers, A-B ranges, or `all`.
3464 .. option:: cgroup=str
3466 Add job to this control group. If it doesn't exist, it will be created. The
3467 system must have a mounted cgroup blkio mount point for this to work. If
3468 your system doesn't have it mounted, you can do so with::
3470 # mount -t cgroup -o blkio none /cgroup
3472 .. option:: cgroup_weight=int
3474 Set the weight of the cgroup to this value. See the documentation that comes
3475 with the kernel, allowed values are in the range of 100..1000.
3477 .. option:: cgroup_nodelete=bool
3479 Normally fio will delete the cgroups it has created after the job
3480 completion. To override this behavior and to leave cgroups around after the
3481 job completion, set ``cgroup_nodelete=1``. This can be useful if one wants
3482 to inspect various cgroup files after job completion. Default: false.
3484 .. option:: flow_id=int
3486 The ID of the flow. If not specified, it defaults to being a global
3487 flow. See :option:`flow`.
3489 .. option:: flow=int
3491 Weight in token-based flow control. If this value is used, then fio
3492 regulates the activity between two or more jobs sharing the same
3493 flow_id. Fio attempts to keep each job activity proportional to other
3494 jobs' activities in the same flow_id group, with respect to requested
3495 weight per job. That is, if one job has `flow=3', another job has
3496 `flow=2' and another with `flow=1`, then there will be a roughly 3:2:1
3497 ratio in how much one runs vs the others.
3499 .. option:: flow_sleep=int
3501 The period of time, in microseconds, to wait after the flow counter
3502 has exceeded its proportion before retrying operations.
3504 .. option:: stonewall, wait_for_previous
3506 Wait for preceding jobs in the job file to exit, before starting this
3507 one. Can be used to insert serialization points in the job file. A stone
3508 wall also implies starting a new reporting group, see
3509 :option:`group_reporting`.
3513 By default, fio will continue running all other jobs when one job finishes.
3514 Sometimes this is not the desired action. Setting ``exitall`` will instead
3515 make fio terminate all jobs in the same group, as soon as one job of that
3518 .. option:: exit_what=str
3520 By default, fio will continue running all other jobs when one job finishes.
3521 Sometimes this is not the desired action. Setting ``exitall`` will
3522 instead make fio terminate all jobs in the same group. The option
3523 ``exit_what`` allows to control which jobs get terminated when ``exitall`` is
3524 enabled. The default is ``group`` and does not change the behaviour of
3525 ``exitall``. The setting ``all`` terminates all jobs. The setting ``stonewall``
3526 terminates all currently running jobs across all groups and continues execution
3527 with the next stonewalled group.
3529 .. option:: exec_prerun=str
3531 Before running this job, issue the command specified through
3532 :manpage:`system(3)`. Output is redirected in a file called
3533 :file:`jobname.prerun.txt`.
3535 .. option:: exec_postrun=str
3537 After the job completes, issue the command specified though
3538 :manpage:`system(3)`. Output is redirected in a file called
3539 :file:`jobname.postrun.txt`.
3543 Instead of running as the invoking user, set the user ID to this value
3544 before the thread/process does any work.
3548 Set group ID, see :option:`uid`.
3554 .. option:: verify_only
3556 Do not perform specified workload, only verify data still matches previous
3557 invocation of this workload. This option allows one to check data multiple
3558 times at a later date without overwriting it. This option makes sense only
3559 for workloads that write data, and does not support workloads with the
3560 :option:`time_based` option set.
3562 .. option:: do_verify=bool
3564 Run the verify phase after a write phase. Only valid if :option:`verify` is
3567 .. option:: verify=str
3569 If writing to a file, fio can verify the file contents after each iteration
3570 of the job. Each verification method also implies verification of special
3571 header, which is written to the beginning of each block. This header also
3572 includes meta information, like offset of the block, block number, timestamp
3573 when block was written, etc. :option:`verify` can be combined with
3574 :option:`verify_pattern` option. The allowed values are:
3577 Use an md5 sum of the data area and store it in the header of
3581 Use an experimental crc64 sum of the data area and store it in the
3582 header of each block.
3585 Use a crc32c sum of the data area and store it in the header of
3586 each block. This will automatically use hardware acceleration
3587 (e.g. SSE4.2 on an x86 or CRC crypto extensions on ARM64) but will
3588 fall back to software crc32c if none is found. Generally the
3589 fastest checksum fio supports when hardware accelerated.
3595 Use a crc32 sum of the data area and store it in the header of each
3599 Use a crc16 sum of the data area and store it in the header of each
3603 Use a crc7 sum of the data area and store it in the header of each
3607 Use xxhash as the checksum function. Generally the fastest software
3608 checksum that fio supports.
3611 Use sha512 as the checksum function.
3614 Use sha256 as the checksum function.
3617 Use optimized sha1 as the checksum function.
3620 Use optimized sha3-224 as the checksum function.
3623 Use optimized sha3-256 as the checksum function.
3626 Use optimized sha3-384 as the checksum function.
3629 Use optimized sha3-512 as the checksum function.
3632 This option is deprecated, since now meta information is included in
3633 generic verification header and meta verification happens by
3634 default. For detailed information see the description of the
3635 :option:`verify` setting. This option is kept because of
3636 compatibility's sake with old configurations. Do not use it.
3639 Verify a strict pattern. Normally fio includes a header with some
3640 basic information and checksumming, but if this option is set, only
3641 the specific pattern set with :option:`verify_pattern` is verified.
3644 Only pretend to verify. Useful for testing internals with
3645 :option:`ioengine`\=null, not for much else.
3647 This option can be used for repeated burn-in tests of a system to make sure
3648 that the written data is also correctly read back. If the data direction
3649 given is a read or random read, fio will assume that it should verify a
3650 previously written file. If the data direction includes any form of write,
3651 the verify will be of the newly written data.
3653 To avoid false verification errors, do not use the norandommap option when
3654 verifying data with async I/O engines and I/O depths > 1. Or use the
3655 norandommap and the lfsr random generator together to avoid writing to the
3656 same offset with multiple outstanding I/Os.
3658 .. option:: verify_offset=int
3660 Swap the verification header with data somewhere else in the block before
3661 writing. It is swapped back before verifying.
3663 .. option:: verify_interval=int
3665 Write the verification header at a finer granularity than the
3666 :option:`blocksize`. It will be written for chunks the size of
3667 ``verify_interval``. :option:`blocksize` should divide this evenly.
3669 .. option:: verify_pattern=str
3671 If set, fio will fill the I/O buffers with this pattern. Fio defaults to
3672 filling with totally random bytes, but sometimes it's interesting to fill
3673 with a known pattern for I/O verification purposes. Depending on the width
3674 of the pattern, fio will fill 1/2/3/4 bytes of the buffer at the time (it can
3675 be either a decimal or a hex number). The ``verify_pattern`` if larger than
3676 a 32-bit quantity has to be a hex number that starts with either "0x" or
3677 "0X". Use with :option:`verify`. Also, ``verify_pattern`` supports %o
3678 format, which means that for each block offset will be written and then
3679 verified back, e.g.::
3683 Or use combination of everything::
3685 verify_pattern=0xff%o"abcd"-12
3687 .. option:: verify_fatal=bool
3689 Normally fio will keep checking the entire contents before quitting on a
3690 block verification failure. If this option is set, fio will exit the job on
3691 the first observed failure. Default: false.
3693 .. option:: verify_dump=bool
3695 If set, dump the contents of both the original data block and the data block
3696 we read off disk to files. This allows later analysis to inspect just what
3697 kind of data corruption occurred. Off by default.
3699 .. option:: verify_async=int
3701 Fio will normally verify I/O inline from the submitting thread. This option
3702 takes an integer describing how many async offload threads to create for I/O
3703 verification instead, causing fio to offload the duty of verifying I/O
3704 contents to one or more separate threads. If using this offload option, even
3705 sync I/O engines can benefit from using an :option:`iodepth` setting higher
3706 than 1, as it allows them to have I/O in flight while verifies are running.
3707 Defaults to 0 async threads, i.e. verification is not asynchronous.
3709 .. option:: verify_async_cpus=str
3711 Tell fio to set the given CPU affinity on the async I/O verification
3712 threads. See :option:`cpus_allowed` for the format used.
3714 .. option:: verify_backlog=int
3716 Fio will normally verify the written contents of a job that utilizes verify
3717 once that job has completed. In other words, everything is written then
3718 everything is read back and verified. You may want to verify continually
3719 instead for a variety of reasons. Fio stores the meta data associated with
3720 an I/O block in memory, so for large verify workloads, quite a bit of memory
3721 would be used up holding this meta data. If this option is enabled, fio will
3722 write only N blocks before verifying these blocks.
3724 .. option:: verify_backlog_batch=int
3726 Control how many blocks fio will verify if :option:`verify_backlog` is
3727 set. If not set, will default to the value of :option:`verify_backlog`
3728 (meaning the entire queue is read back and verified). If
3729 ``verify_backlog_batch`` is less than :option:`verify_backlog` then not all
3730 blocks will be verified, if ``verify_backlog_batch`` is larger than
3731 :option:`verify_backlog`, some blocks will be verified more than once.
3733 .. option:: verify_state_save=bool
3735 When a job exits during the write phase of a verify workload, save its
3736 current state. This allows fio to replay up until that point, if the verify
3737 state is loaded for the verify read phase. The format of the filename is,
3740 <type>-<jobname>-<jobindex>-verify.state.
3742 <type> is "local" for a local run, "sock" for a client/server socket
3743 connection, and "ip" (192.168.0.1, for instance) for a networked
3744 client/server connection. Defaults to true.
3746 .. option:: verify_state_load=bool
3748 If a verify termination trigger was used, fio stores the current write state
3749 of each thread. This can be used at verification time so that fio knows how
3750 far it should verify. Without this information, fio will run a full
3751 verification pass, according to the settings in the job file used. Default
3754 .. option:: trim_percentage=int
3756 Number of verify blocks to discard/trim.
3758 .. option:: trim_verify_zero=bool
3760 Verify that trim/discarded blocks are returned as zeros.
3762 .. option:: trim_backlog=int
3764 Trim after this number of blocks are written.
3766 .. option:: trim_backlog_batch=int
3768 Trim this number of I/O blocks.
3770 .. option:: experimental_verify=bool
3772 Enable experimental verification. Standard verify records I/O metadata
3773 for later use during the verification phase. Experimental verify
3774 instead resets the file after the write phase and then replays I/Os for
3775 the verification phase.
3780 .. option:: steadystate=str:float, ss=str:float
3782 Define the criterion and limit for assessing steady state performance. The
3783 first parameter designates the criterion whereas the second parameter sets
3784 the threshold. When the criterion falls below the threshold for the
3785 specified duration, the job will stop. For example, `iops_slope:0.1%` will
3786 direct fio to terminate the job when the least squares regression slope
3787 falls below 0.1% of the mean IOPS. If :option:`group_reporting` is enabled
3788 this will apply to all jobs in the group. Below is the list of available
3789 steady state assessment criteria. All assessments are carried out using only
3790 data from the rolling collection window. Threshold limits can be expressed
3791 as a fixed value or as a percentage of the mean in the collection window.
3793 When using this feature, most jobs should include the :option:`time_based`
3794 and :option:`runtime` options or the :option:`loops` option so that fio does not
3795 stop running after it has covered the full size of the specified file(s) or device(s).
3798 Collect IOPS data. Stop the job if all individual IOPS measurements
3799 are within the specified limit of the mean IOPS (e.g., ``iops:2``
3800 means that all individual IOPS values must be within 2 of the mean,
3801 whereas ``iops:0.2%`` means that all individual IOPS values must be
3802 within 0.2% of the mean IOPS to terminate the job).
3805 Collect IOPS data and calculate the least squares regression
3806 slope. Stop the job if the slope falls below the specified limit.
3809 Collect bandwidth data. Stop the job if all individual bandwidth
3810 measurements are within the specified limit of the mean bandwidth.
3813 Collect bandwidth data and calculate the least squares regression
3814 slope. Stop the job if the slope falls below the specified limit.
3816 .. option:: steadystate_duration=time, ss_dur=time
3818 A rolling window of this duration will be used to judge whether steady state
3819 has been reached. Data will be collected once per second. The default is 0
3820 which disables steady state detection. When the unit is omitted, the
3821 value is interpreted in seconds.
3823 .. option:: steadystate_ramp_time=time, ss_ramp=time
3825 Allow the job to run for the specified duration before beginning data
3826 collection for checking the steady state job termination criterion. The
3827 default is 0. When the unit is omitted, the value is interpreted in seconds.
3830 Measurements and reporting
3831 ~~~~~~~~~~~~~~~~~~~~~~~~~~
3833 .. option:: per_job_logs=bool
3835 If set, this generates bw/clat/iops log with per file private filenames. If
3836 not set, jobs with identical names will share the log filename. Default:
3839 .. option:: group_reporting
3841 It may sometimes be interesting to display statistics for groups of jobs as
3842 a whole instead of for each individual job. This is especially true if
3843 :option:`numjobs` is used; looking at individual thread/process output
3844 quickly becomes unwieldy. To see the final report per-group instead of
3845 per-job, use :option:`group_reporting`. Jobs in a file will be part of the
3846 same reporting group, unless if separated by a :option:`stonewall`, or by
3847 using :option:`new_group`.
3849 .. option:: new_group
3851 Start a new reporting group. See: :option:`group_reporting`. If not given,
3852 all jobs in a file will be part of the same reporting group, unless
3853 separated by a :option:`stonewall`.
3855 .. option:: stats=bool
3857 By default, fio collects and shows final output results for all jobs
3858 that run. If this option is set to 0, then fio will ignore it in
3859 the final stat output.
3861 .. option:: write_bw_log=str
3863 If given, write a bandwidth log for this job. Can be used to store data of
3864 the bandwidth of the jobs in their lifetime.
3866 If no str argument is given, the default filename of
3867 :file:`jobname_type.x.log` is used. Even when the argument is given, fio
3868 will still append the type of log. So if one specifies::
3872 The actual log name will be :file:`foo_bw.x.log` where `x` is the index
3873 of the job (`1..N`, where `N` is the number of jobs). If
3874 :option:`per_job_logs` is false, then the filename will not include the
3877 The included :command:`fio_generate_plots` script uses :command:`gnuplot` to turn these
3878 text files into nice graphs. See `Log File Formats`_ for how data is
3879 structured within the file.
3881 .. option:: write_lat_log=str
3883 Same as :option:`write_bw_log`, except this option creates I/O
3884 submission (e.g., :file:`name_slat.x.log`), completion (e.g.,
3885 :file:`name_clat.x.log`), and total (e.g., :file:`name_lat.x.log`)
3886 latency files instead. See :option:`write_bw_log` for details about
3887 the filename format and `Log File Formats`_ for how data is structured
3890 .. option:: write_hist_log=str
3892 Same as :option:`write_bw_log` but writes an I/O completion latency
3893 histogram file (e.g., :file:`name_hist.x.log`) instead. Note that this
3894 file will be empty unless :option:`log_hist_msec` has also been set.
3895 See :option:`write_bw_log` for details about the filename format and
3896 `Log File Formats`_ for how data is structured within the file.
3898 .. option:: write_iops_log=str
3900 Same as :option:`write_bw_log`, but writes an IOPS file (e.g.
3901 :file:`name_iops.x.log`) instead. Because fio defaults to individual
3902 I/O logging, the value entry in the IOPS log will be 1 unless windowed
3903 logging (see :option:`log_avg_msec`) has been enabled. See
3904 :option:`write_bw_log` for details about the filename format and `Log
3905 File Formats`_ for how data is structured within the file.
3907 .. option:: log_entries=int
3909 By default, fio will log an entry in the iops, latency, or bw log for
3910 every I/O that completes. The initial number of I/O log entries is 1024.
3911 When the log entries are all used, new log entries are dynamically
3912 allocated. This dynamic log entry allocation may negatively impact
3913 time-related statistics such as I/O tail latencies (e.g. 99.9th percentile
3914 completion latency). This option allows specifying a larger initial
3915 number of log entries to avoid run-time allocations of new log entries,
3916 resulting in more precise time-related I/O statistics.
3917 Also see :option:`log_avg_msec`. Defaults to 1024.
3919 .. option:: log_avg_msec=int
3921 By default, fio will log an entry in the iops, latency, or bw log for every
3922 I/O that completes. When writing to the disk log, that can quickly grow to a
3923 very large size. Setting this option makes fio average the each log entry
3924 over the specified period of time, reducing the resolution of the log. See
3925 :option:`log_max_value` as well. Defaults to 0, logging all entries.
3926 Also see `Log File Formats`_.
3928 .. option:: log_hist_msec=int
3930 Same as :option:`log_avg_msec`, but logs entries for completion latency
3931 histograms. Computing latency percentiles from averages of intervals using
3932 :option:`log_avg_msec` is inaccurate. Setting this option makes fio log
3933 histogram entries over the specified period of time, reducing log sizes for
3934 high IOPS devices while retaining percentile accuracy. See
3935 :option:`log_hist_coarseness` and :option:`write_hist_log` as well.
3936 Defaults to 0, meaning histogram logging is disabled.
3938 .. option:: log_hist_coarseness=int
3940 Integer ranging from 0 to 6, defining the coarseness of the resolution of
3941 the histogram logs enabled with :option:`log_hist_msec`. For each increment
3942 in coarseness, fio outputs half as many bins. Defaults to 0, for which
3943 histogram logs contain 1216 latency bins. See :option:`write_hist_log`
3944 and `Log File Formats`_.
3946 .. option:: log_max_value=bool
3948 If :option:`log_avg_msec` is set, fio logs the average over that window. If
3949 you instead want to log the maximum value, set this option to 1. Defaults to
3950 0, meaning that averaged values are logged.
3952 .. option:: log_offset=bool
3954 If this is set, the iolog options will include the byte offset for the I/O
3955 entry as well as the other data values. Defaults to 0 meaning that
3956 offsets are not present in logs. Also see `Log File Formats`_.
3958 .. option:: log_compression=int
3960 If this is set, fio will compress the I/O logs as it goes, to keep the
3961 memory footprint lower. When a log reaches the specified size, that chunk is
3962 removed and compressed in the background. Given that I/O logs are fairly
3963 highly compressible, this yields a nice memory savings for longer runs. The
3964 downside is that the compression will consume some background CPU cycles, so
3965 it may impact the run. This, however, is also true if the logging ends up
3966 consuming most of the system memory. So pick your poison. The I/O logs are
3967 saved normally at the end of a run, by decompressing the chunks and storing
3968 them in the specified log file. This feature depends on the availability of
3971 .. option:: log_compression_cpus=str
3973 Define the set of CPUs that are allowed to handle online log compression for
3974 the I/O jobs. This can provide better isolation between performance
3975 sensitive jobs, and background compression work. See
3976 :option:`cpus_allowed` for the format used.
3978 .. option:: log_store_compressed=bool
3980 If set, fio will store the log files in a compressed format. They can be
3981 decompressed with fio, using the :option:`--inflate-log` command line
3982 parameter. The files will be stored with a :file:`.fz` suffix.
3984 .. option:: log_unix_epoch=bool
3986 If set, fio will log Unix timestamps to the log files produced by enabling
3987 write_type_log for each log type, instead of the default zero-based
3990 .. option:: log_alternate_epoch=bool
3992 If set, fio will log timestamps based on the epoch used by the clock specified
3993 in the log_alternate_epoch_clock_id option, to the log files produced by
3994 enabling write_type_log for each log type, instead of the default zero-based
3997 .. option:: log_alternate_epoch_clock_id=int
3999 Specifies the clock_id to be used by clock_gettime to obtain the alternate epoch
4000 if either log_unix_epoch or log_alternate_epoch are true. Otherwise has no
4001 effect. Default value is 0, or CLOCK_REALTIME.
4003 .. option:: block_error_percentiles=bool
4005 If set, record errors in trim block-sized units from writes and trims and
4006 output a histogram of how many trims it took to get to errors, and what kind
4007 of error was encountered.
4009 .. option:: bwavgtime=int
4011 Average the calculated bandwidth over the given time. Value is specified in
4012 milliseconds. If the job also does bandwidth logging through
4013 :option:`write_bw_log`, then the minimum of this option and
4014 :option:`log_avg_msec` will be used. Default: 500ms.
4016 .. option:: iopsavgtime=int
4018 Average the calculated IOPS over the given time. Value is specified in
4019 milliseconds. If the job also does IOPS logging through
4020 :option:`write_iops_log`, then the minimum of this option and
4021 :option:`log_avg_msec` will be used. Default: 500ms.
4023 .. option:: disk_util=bool
4025 Generate disk utilization statistics, if the platform supports it.
4028 .. option:: disable_lat=bool
4030 Disable measurements of total latency numbers. Useful only for cutting back
4031 the number of calls to :manpage:`gettimeofday(2)`, as that does impact
4032 performance at really high IOPS rates. Note that to really get rid of a
4033 large amount of these calls, this option must be used with
4034 :option:`disable_slat` and :option:`disable_bw_measurement` as well.
4036 .. option:: disable_clat=bool
4038 Disable measurements of completion latency numbers. See
4039 :option:`disable_lat`.
4041 .. option:: disable_slat=bool
4043 Disable measurements of submission latency numbers. See
4044 :option:`disable_lat`.
4046 .. option:: disable_bw_measurement=bool, disable_bw=bool
4048 Disable measurements of throughput/bandwidth numbers. See
4049 :option:`disable_lat`.
4051 .. option:: slat_percentiles=bool
4053 Report submission latency percentiles. Submission latency is not recorded
4054 for synchronous ioengines.
4056 .. option:: clat_percentiles=bool
4058 Report completion latency percentiles.
4060 .. option:: lat_percentiles=bool
4062 Report total latency percentiles. Total latency is the sum of submission
4063 latency and completion latency.
4065 .. option:: percentile_list=float_list
4067 Overwrite the default list of percentiles for latencies and the block error
4068 histogram. Each number is a floating point number in the range (0,100], and
4069 the maximum length of the list is 20. Use ``:`` to separate the numbers. For
4070 example, ``--percentile_list=99.5:99.9`` will cause fio to report the
4071 latency durations below which 99.5% and 99.9% of the observed latencies fell,
4074 .. option:: significant_figures=int
4076 If using :option:`--output-format` of `normal`, set the significant
4077 figures to this value. Higher values will yield more precise IOPS and
4078 throughput units, while lower values will round. Requires a minimum
4079 value of 1 and a maximum value of 10. Defaults to 4.
4085 .. option:: exitall_on_error
4087 When one job finishes in error, terminate the rest. The default is to wait
4088 for each job to finish.
4090 .. option:: continue_on_error=str
4092 Normally fio will exit the job on the first observed failure. If this option
4093 is set, fio will continue the job when there is a 'non-fatal error' (EIO or
4094 EILSEQ) until the runtime is exceeded or the I/O size specified is
4095 completed. If this option is used, there are two more stats that are
4096 appended, the total error count and the first error. The error field given
4097 in the stats is the first error that was hit during the run.
4099 Note: a write error from the device may go unnoticed by fio when using
4100 buffered IO, as the write() (or similar) system call merely dirties the
4101 kernel pages, unless :option:`sync` or :option:`direct` is used. Device IO
4102 errors occur when the dirty data is actually written out to disk. If fully
4103 sync writes aren't desirable, :option:`fsync` or :option:`fdatasync` can be
4104 used as well. This is specific to writes, as reads are always synchronous.
4106 The allowed values are:
4109 Exit on any I/O or verify errors.
4112 Continue on read errors, exit on all others.
4115 Continue on write errors, exit on all others.
4118 Continue on any I/O error, exit on all others.
4121 Continue on verify errors, exit on all others.
4124 Continue on all errors.
4127 Backward-compatible alias for 'none'.
4130 Backward-compatible alias for 'all'.
4132 .. option:: ignore_error=str
4134 Sometimes you want to ignore some errors during test in that case you can
4135 specify error list for each error type, instead of only being able to
4136 ignore the default 'non-fatal error' using :option:`continue_on_error`.
4137 ``ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST`` errors for
4138 given error type is separated with ':'. Error may be symbol ('ENOSPC',
4139 'ENOMEM') or integer. Example::
4141 ignore_error=EAGAIN,ENOSPC:122
4143 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from
4144 WRITE. This option works by overriding :option:`continue_on_error` with
4145 the list of errors for each error type if any.
4147 .. option:: error_dump=bool
4149 If set dump every error even if it is non fatal, true by default. If
4150 disabled only fatal error will be dumped.
4152 Running predefined workloads
4153 ----------------------------
4155 Fio includes predefined profiles that mimic the I/O workloads generated by
4158 .. option:: profile=str
4160 The predefined workload to run. Current profiles are:
4163 Threaded I/O bench (tiotest/tiobench) like workload.
4166 Aerospike Certification Tool (ACT) like workload.
4168 To view a profile's additional options use :option:`--cmdhelp` after specifying
4169 the profile. For example::
4171 $ fio --profile=act --cmdhelp
4176 .. option:: device-names=str
4181 .. option:: load=int
4184 ACT load multiplier. Default: 1.
4186 .. option:: test-duration=time
4189 How long the entire test takes to run. When the unit is omitted, the value
4190 is given in seconds. Default: 24h.
4192 .. option:: threads-per-queue=int
4195 Number of read I/O threads per device. Default: 8.
4197 .. option:: read-req-num-512-blocks=int
4200 Number of 512B blocks to read at the time. Default: 3.
4202 .. option:: large-block-op-kbytes=int
4205 Size of large block ops in KiB (writes). Default: 131072.
4210 Set to run ACT prep phase.
4212 Tiobench profile options
4213 ~~~~~~~~~~~~~~~~~~~~~~~~
4215 .. option:: size=str
4220 .. option:: block=int
4223 Block size in bytes. Default: 4096.
4225 .. option:: numruns=int
4235 .. option:: threads=int
4240 Interpreting the output
4241 -----------------------
4244 Example output was based on the following:
4245 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --time_based \
4246 --rate=1256k --bs=14K --name=quick --runtime=1s --name=mixed \
4247 --runtime=2m --rw=rw
4249 Fio spits out a lot of output. While running, fio will display the status of the
4250 jobs created. An example of that would be::
4252 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]
4254 The characters inside the first set of square brackets denote the current status of
4255 each thread. The first character is the first job defined in the job file, and so
4256 forth. The possible values (in typical life cycle order) are:
4258 +------+-----+-----------------------------------------------------------+
4260 +======+=====+===========================================================+
4261 | P | | Thread setup, but not started. |
4262 +------+-----+-----------------------------------------------------------+
4263 | C | | Thread created. |
4264 +------+-----+-----------------------------------------------------------+
4265 | I | | Thread initialized, waiting or generating necessary data. |
4266 +------+-----+-----------------------------------------------------------+
4267 | | p | Thread running pre-reading file(s). |
4268 +------+-----+-----------------------------------------------------------+
4269 | | / | Thread is in ramp period. |
4270 +------+-----+-----------------------------------------------------------+
4271 | | R | Running, doing sequential reads. |
4272 +------+-----+-----------------------------------------------------------+
4273 | | r | Running, doing random reads. |
4274 +------+-----+-----------------------------------------------------------+
4275 | | W | Running, doing sequential writes. |
4276 +------+-----+-----------------------------------------------------------+
4277 | | w | Running, doing random writes. |
4278 +------+-----+-----------------------------------------------------------+
4279 | | M | Running, doing mixed sequential reads/writes. |
4280 +------+-----+-----------------------------------------------------------+
4281 | | m | Running, doing mixed random reads/writes. |
4282 +------+-----+-----------------------------------------------------------+
4283 | | D | Running, doing sequential trims. |
4284 +------+-----+-----------------------------------------------------------+
4285 | | d | Running, doing random trims. |
4286 +------+-----+-----------------------------------------------------------+
4287 | | F | Running, currently waiting for :manpage:`fsync(2)`. |
4288 +------+-----+-----------------------------------------------------------+
4289 | | V | Running, doing verification of written data. |
4290 +------+-----+-----------------------------------------------------------+
4291 | f | | Thread finishing. |
4292 +------+-----+-----------------------------------------------------------+
4293 | E | | Thread exited, not reaped by main thread yet. |
4294 +------+-----+-----------------------------------------------------------+
4295 | _ | | Thread reaped. |
4296 +------+-----+-----------------------------------------------------------+
4297 | X | | Thread reaped, exited with an error. |
4298 +------+-----+-----------------------------------------------------------+
4299 | K | | Thread reaped, exited due to signal. |
4300 +------+-----+-----------------------------------------------------------+
4303 Example output was based on the following:
4304 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --runtime=58m \
4305 --time_based --rate=2512k --bs=256K --numjobs=10 \
4306 --name=readers --rw=read --name=writers --rw=write
4308 Fio will condense the thread string as not to take up more space on the command
4309 line than needed. For instance, if you have 10 readers and 10 writers running,
4310 the output would look like this::
4312 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]
4314 Note that the status string is displayed in order, so it's possible to tell which of
4315 the jobs are currently doing what. In the example above this means that jobs 1--10
4316 are readers and 11--20 are writers.
4318 The other values are fairly self explanatory -- number of threads currently
4319 running and doing I/O, the number of currently open files (f=), the estimated
4320 completion percentage, the rate of I/O since last check (read speed listed first,
4321 then write speed and optionally trim speed) in terms of bandwidth and IOPS,
4322 and time to completion for the current running group. It's impossible to estimate
4323 runtime of the following groups (if any).
4326 Example output was based on the following:
4327 TZ=UTC fio --iodepth=16 --ioengine=posixaio --filename=/tmp/fiofile \
4328 --direct=1 --size=100M --time_based --runtime=50s --rate_iops=89 \
4329 --bs=7K --name=Client1 --rw=write
4331 When fio is done (or interrupted by :kbd:`Ctrl-C`), it will show the data for
4332 each thread, group of threads, and disks in that order. For each overall thread (or
4333 group) the output looks like::
4335 Client1: (groupid=0, jobs=1): err= 0: pid=16109: Sat Jun 24 12:07:54 2017
4336 write: IOPS=88, BW=623KiB/s (638kB/s)(30.4MiB/50032msec)
4337 slat (nsec): min=500, max=145500, avg=8318.00, stdev=4781.50
4338 clat (usec): min=170, max=78367, avg=4019.02, stdev=8293.31
4339 lat (usec): min=174, max=78375, avg=4027.34, stdev=8291.79
4340 clat percentiles (usec):
4341 | 1.00th=[ 302], 5.00th=[ 326], 10.00th=[ 343], 20.00th=[ 363],
4342 | 30.00th=[ 392], 40.00th=[ 404], 50.00th=[ 416], 60.00th=[ 445],
4343 | 70.00th=[ 816], 80.00th=[ 6718], 90.00th=[12911], 95.00th=[21627],
4344 | 99.00th=[43779], 99.50th=[51643], 99.90th=[68682], 99.95th=[72877],
4346 bw ( KiB/s): min= 532, max= 686, per=0.10%, avg=622.87, stdev=24.82, samples= 100
4347 iops : min= 76, max= 98, avg=88.98, stdev= 3.54, samples= 100
4348 lat (usec) : 250=0.04%, 500=64.11%, 750=4.81%, 1000=2.79%
4349 lat (msec) : 2=4.16%, 4=1.84%, 10=4.90%, 20=11.33%, 50=5.37%
4350 lat (msec) : 100=0.65%
4351 cpu : usr=0.27%, sys=0.18%, ctx=12072, majf=0, minf=21
4352 IO depths : 1=85.0%, 2=13.1%, 4=1.8%, 8=0.1%, 16=0.0%, 32=0.0%, >=64=0.0%
4353 submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
4354 complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
4355 issued rwt: total=0,4450,0, short=0,0,0, dropped=0,0,0
4356 latency : target=0, window=0, percentile=100.00%, depth=8
4358 The job name (or first job's name when using :option:`group_reporting`) is printed,
4359 along with the group id, count of jobs being aggregated, last error id seen (which
4360 is 0 when there are no errors), pid/tid of that thread and the time the job/group
4361 completed. Below are the I/O statistics for each data direction performed (showing
4362 writes in the example above). In the order listed, they denote:
4365 The string before the colon shows the I/O direction the statistics
4366 are for. **IOPS** is the average I/Os performed per second. **BW**
4367 is the average bandwidth rate shown as: value in power of 2 format
4368 (value in power of 10 format). The last two values show: (**total
4369 I/O performed** in power of 2 format / **runtime** of that thread).
4372 Submission latency (**min** being the minimum, **max** being the
4373 maximum, **avg** being the average, **stdev** being the standard
4374 deviation). This is the time from when fio initialized the I/O
4375 to submission. For synchronous ioengines this includes the time
4376 up until just before the ioengine's queue function is called.
4377 For asynchronous ioengines this includes the time up through the
4378 completion of the ioengine's queue function (and commit function
4379 if it is defined). For sync I/O this row is not displayed as the
4380 slat is negligible. This value can be in nanoseconds,
4381 microseconds or milliseconds --- fio will choose the most
4382 appropriate base and print that (in the example above
4383 nanoseconds was the best scale). Note: in :option:`--minimal`
4384 mode latencies are always expressed in microseconds.
4387 Completion latency. Same names as slat, this denotes the time from
4388 submission to completion of the I/O pieces. For sync I/O, this
4389 represents the time from when the I/O was submitted to the
4390 operating system to when it was completed. For asynchronous
4391 ioengines this is the time from when the ioengine's queue (and
4392 commit if available) functions were completed to when the I/O's
4393 completion was reaped by fio.
4396 Total latency. Same names as slat and clat, this denotes the time from
4397 when fio created the I/O unit to completion of the I/O operation.
4398 It is the sum of submission and completion latency.
4401 Bandwidth statistics based on samples. Same names as the xlat stats,
4402 but also includes the number of samples taken (**samples**) and an
4403 approximate percentage of total aggregate bandwidth this thread
4404 received in its group (**per**). This last value is only really
4405 useful if the threads in this group are on the same disk, since they
4406 are then competing for disk access.
4409 IOPS statistics based on samples. Same names as bw.
4411 **lat (nsec/usec/msec)**
4412 The distribution of I/O completion latencies. This is the time from when
4413 I/O leaves fio and when it gets completed. Unlike the separate
4414 read/write/trim sections above, the data here and in the remaining
4415 sections apply to all I/Os for the reporting group. 250=0.04% means that
4416 0.04% of the I/Os completed in under 250us. 500=64.11% means that 64.11%
4417 of the I/Os required 250 to 499us for completion.
4420 CPU usage. User and system time, along with the number of context
4421 switches this thread went through, usage of system and user time, and
4422 finally the number of major and minor page faults. The CPU utilization
4423 numbers are averages for the jobs in that reporting group, while the
4424 context and fault counters are summed.
4427 The distribution of I/O depths over the job lifetime. The numbers are
4428 divided into powers of 2 and each entry covers depths from that value
4429 up to those that are lower than the next entry -- e.g., 16= covers
4430 depths from 16 to 31. Note that the range covered by a depth
4431 distribution entry can be different to the range covered by the
4432 equivalent submit/complete distribution entry.
4435 How many pieces of I/O were submitting in a single submit call. Each
4436 entry denotes that amount and below, until the previous entry -- e.g.,
4437 16=100% means that we submitted anywhere between 9 to 16 I/Os per submit
4438 call. Note that the range covered by a submit distribution entry can
4439 be different to the range covered by the equivalent depth distribution
4443 Like the above submit number, but for completions instead.
4446 The number of read/write/trim requests issued, and how many of them were
4450 These values are for :option:`latency_target` and related options. When
4451 these options are engaged, this section describes the I/O depth required
4452 to meet the specified latency target.
4455 Example output was based on the following:
4456 TZ=UTC fio --ioengine=null --iodepth=2 --size=100M --numjobs=2 \
4457 --rate_process=poisson --io_limit=32M --name=read --bs=128k \
4458 --rate=11M --name=write --rw=write --bs=2k --rate=700k
4460 After each client has been listed, the group statistics are printed. They
4461 will look like this::
4463 Run status group 0 (all jobs):
4464 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
4465 WRITE: bw=1231KiB/s (1261kB/s), 616KiB/s-621KiB/s (630kB/s-636kB/s), io=64.0MiB (67.1MB), run=52747-53223msec
4467 For each data direction it prints:
4470 Aggregate bandwidth of threads in this group followed by the
4471 minimum and maximum bandwidth of all the threads in this group.
4472 Values outside of brackets are power-of-2 format and those
4473 within are the equivalent value in a power-of-10 format.
4475 Aggregate I/O performed of all threads in this group. The
4476 format is the same as bw.
4478 The smallest and longest runtimes of the threads in this group.
4480 And finally, the disk statistics are printed. This is Linux specific. They will look like this::
4482 Disk stats (read/write):
4483 sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00%
4485 Each value is printed for both reads and writes, with reads first. The
4489 Number of I/Os performed by all groups.
4491 Number of merges performed by the I/O scheduler.
4493 Number of ticks we kept the disk busy.
4495 Total time spent in the disk queue.
4497 The disk utilization. A value of 100% means we kept the disk
4498 busy constantly, 50% would be a disk idling half of the time.
4500 It is also possible to get fio to dump the current output while it is running,
4501 without terminating the job. To do that, send fio the **USR1** signal. You can
4502 also get regularly timed dumps by using the :option:`--status-interval`
4503 parameter, or by creating a file in :file:`/tmp` named
4504 :file:`fio-dump-status`. If fio sees this file, it will unlink it and dump the
4505 current output status.
4511 For scripted usage where you typically want to generate tables or graphs of the
4512 results, fio can output the results in a semicolon separated format. The format
4513 is one long line of values, such as::
4515 2;card0;0;0;7139336;121836;60004;1;10109;27.932460;116.933948;220;126861;3495.446807;1085.368601;226;126864;3523.635629;1089.012448;24063;99944;50.275485%;59818.274627;5540.657370;7155060;122104;60004;1;8338;29.086342;117.839068;388;128077;5032.488518;1234.785715;391;128085;5061.839412;1236.909129;23436;100928;50.287926%;59964.832030;5644.844189;14.595833%;19.394167%;123706;0;7313;0.1%;0.1%;0.1%;0.1%;0.1%;0.1%;100.0%;0.00%;0.00%;0.00%;0.00%;0.00%;0.00%;0.01%;0.02%;0.05%;0.16%;6.04%;40.40%;52.68%;0.64%;0.01%;0.00%;0.01%;0.00%;0.00%;0.00%;0.00%;0.00%
4516 A description of this job goes here.
4518 The job description (if provided) follows on a second line for terse v2.
4519 It appears on the same line for other terse versions.
4521 To enable terse output, use the :option:`--minimal` or
4522 :option:`--output-format`\=terse command line options. The
4523 first value is the version of the terse output format. If the output has to be
4524 changed for some reason, this number will be incremented by 1 to signify that
4527 Split up, the format is as follows (comments in brackets denote when a
4528 field was introduced or whether it's specific to some terse version):
4532 terse version, fio version [v3], jobname, groupid, error
4536 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
4537 Submission latency: min, max, mean, stdev (usec)
4538 Completion latency: min, max, mean, stdev (usec)
4539 Completion latency percentiles: 20 fields (see below)
4540 Total latency: min, max, mean, stdev (usec)
4541 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
4542 IOPS [v5]: min, max, mean, stdev, number of samples
4548 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
4549 Submission latency: min, max, mean, stdev (usec)
4550 Completion latency: min, max, mean, stdev (usec)
4551 Completion latency percentiles: 20 fields (see below)
4552 Total latency: min, max, mean, stdev (usec)
4553 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
4554 IOPS [v5]: min, max, mean, stdev, number of samples
4556 TRIM status [all but version 3]:
4558 Fields are similar to READ/WRITE status.
4562 user, system, context switches, major faults, minor faults
4566 <=1, 2, 4, 8, 16, 32, >=64
4568 I/O latencies microseconds::
4570 <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
4572 I/O latencies milliseconds::
4574 <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
4576 Disk utilization [v3]::
4578 disk name, read ios, write ios, read merges, write merges, read ticks, write ticks,
4579 time spent in queue, disk utilization percentage
4581 Additional Info (dependent on continue_on_error, default off)::
4583 total # errors, first error code
4585 Additional Info (dependent on description being set)::
4589 Completion latency percentiles can be a grouping of up to 20 sets, so for the
4590 terse output fio writes all of them. Each field will look like this::
4594 which is the Xth percentile, and the `usec` latency associated with it.
4596 For `Disk utilization`, all disks used by fio are shown. So for each disk there
4597 will be a disk utilization section.
4599 Below is a single line containing short names for each of the fields in the
4600 minimal output v3, separated by semicolons::
4602 terse_version_3;fio_version;jobname;groupid;error;read_kb;read_bandwidth_kb;read_iops;read_runtime_ms;read_slat_min_us;read_slat_max_us;read_slat_mean_us;read_slat_dev_us;read_clat_min_us;read_clat_max_us;read_clat_mean_us;read_clat_dev_us;read_clat_pct01;read_clat_pct02;read_clat_pct03;read_clat_pct04;read_clat_pct05;read_clat_pct06;read_clat_pct07;read_clat_pct08;read_clat_pct09;read_clat_pct10;read_clat_pct11;read_clat_pct12;read_clat_pct13;read_clat_pct14;read_clat_pct15;read_clat_pct16;read_clat_pct17;read_clat_pct18;read_clat_pct19;read_clat_pct20;read_tlat_min_us;read_lat_max_us;read_lat_mean_us;read_lat_dev_us;read_bw_min_kb;read_bw_max_kb;read_bw_agg_pct;read_bw_mean_kb;read_bw_dev_kb;write_kb;write_bandwidth_kb;write_iops;write_runtime_ms;write_slat_min_us;write_slat_max_us;write_slat_mean_us;write_slat_dev_us;write_clat_min_us;write_clat_max_us;write_clat_mean_us;write_clat_dev_us;write_clat_pct01;write_clat_pct02;write_clat_pct03;write_clat_pct04;write_clat_pct05;write_clat_pct06;write_clat_pct07;write_clat_pct08;write_clat_pct09;write_clat_pct10;write_clat_pct11;write_clat_pct12;write_clat_pct13;write_clat_pct14;write_clat_pct15;write_clat_pct16;write_clat_pct17;write_clat_pct18;write_clat_pct19;write_clat_pct20;write_tlat_min_us;write_lat_max_us;write_lat_mean_us;write_lat_dev_us;write_bw_min_kb;write_bw_max_kb;write_bw_agg_pct;write_bw_mean_kb;write_bw_dev_kb;cpu_user;cpu_sys;cpu_csw;cpu_mjf;cpu_minf;iodepth_1;iodepth_2;iodepth_4;iodepth_8;iodepth_16;iodepth_32;iodepth_64;lat_2us;lat_4us;lat_10us;lat_20us;lat_50us;lat_100us;lat_250us;lat_500us;lat_750us;lat_1000us;lat_2ms;lat_4ms;lat_10ms;lat_20ms;lat_50ms;lat_100ms;lat_250ms;lat_500ms;lat_750ms;lat_1000ms;lat_2000ms;lat_over_2000ms;disk_name;disk_read_iops;disk_write_iops;disk_read_merges;disk_write_merges;disk_read_ticks;write_ticks;disk_queue_time;disk_util
4604 In client/server mode terse output differs from what appears when jobs are run
4605 locally. Disk utilization data is omitted from the standard terse output and
4606 for v3 and later appears on its own separate line at the end of each terse
4613 The `json` output format is intended to be both human readable and convenient
4614 for automated parsing. For the most part its sections mirror those of the
4615 `normal` output. The `runtime` value is reported in msec and the `bw` value is
4616 reported in 1024 bytes per second units.
4622 The `json+` output format is identical to the `json` output format except that it
4623 adds a full dump of the completion latency bins. Each `bins` object contains a
4624 set of (key, value) pairs where keys are latency durations and values count how
4625 many I/Os had completion latencies of the corresponding duration. For example,
4628 "bins" : { "87552" : 1, "89600" : 1, "94720" : 1, "96768" : 1, "97792" : 1, "99840" : 1, "100864" : 2, "103936" : 6, "104960" : 534, "105984" : 5995, "107008" : 7529, ... }
4630 This data indicates that one I/O required 87,552ns to complete, two I/Os required
4631 100,864ns to complete, and 7529 I/Os required 107,008ns to complete.
4633 Also included with fio is a Python script `fio_jsonplus_clat2csv` that takes
4634 json+ output and generates CSV-formatted latency data suitable for plotting.
4636 The latency durations actually represent the midpoints of latency intervals.
4637 For details refer to :file:`stat.h`.
4643 There are two trace file format that you can encounter. The older (v1) format is
4644 unsupported since version 1.20-rc3 (March 2008). It will still be described
4645 below in case that you get an old trace and want to understand it.
4647 In any case the trace is a simple text file with a single action per line.
4650 Trace file format v1
4651 ~~~~~~~~~~~~~~~~~~~~
4653 Each line represents a single I/O action in the following format::
4657 where `rw=0/1` for read/write, and the `offset` and `length` entries being in bytes.
4659 This format is not supported in fio versions >= 1.20-rc3.
4662 Trace file format v2
4663 ~~~~~~~~~~~~~~~~~~~~
4665 The second version of the trace file format was added in fio version 1.17. It
4666 allows one to access more than one file per trace and has a bigger set of possible
4669 The first line of the trace file has to be::
4673 Following this can be lines in two different formats, which are described below.
4675 The file management format::
4679 The `filename` is given as an absolute path. The `action` can be one of these:
4682 Add the given `filename` to the trace.
4684 Open the file with the given `filename`. The `filename` has to have
4685 been added with the **add** action before.
4687 Close the file with the given `filename`. The file has to have been
4691 The file I/O action format::
4693 filename action offset length
4695 The `filename` is given as an absolute path, and has to have been added and
4696 opened before it can be used with this format. The `offset` and `length` are
4697 given in bytes. The `action` can be one of these:
4700 Wait for `offset` microseconds. Everything below 100 is discarded.
4701 The time is relative to the previous `wait` statement. Note that
4702 action `wait` is not allowed as of version 3, as the same behavior
4703 can be achieved using timestamps.
4705 Read `length` bytes beginning from `offset`.
4707 Write `length` bytes beginning from `offset`.
4709 :manpage:`fsync(2)` the file.
4711 :manpage:`fdatasync(2)` the file.
4713 Trim the given file from the given `offset` for `length` bytes.
4716 Trace file format v3
4717 ~~~~~~~~~~~~~~~~~~~~
4719 The third version of the trace file format was added in fio version 3.31. It
4720 forces each action to have a timestamp associated with it.
4722 The first line of the trace file has to be::
4726 Following this can be lines in two different formats, which are described below.
4728 The file management format::
4730 timestamp filename action
4732 The file I/O action format::
4734 timestamp filename action offset length
4736 The `timestamp` is relative to the beginning of the run (ie starts at 0). The
4737 `filename`, `action`, `offset` and `length` are identical to version 2, except
4738 that version 3 does not allow the `wait` action.
4741 I/O Replay - Merging Traces
4742 ---------------------------
4744 Colocation is a common practice used to get the most out of a machine.
4745 Knowing which workloads play nicely with each other and which ones don't is
4746 a much harder task. While fio can replay workloads concurrently via multiple
4747 jobs, it leaves some variability up to the scheduler making results harder to
4748 reproduce. Merging is a way to make the order of events consistent.
4750 Merging is integrated into I/O replay and done when a
4751 :option:`merge_blktrace_file` is specified. The list of files passed to
4752 :option:`read_iolog` go through the merge process and output a single file
4753 stored to the specified file. The output file is passed on as if it were the
4754 only file passed to :option:`read_iolog`. An example would look like::
4756 $ fio --read_iolog="<file1>:<file2>" --merge_blktrace_file="<output_file>"
4758 Creating only the merged file can be done by passing the command line argument
4759 :option:`--merge-blktrace-only`.
4761 Scaling traces can be done to see the relative impact of any particular trace
4762 being slowed down or sped up. :option:`merge_blktrace_scalars` takes in a colon
4763 separated list of percentage scalars. It is index paired with the files passed
4764 to :option:`read_iolog`.
4766 With scaling, it may be desirable to match the running time of all traces.
4767 This can be done with :option:`merge_blktrace_iters`. It is index paired with
4768 :option:`read_iolog` just like :option:`merge_blktrace_scalars`.
4770 In an example, given two traces, A and B, each 60s long. If we want to see
4771 the impact of trace A issuing IOs twice as fast and repeat trace A over the
4772 runtime of trace B, the following can be done::
4774 $ fio --read_iolog="<trace_a>:"<trace_b>" --merge_blktrace_file"<output_file>" --merge_blktrace_scalars="50:100" --merge_blktrace_iters="2:1"
4776 This runs trace A at 2x the speed twice for approximately the same runtime as
4777 a single run of trace B.
4780 CPU idleness profiling
4781 ----------------------
4783 In some cases, we want to understand CPU overhead in a test. For example, we
4784 test patches for the specific goodness of whether they reduce CPU usage.
4785 Fio implements a balloon approach to create a thread per CPU that runs at idle
4786 priority, meaning that it only runs when nobody else needs the cpu.
4787 By measuring the amount of work completed by the thread, idleness of each CPU
4788 can be derived accordingly.
4790 An unit work is defined as touching a full page of unsigned characters. Mean and
4791 standard deviation of time to complete an unit work is reported in "unit work"
4792 section. Options can be chosen to report detailed percpu idleness or overall
4793 system idleness by aggregating percpu stats.
4796 Verification and triggers
4797 -------------------------
4799 Fio is usually run in one of two ways, when data verification is done. The first
4800 is a normal write job of some sort with verify enabled. When the write phase has
4801 completed, fio switches to reads and verifies everything it wrote. The second
4802 model is running just the write phase, and then later on running the same job
4803 (but with reads instead of writes) to repeat the same I/O patterns and verify
4804 the contents. Both of these methods depend on the write phase being completed,
4805 as fio otherwise has no idea how much data was written.
4807 With verification triggers, fio supports dumping the current write state to
4808 local files. Then a subsequent read verify workload can load this state and know
4809 exactly where to stop. This is useful for testing cases where power is cut to a
4810 server in a managed fashion, for instance.
4812 A verification trigger consists of two things:
4814 1) Storing the write state of each job.
4815 2) Executing a trigger command.
4817 The write state is relatively small, on the order of hundreds of bytes to single
4818 kilobytes. It contains information on the number of completions done, the last X
4821 A trigger is invoked either through creation ('touch') of a specified file in
4822 the system, or through a timeout setting. If fio is run with
4823 :option:`--trigger-file`\= :file:`/tmp/trigger-file`, then it will continually
4824 check for the existence of :file:`/tmp/trigger-file`. When it sees this file, it
4825 will fire off the trigger (thus saving state, and executing the trigger
4828 For client/server runs, there's both a local and remote trigger. If fio is
4829 running as a server backend, it will send the job states back to the client for
4830 safe storage, then execute the remote trigger, if specified. If a local trigger
4831 is specified, the server will still send back the write state, but the client
4832 will then execute the trigger.
4834 Verification trigger example
4835 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4837 Let's say we want to run a powercut test on the remote Linux machine 'server'.
4838 Our write workload is in :file:`write-test.fio`. We want to cut power to 'server' at
4839 some point during the run, and we'll run this test from the safety or our local
4840 machine, 'localbox'. On the server, we'll start the fio backend normally::
4842 server# fio --server
4844 and on the client, we'll fire off the workload::
4846 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger-remote="bash -c \"echo b > /proc/sysrq-triger\""
4848 We set :file:`/tmp/my-trigger` as the trigger file, and we tell fio to execute::
4850 echo b > /proc/sysrq-trigger
4852 on the server once it has received the trigger and sent us the write state. This
4853 will work, but it's not **really** cutting power to the server, it's merely
4854 abruptly rebooting it. If we have a remote way of cutting power to the server
4855 through IPMI or similar, we could do that through a local trigger command
4856 instead. Let's assume we have a script that does IPMI reboot of a given hostname,
4857 ipmi-reboot. On localbox, we could then have run fio with a local trigger
4860 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger="ipmi-reboot server"
4862 For this case, fio would wait for the server to send us the write state, then
4863 execute ``ipmi-reboot server`` when that happened.
4865 Loading verify state
4866 ~~~~~~~~~~~~~~~~~~~~
4868 To load stored write state, a read verification job file must contain the
4869 :option:`verify_state_load` option. If that is set, fio will load the previously
4870 stored state. For a local fio run this is done by loading the files directly,
4871 and on a client/server run, the server backend will ask the client to send the
4872 files over and load them from there.
4878 Fio supports a variety of log file formats, for logging latencies, bandwidth,
4879 and IOPS. The logs share a common format, which looks like this:
4881 *time* (`msec`), *value*, *data direction*, *block size* (`bytes`),
4882 *offset* (`bytes`), *command priority*
4884 *Time* for the log entry is always in milliseconds. The *value* logged depends
4885 on the type of log, it will be one of the following:
4888 Value is latency in nsecs
4894 *Data direction* is one of the following:
4903 The entry's *block size* is always in bytes. The *offset* is the position in bytes
4904 from the start of the file for that particular I/O. The logging of the offset can be
4905 toggled with :option:`log_offset`.
4907 *Command priority* is 0 for normal priority and 1 for high priority. This is controlled
4908 by the ioengine specific :option:`cmdprio_percentage`.
4910 Fio defaults to logging every individual I/O but when windowed logging is set
4911 through :option:`log_avg_msec`, either the average (by default) or the maximum
4912 (:option:`log_max_value` is set) *value* seen over the specified period of time
4913 is recorded. Each *data direction* seen within the window period will aggregate
4914 its values in a separate row. Further, when using windowed logging the *block
4915 size* and *offset* entries will always contain 0.
4921 Normally fio is invoked as a stand-alone application on the machine where the
4922 I/O workload should be generated. However, the backend and frontend of fio can
4923 be run separately i.e., the fio server can generate an I/O workload on the "Device
4924 Under Test" while being controlled by a client on another machine.
4926 Start the server on the machine which has access to the storage DUT::
4930 where `args` defines what fio listens to. The arguments are of the form
4931 ``type,hostname`` or ``IP,port``. *type* is either ``ip`` (or ip4) for TCP/IP
4932 v4, ``ip6`` for TCP/IP v6, or ``sock`` for a local unix domain socket.
4933 *hostname* is either a hostname or IP address, and *port* is the port to listen
4934 to (only valid for TCP/IP, not a local socket). Some examples:
4938 Start a fio server, listening on all interfaces on the default port (8765).
4940 2) ``fio --server=ip:hostname,4444``
4942 Start a fio server, listening on IP belonging to hostname and on port 4444.
4944 3) ``fio --server=ip6:::1,4444``
4946 Start a fio server, listening on IPv6 localhost ::1 and on port 4444.
4948 4) ``fio --server=,4444``
4950 Start a fio server, listening on all interfaces on port 4444.
4952 5) ``fio --server=1.2.3.4``
4954 Start a fio server, listening on IP 1.2.3.4 on the default port.
4956 6) ``fio --server=sock:/tmp/fio.sock``
4958 Start a fio server, listening on the local socket :file:`/tmp/fio.sock`.
4960 Once a server is running, a "client" can connect to the fio server with::
4962 fio <local-args> --client=<server> <remote-args> <job file(s)>
4964 where `local-args` are arguments for the client where it is running, `server`
4965 is the connect string, and `remote-args` and `job file(s)` are sent to the
4966 server. The `server` string follows the same format as it does on the server
4967 side, to allow IP/hostname/socket and port strings.
4969 Fio can connect to multiple servers this way::
4971 fio --client=<server1> <job file(s)> --client=<server2> <job file(s)>
4973 If the job file is located on the fio server, then you can tell the server to
4974 load a local file as well. This is done by using :option:`--remote-config` ::
4976 fio --client=server --remote-config /path/to/file.fio
4978 Then fio will open this local (to the server) job file instead of being passed
4979 one from the client.
4981 If you have many servers (example: 100 VMs/containers), you can input a pathname
4982 of a file containing host IPs/names as the parameter value for the
4983 :option:`--client` option. For example, here is an example :file:`host.list`
4984 file containing 2 hostnames::
4986 host1.your.dns.domain
4987 host2.your.dns.domain
4989 The fio command would then be::
4991 fio --client=host.list <job file(s)>
4993 In this mode, you cannot input server-specific parameters or job files -- all
4994 servers receive the same job file.
4996 In order to let ``fio --client`` runs use a shared filesystem from multiple
4997 hosts, ``fio --client`` now prepends the IP address of the server to the
4998 filename. For example, if fio is using the directory :file:`/mnt/nfs/fio` and is
4999 writing filename :file:`fileio.tmp`, with a :option:`--client` `hostfile`
5000 containing two hostnames ``h1`` and ``h2`` with IP addresses 192.168.10.120 and
5001 192.168.10.121, then fio will create two files::
5003 /mnt/nfs/fio/192.168.10.120.fileio.tmp
5004 /mnt/nfs/fio/192.168.10.121.fileio.tmp
5006 Terse output in client/server mode will differ slightly from what is produced
5007 when fio is run in stand-alone mode. See the terse output section for details.