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 Show available debug options.
99 .. option:: --parse-only
101 Parse options only, don't start any I/O.
103 .. option:: --output=filename
105 Write output to file `filename`.
107 .. option:: --output-format=format
109 Set the reporting `format` to `normal`, `terse`, `json`, or `json+`. Multiple
110 formats can be selected, separated by a comma. `terse` is a CSV based
111 format. `json+` is like `json`, except it adds a full dump of the latency
114 .. option:: --bandwidth-log
116 Generate aggregate bandwidth logs.
118 .. option:: --minimal
120 Print statistics in a terse, semicolon-delimited format.
122 .. option:: --append-terse
124 Print statistics in selected mode AND terse, semicolon-delimited format.
125 **Deprecated**, use :option:`--output-format` instead to select multiple
128 .. option:: --terse-version=version
130 Set terse `version` output format (default 3, or 2 or 4 or 5).
132 .. option:: --version
134 Print version information and exit.
138 Print a summary of the command line options and exit.
140 .. option:: --cpuclock-test
142 Perform test and validation of internal CPU clock.
144 .. option:: --crctest=[test]
146 Test the speed of the built-in checksumming functions. If no argument is
147 given, all of them are tested. Alternatively, a comma separated list can
148 be passed, in which case the given ones are tested.
150 .. option:: --cmdhelp=command
152 Print help information for `command`. May be ``all`` for all commands.
154 .. option:: --enghelp=[ioengine[,command]]
156 List all commands defined by `ioengine`, or print help for `command`
157 defined by `ioengine`. If no `ioengine` is given, list all
160 .. option:: --showcmd=jobfile
162 Convert `jobfile` to a set of command-line options.
164 .. option:: --readonly
166 Turn on safety read-only checks, preventing writes. The ``--readonly``
167 option is an extra safety guard to prevent users from accidentally starting
168 a write workload when that is not desired. Fio will only write if
169 `rw=write/randwrite/rw/randrw` is given. This extra safety net can be used
170 as an extra precaution as ``--readonly`` will also enable a write check in
171 the I/O engine core to prevent writes due to unknown user space bug(s).
173 .. option:: --eta=when
175 Specifies when real-time ETA estimate should be printed. `when` may be
176 `always`, `never` or `auto`. `auto` is the default, it prints ETA
177 when requested if the output is a TTY. `always` disregards the output
178 type, and prints ETA when requested. `never` never prints ETA.
180 .. option:: --eta-interval=time
182 By default, fio requests client ETA status roughly every second. With
183 this option, the interval is configurable. Fio imposes a minimum
184 allowed time to avoid flooding the console, less than 250 msec is
187 .. option:: --eta-newline=time
189 Force a new line for every `time` period passed. When the unit is omitted,
190 the value is interpreted in seconds.
192 .. option:: --status-interval=time
194 Force a full status dump of cumulative (from job start) values at `time`
195 intervals. This option does *not* provide per-period measurements. So
196 values such as bandwidth are running averages. When the time unit is omitted,
197 `time` is interpreted in seconds.
199 .. option:: --section=name
201 Only run specified section `name` in job file. Multiple sections can be specified.
202 The ``--section`` option allows one to combine related jobs into one file.
203 E.g. one job file could define light, moderate, and heavy sections. Tell
204 fio to run only the "heavy" section by giving ``--section=heavy``
205 command line option. One can also specify the "write" operations in one
206 section and "verify" operation in another section. The ``--section`` option
207 only applies to job sections. The reserved *global* section is always
210 .. option:: --alloc-size=kb
212 Set the internal smalloc pool size to `kb` in KiB. The
213 ``--alloc-size`` switch allows one to use a larger pool size for smalloc.
214 If running large jobs with randommap enabled, fio can run out of memory.
215 Smalloc is an internal allocator for shared structures from a fixed size
216 memory pool and can grow to 16 pools. The pool size defaults to 16MiB.
218 NOTE: While running :file:`.fio_smalloc.*` backing store files are visible
221 .. option:: --warnings-fatal
223 All fio parser warnings are fatal, causing fio to exit with an
226 .. option:: --max-jobs=nr
228 Set the maximum number of threads/processes to support to `nr`.
229 NOTE: On Linux, it may be necessary to increase the shared-memory
230 limit (:file:`/proc/sys/kernel/shmmax`) if fio runs into errors while
233 .. option:: --server=args
235 Start a backend server, with `args` specifying what to listen to.
236 See `Client/Server`_ section.
238 .. option:: --daemonize=pidfile
240 Background a fio server, writing the pid to the given `pidfile` file.
242 .. option:: --client=hostname
244 Instead of running the jobs locally, send and run them on the given `hostname`
245 or set of `hostname`\s. See `Client/Server`_ section.
247 .. option:: --remote-config=file
249 Tell fio server to load this local `file`.
251 .. option:: --idle-prof=option
253 Report CPU idleness. `option` is one of the following:
256 Run unit work calibration only and exit.
259 Show aggregate system idleness and unit work.
262 As **system** but also show per CPU idleness.
264 .. option:: --inflate-log=log
266 Inflate and output compressed `log`.
268 .. option:: --trigger-file=file
270 Execute trigger command when `file` exists.
272 .. option:: --trigger-timeout=time
274 Execute trigger at this `time`.
276 .. option:: --trigger=command
278 Set this `command` as local trigger.
280 .. option:: --trigger-remote=command
282 Set this `command` as remote trigger.
284 .. option:: --aux-path=path
286 Use this `path` for fio state generated files.
288 Any parameters following the options will be assumed to be job files, unless
289 they match a job file parameter. Multiple job files can be listed and each job
290 file will be regarded as a separate group. Fio will :option:`stonewall`
291 execution between each group.
297 As previously described, fio accepts one or more job files describing what it is
298 supposed to do. The job file format is the classic ini file, where the names
299 enclosed in [] brackets define the job name. You are free to use any ASCII name
300 you want, except *global* which has special meaning. Following the job name is
301 a sequence of zero or more parameters, one per line, that define the behavior of
302 the job. If the first character in a line is a ';' or a '#', the entire line is
303 discarded as a comment.
305 A *global* section sets defaults for the jobs described in that file. A job may
306 override a *global* section parameter, and a job file may even have several
307 *global* sections if so desired. A job is only affected by a *global* section
310 The :option:`--cmdhelp` option also lists all options. If used with a `command`
311 argument, :option:`--cmdhelp` will detail the given `command`.
313 See the `examples/` directory for inspiration on how to write job files. Note
314 the copyright and license requirements currently apply to `examples/` files.
316 So let's look at a really simple job file that defines two processes, each
317 randomly reading from a 128MiB file:
321 ; -- start job file --
332 As you can see, the job file sections themselves are empty as all the described
333 parameters are shared. As no :option:`filename` option is given, fio makes up a
334 `filename` for each of the jobs as it sees fit. On the command line, this job
335 would look as follows::
337 $ fio --name=global --rw=randread --size=128m --name=job1 --name=job2
340 Let's look at an example that has a number of processes writing randomly to
345 ; -- start job file --
356 Here we have no *global* section, as we only have one job defined anyway. We
357 want to use async I/O here, with a depth of 4 for each file. We also increased
358 the buffer size used to 32KiB and define numjobs to 4 to fork 4 identical
359 jobs. The result is 4 processes each randomly writing to their own 64MiB
360 file. Instead of using the above job file, you could have given the parameters
361 on the command line. For this case, you would specify::
363 $ fio --name=random-writers --ioengine=libaio --iodepth=4 --rw=randwrite --bs=32k --direct=0 --size=64m --numjobs=4
365 When fio is utilized as a basis of any reasonably large test suite, it might be
366 desirable to share a set of standardized settings across multiple job files.
367 Instead of copy/pasting such settings, any section may pull in an external
368 :file:`filename.fio` file with *include filename* directive, as in the following
371 ; -- start job file including.fio --
375 include glob-include.fio
382 include test-include.fio
383 ; -- end job file including.fio --
387 ; -- start job file glob-include.fio --
390 ; -- end job file glob-include.fio --
394 ; -- start job file test-include.fio --
397 ; -- end job file test-include.fio --
399 Settings pulled into a section apply to that section only (except *global*
400 section). Include directives may be nested in that any included file may contain
401 further include directive(s). Include files may not contain [] sections.
404 Environment variables
405 ~~~~~~~~~~~~~~~~~~~~~
407 Fio also supports environment variable expansion in job files. Any sub-string of
408 the form ``${VARNAME}`` as part of an option value (in other words, on the right
409 of the '='), will be expanded to the value of the environment variable called
410 `VARNAME`. If no such environment variable is defined, or `VARNAME` is the
411 empty string, the empty string will be substituted.
413 As an example, let's look at a sample fio invocation and job file::
415 $ SIZE=64m NUMJOBS=4 fio jobfile.fio
419 ; -- start job file --
426 This will expand to the following equivalent job file at runtime:
430 ; -- start job file --
437 Fio ships with a few example job files, you can also look there for inspiration.
442 Additionally, fio has a set of reserved keywords that will be replaced
443 internally with the appropriate value. Those keywords are:
447 The architecture page size of the running system.
451 Megabytes of total memory in the system.
455 Number of online available CPUs.
457 These can be used on the command line or in the job file, and will be
458 automatically substituted with the current system values when the job is
459 run. Simple math is also supported on these keywords, so you can perform actions
464 and get that properly expanded to 8 times the size of memory in the machine.
470 This section describes in details each parameter associated with a job. Some
471 parameters take an option of a given type, such as an integer or a
472 string. Anywhere a numeric value is required, an arithmetic expression may be
473 used, provided it is surrounded by parentheses. Supported operators are:
482 For time values in expressions, units are microseconds by default. This is
483 different than for time values not in expressions (not enclosed in
484 parentheses). The following types are used:
491 String: A sequence of alphanumeric characters.
494 Integer with possible time suffix. Without a unit value is interpreted as
495 seconds unless otherwise specified. Accepts a suffix of 'd' for days, 'h' for
496 hours, 'm' for minutes, 's' for seconds, 'ms' (or 'msec') for milliseconds and
497 'us' (or 'usec') for microseconds. For example, use 10m for 10 minutes.
502 Integer. A whole number value, which may contain an integer prefix
503 and an integer suffix:
505 [*integer prefix*] **number** [*integer suffix*]
507 The optional *integer prefix* specifies the number's base. The default
508 is decimal. *0x* specifies hexadecimal.
510 The optional *integer suffix* specifies the number's units, and includes an
511 optional unit prefix and an optional unit. For quantities of data, the
512 default unit is bytes. For quantities of time, the default unit is seconds
513 unless otherwise specified.
515 With :option:`kb_base`\=1000, fio follows international standards for unit
516 prefixes. To specify power-of-10 decimal values defined in the
517 International System of Units (SI):
519 * *K* -- means kilo (K) or 1000
520 * *M* -- means mega (M) or 1000**2
521 * *G* -- means giga (G) or 1000**3
522 * *T* -- means tera (T) or 1000**4
523 * *P* -- means peta (P) or 1000**5
525 To specify power-of-2 binary values defined in IEC 80000-13:
527 * *Ki* -- means kibi (Ki) or 1024
528 * *Mi* -- means mebi (Mi) or 1024**2
529 * *Gi* -- means gibi (Gi) or 1024**3
530 * *Ti* -- means tebi (Ti) or 1024**4
531 * *Pi* -- means pebi (Pi) or 1024**5
533 With :option:`kb_base`\=1024 (the default), the unit prefixes are opposite
534 from those specified in the SI and IEC 80000-13 standards to provide
535 compatibility with old scripts. For example, 4k means 4096.
537 For quantities of data, an optional unit of 'B' may be included
538 (e.g., 'kB' is the same as 'k').
540 The *integer suffix* is not case sensitive (e.g., m/mi mean mebi/mega,
541 not milli). 'b' and 'B' both mean byte, not bit.
543 Examples with :option:`kb_base`\=1000:
545 * *4 KiB*: 4096, 4096b, 4096B, 4ki, 4kib, 4kiB, 4Ki, 4KiB
546 * *1 MiB*: 1048576, 1mi, 1024ki
547 * *1 MB*: 1000000, 1m, 1000k
548 * *1 TiB*: 1099511627776, 1ti, 1024gi, 1048576mi
549 * *1 TB*: 1000000000, 1t, 1000m, 1000000k
551 Examples with :option:`kb_base`\=1024 (default):
553 * *4 KiB*: 4096, 4096b, 4096B, 4k, 4kb, 4kB, 4K, 4KB
554 * *1 MiB*: 1048576, 1m, 1024k
555 * *1 MB*: 1000000, 1mi, 1000ki
556 * *1 TiB*: 1099511627776, 1t, 1024g, 1048576m
557 * *1 TB*: 1000000000, 1ti, 1000mi, 1000000ki
559 To specify times (units are not case sensitive):
563 * *M* -- means minutes
564 * *s* -- or sec means seconds (default)
565 * *ms* -- or *msec* means milliseconds
566 * *us* -- or *usec* means microseconds
568 If the option accepts an upper and lower range, use a colon ':' or
569 minus '-' to separate such values. See :ref:`irange <irange>`.
570 If the lower value specified happens to be larger than the upper value
571 the two values are swapped.
576 Boolean. Usually parsed as an integer, however only defined for
577 true and false (1 and 0).
582 Integer range with suffix. Allows value range to be given, such as
583 1024-4096. A colon may also be used as the separator, e.g. 1k:4k. If the
584 option allows two sets of ranges, they can be specified with a ',' or '/'
585 delimiter: 1k-4k/8k-32k. Also see :ref:`int <int>`.
588 A list of floating point numbers, separated by a ':' character.
590 With the above in mind, here follows the complete list of fio job parameters.
596 .. option:: kb_base=int
598 Select the interpretation of unit prefixes in input parameters.
601 Inputs comply with IEC 80000-13 and the International
602 System of Units (SI). Use:
604 - power-of-2 values with IEC prefixes (e.g., KiB)
605 - power-of-10 values with SI prefixes (e.g., kB)
608 Compatibility mode (default). To avoid breaking old scripts:
610 - power-of-2 values with SI prefixes
611 - power-of-10 values with IEC prefixes
613 See :option:`bs` for more details on input parameters.
615 Outputs always use correct prefixes. Most outputs include both
618 bw=2383.3kB/s (2327.4KiB/s)
620 If only one value is reported, then kb_base selects the one to use:
622 **1000** -- SI prefixes
624 **1024** -- IEC prefixes
626 .. option:: unit_base=int
628 Base unit for reporting. Allowed values are:
631 Use auto-detection (default).
643 ASCII name of the job. This may be used to override the name printed by fio
644 for this job. Otherwise the job name is used. On the command line this
645 parameter has the special purpose of also signaling the start of a new job.
647 .. option:: description=str
649 Text description of the job. Doesn't do anything except dump this text
650 description when this job is run. It's not parsed.
652 .. option:: loops=int
654 Run the specified number of iterations of this job. Used to repeat the same
655 workload a given number of times. Defaults to 1.
657 .. option:: numjobs=int
659 Create the specified number of clones of this job. Each clone of job
660 is spawned as an independent thread or process. May be used to setup a
661 larger number of threads/processes doing the same thing. Each thread is
662 reported separately; to see statistics for all clones as a whole, use
663 :option:`group_reporting` in conjunction with :option:`new_group`.
664 See :option:`--max-jobs`. Default: 1.
667 Time related parameters
668 ~~~~~~~~~~~~~~~~~~~~~~~
670 .. option:: runtime=time
672 Tell fio to terminate processing after the specified period of time. It
673 can be quite hard to determine for how long a specified job will run, so
674 this parameter is handy to cap the total runtime to a given time. When
675 the unit is omitted, the value is intepreted in seconds.
677 .. option:: time_based
679 If set, fio will run for the duration of the :option:`runtime` specified
680 even if the file(s) are completely read or written. It will simply loop over
681 the same workload as many times as the :option:`runtime` allows.
683 .. option:: startdelay=irange(time)
685 Delay the start of job for the specified amount of time. Can be a single
686 value or a range. When given as a range, each thread will choose a value
687 randomly from within the range. Value is in seconds if a unit is omitted.
689 .. option:: ramp_time=time
691 If set, fio will run the specified workload for this amount of time before
692 logging any performance numbers. Useful for letting performance settle
693 before logging results, thus minimizing the runtime required for stable
694 results. Note that the ``ramp_time`` is considered lead in time for a job,
695 thus it will increase the total runtime if a special timeout or
696 :option:`runtime` is specified. When the unit is omitted, the value is
699 .. option:: clocksource=str
701 Use the given clocksource as the base of timing. The supported options are:
704 :manpage:`gettimeofday(2)`
707 :manpage:`clock_gettime(2)`
710 Internal CPU clock source
712 cpu is the preferred clocksource if it is reliable, as it is very fast (and
713 fio is heavy on time calls). Fio will automatically use this clocksource if
714 it's supported and considered reliable on the system it is running on,
715 unless another clocksource is specifically set. For x86/x86-64 CPUs, this
716 means supporting TSC Invariant.
718 .. option:: gtod_reduce=bool
720 Enable all of the :manpage:`gettimeofday(2)` reducing options
721 (:option:`disable_clat`, :option:`disable_slat`, :option:`disable_bw_measurement`) plus
722 reduce precision of the timeout somewhat to really shrink the
723 :manpage:`gettimeofday(2)` call count. With this option enabled, we only do
724 about 0.4% of the :manpage:`gettimeofday(2)` calls we would have done if all
725 time keeping was enabled.
727 .. option:: gtod_cpu=int
729 Sometimes it's cheaper to dedicate a single thread of execution to just
730 getting the current time. Fio (and databases, for instance) are very
731 intensive on :manpage:`gettimeofday(2)` calls. With this option, you can set
732 one CPU aside for doing nothing but logging current time to a shared memory
733 location. Then the other threads/processes that run I/O workloads need only
734 copy that segment, instead of entering the kernel with a
735 :manpage:`gettimeofday(2)` call. The CPU set aside for doing these time
736 calls will be excluded from other uses. Fio will manually clear it from the
737 CPU mask of other jobs.
743 .. option:: directory=str
745 Prefix filenames with this directory. Used to place files in a different
746 location than :file:`./`. You can specify a number of directories by
747 separating the names with a ':' character. These directories will be
748 assigned equally distributed to job clones created by :option:`numjobs` as
749 long as they are using generated filenames. If specific `filename(s)` are
750 set fio will use the first listed directory, and thereby matching the
751 `filename` semantic which generates a file each clone if not specified, but
752 let all clones use the same if set.
754 See the :option:`filename` option for information on how to escape "``:``" and
755 "``\``" characters within the directory path itself.
757 .. option:: filename=str
759 Fio normally makes up a `filename` based on the job name, thread number, and
760 file number (see :option:`filename_format`). If you want to share files
761 between threads in a job or several
762 jobs with fixed file paths, specify a `filename` for each of them to override
763 the default. If the ioengine is file based, you can specify a number of files
764 by separating the names with a ':' colon. So if you wanted a job to open
765 :file:`/dev/sda` and :file:`/dev/sdb` as the two working files, you would use
766 ``filename=/dev/sda:/dev/sdb``. This also means that whenever this option is
767 specified, :option:`nrfiles` is ignored. The size of regular files specified
768 by this option will be :option:`size` divided by number of files unless an
769 explicit size is specified by :option:`filesize`.
771 Each colon and backslash in the wanted path must be escaped with a ``\``
772 character. For instance, if the path is :file:`/dev/dsk/foo@3,0:c` then you
773 would use ``filename=/dev/dsk/foo@3,0\:c`` and if the path is
774 :file:`F:\\filename` then you would use ``filename=F\:\\filename``.
776 On Windows, disk devices are accessed as :file:`\\\\.\\PhysicalDrive0` for
777 the first device, :file:`\\\\.\\PhysicalDrive1` for the second etc.
778 Note: Windows and FreeBSD prevent write access to areas
779 of the disk containing in-use data (e.g. filesystems).
781 The filename "`-`" is a reserved name, meaning *stdin* or *stdout*. Which
782 of the two depends on the read/write direction set.
784 .. option:: filename_format=str
786 If sharing multiple files between jobs, it is usually necessary to have fio
787 generate the exact names that you want. By default, fio will name a file
788 based on the default file format specification of
789 :file:`jobname.jobnumber.filenumber`. With this option, that can be
790 customized. Fio will recognize and replace the following keywords in this
794 The name of the worker thread or process.
796 The incremental number of the worker thread or process.
798 The incremental number of the file for that worker thread or
801 To have dependent jobs share a set of files, this option can be set to have
802 fio generate filenames that are shared between the two. For instance, if
803 :file:`testfiles.$filenum` is specified, file number 4 for any job will be
804 named :file:`testfiles.4`. The default of :file:`$jobname.$jobnum.$filenum`
805 will be used if no other format specifier is given.
807 If you specify a path then the directories will be created up to the
808 main directory for the file. So for example if you specify
809 ``filename_format=a/b/c/$jobnum`` then the directories a/b/c will be
810 created before the file setup part of the job. If you specify
811 :option:`directory` then the path will be relative that directory,
812 otherwise it is treated as the absolute path.
814 .. option:: unique_filename=bool
816 To avoid collisions between networked clients, fio defaults to prefixing any
817 generated filenames (with a directory specified) with the source of the
818 client connecting. To disable this behavior, set this option to 0.
820 .. option:: opendir=str
822 Recursively open any files below directory `str`.
824 .. option:: lockfile=str
826 Fio defaults to not locking any files before it does I/O to them. If a file
827 or file descriptor is shared, fio can serialize I/O to that file to make the
828 end result consistent. This is usual for emulating real workloads that share
829 files. The lock modes are:
832 No locking. The default.
834 Only one thread or process may do I/O at a time, excluding all
837 Read-write locking on the file. Many readers may
838 access the file at the same time, but writes get exclusive access.
840 .. option:: nrfiles=int
842 Number of files to use for this job. Defaults to 1. The size of files
843 will be :option:`size` divided by this unless explicit size is specified by
844 :option:`filesize`. Files are created for each thread separately, and each
845 file will have a file number within its name by default, as explained in
846 :option:`filename` section.
849 .. option:: openfiles=int
851 Number of files to keep open at the same time. Defaults to the same as
852 :option:`nrfiles`, can be set smaller to limit the number simultaneous
855 .. option:: file_service_type=str
857 Defines how fio decides which file from a job to service next. The following
861 Choose a file at random.
864 Round robin over opened files. This is the default.
867 Finish one file before moving on to the next. Multiple files can
868 still be open depending on :option:`openfiles`.
871 Use a *Zipf* distribution to decide what file to access.
874 Use a *Pareto* distribution to decide what file to access.
877 Use a *Gaussian* (normal) distribution to decide what file to
883 For *random*, *roundrobin*, and *sequential*, a postfix can be appended to
884 tell fio how many I/Os to issue before switching to a new file. For example,
885 specifying ``file_service_type=random:8`` would cause fio to issue
886 8 I/Os before selecting a new file at random. For the non-uniform
887 distributions, a floating point postfix can be given to influence how the
888 distribution is skewed. See :option:`random_distribution` for a description
889 of how that would work.
891 .. option:: ioscheduler=str
893 Attempt to switch the device hosting the file to the specified I/O scheduler
896 .. option:: create_serialize=bool
898 If true, serialize the file creation for the jobs. This may be handy to
899 avoid interleaving of data files, which may greatly depend on the filesystem
900 used and even the number of processors in the system. Default: true.
902 .. option:: create_fsync=bool
904 :manpage:`fsync(2)` the data file after creation. This is the default.
906 .. option:: create_on_open=bool
908 If true, don't pre-create files but allow the job's open() to create a file
909 when it's time to do I/O. Default: false -- pre-create all necessary files
912 .. option:: create_only=bool
914 If true, fio will only run the setup phase of the job. If files need to be
915 laid out or updated on disk, only that will be done -- the actual job contents
916 are not executed. Default: false.
918 .. option:: allow_file_create=bool
920 If true, fio is permitted to create files as part of its workload. If this
921 option is false, then fio will error out if
922 the files it needs to use don't already exist. Default: true.
924 .. option:: allow_mounted_write=bool
926 If this isn't set, fio will abort jobs that are destructive (e.g. that write)
927 to what appears to be a mounted device or partition. This should help catch
928 creating inadvertently destructive tests, not realizing that the test will
929 destroy data on the mounted file system. Note that some platforms don't allow
930 writing against a mounted device regardless of this option. Default: false.
932 .. option:: pre_read=bool
934 If this is given, files will be pre-read into memory before starting the
935 given I/O operation. This will also clear the :option:`invalidate` flag,
936 since it is pointless to pre-read and then drop the cache. This will only
937 work for I/O engines that are seek-able, since they allow you to read the
938 same data multiple times. Thus it will not work on non-seekable I/O engines
939 (e.g. network, splice). Default: false.
941 .. option:: unlink=bool
943 Unlink the job files when done. Not the default, as repeated runs of that
944 job would then waste time recreating the file set again and again. Default:
947 .. option:: unlink_each_loop=bool
949 Unlink job files after each iteration or loop. Default: false.
951 .. option:: zonesize=int
953 Divide a file into zones of the specified size. See :option:`zoneskip`.
955 .. option:: zonerange=int
957 Give size of an I/O zone. See :option:`zoneskip`.
959 .. option:: zoneskip=int
961 Skip the specified number of bytes when :option:`zonesize` data has been
962 read. The two zone options can be used to only do I/O on zones of a file.
968 .. option:: direct=bool
970 If value is true, use non-buffered I/O. This is usually O_DIRECT. Note that
971 OpenBSD and ZFS on Solaris don't support direct I/O. On Windows the synchronous
972 ioengines don't support direct I/O. Default: false.
974 .. option:: atomic=bool
976 If value is true, attempt to use atomic direct I/O. Atomic writes are
977 guaranteed to be stable once acknowledged by the operating system. Only
978 Linux supports O_ATOMIC right now.
980 .. option:: buffered=bool
982 If value is true, use buffered I/O. This is the opposite of the
983 :option:`direct` option. Defaults to true.
985 .. option:: readwrite=str, rw=str
987 Type of I/O pattern. Accepted values are:
994 Sequential trims (Linux block devices only).
1000 Random trims (Linux block devices only).
1002 Sequential mixed reads and writes.
1004 Random mixed reads and writes.
1006 Sequential trim+write sequences. Blocks will be trimmed first,
1007 then the same blocks will be written to.
1009 Fio defaults to read if the option is not specified. For the mixed I/O
1010 types, the default is to split them 50/50. For certain types of I/O the
1011 result may still be skewed a bit, since the speed may be different.
1013 It is possible to specify the number of I/Os to do before getting a new
1014 offset by appending ``:<nr>`` to the end of the string given. For a
1015 random read, it would look like ``rw=randread:8`` for passing in an offset
1016 modifier with a value of 8. If the suffix is used with a sequential I/O
1017 pattern, then the *<nr>* value specified will be **added** to the generated
1018 offset for each I/O turning sequential I/O into sequential I/O with holes.
1019 For instance, using ``rw=write:4k`` will skip 4k for every write. Also see
1020 the :option:`rw_sequencer` option.
1022 .. option:: rw_sequencer=str
1024 If an offset modifier is given by appending a number to the ``rw=<str>``
1025 line, then this option controls how that number modifies the I/O offset
1026 being generated. Accepted values are:
1029 Generate sequential offset.
1031 Generate the same offset.
1033 ``sequential`` is only useful for random I/O, where fio would normally
1034 generate a new random offset for every I/O. If you append e.g. 8 to randread,
1035 you would get a new random offset for every 8 I/Os. The result would be a
1036 seek for only every 8 I/Os, instead of for every I/O. Use ``rw=randread:8``
1037 to specify that. As sequential I/O is already sequential, setting
1038 ``sequential`` for that would not result in any differences. ``identical``
1039 behaves in a similar fashion, except it sends the same offset 8 number of
1040 times before generating a new offset.
1042 .. option:: unified_rw_reporting=bool
1044 Fio normally reports statistics on a per data direction basis, meaning that
1045 reads, writes, and trims are accounted and reported separately. If this
1046 option is set fio sums the results and report them as "mixed" instead.
1048 .. option:: randrepeat=bool
1050 Seed the random number generator used for random I/O patterns in a
1051 predictable way so the pattern is repeatable across runs. Default: true.
1053 .. option:: allrandrepeat=bool
1055 Seed all random number generators in a predictable way so results are
1056 repeatable across runs. Default: false.
1058 .. option:: randseed=int
1060 Seed the random number generators based on this seed value, to be able to
1061 control what sequence of output is being generated. If not set, the random
1062 sequence depends on the :option:`randrepeat` setting.
1064 .. option:: fallocate=str
1066 Whether pre-allocation is performed when laying down files.
1067 Accepted values are:
1070 Do not pre-allocate space.
1073 Use a platform's native pre-allocation call but fall back to
1074 **none** behavior if it fails/is not implemented.
1077 Pre-allocate via :manpage:`posix_fallocate(3)`.
1080 Pre-allocate via :manpage:`fallocate(2)` with
1081 FALLOC_FL_KEEP_SIZE set.
1084 Backward-compatible alias for **none**.
1087 Backward-compatible alias for **posix**.
1089 May not be available on all supported platforms. **keep** is only available
1090 on Linux. If using ZFS on Solaris this cannot be set to **posix**
1091 because ZFS doesn't support pre-allocation. Default: **native** if any
1092 pre-allocation methods are available, **none** if not.
1094 .. option:: fadvise_hint=str
1096 Use :manpage:`posix_fadvise(2)` to advise the kernel on what I/O patterns
1097 are likely to be issued. Accepted values are:
1100 Backwards-compatible hint for "no hint".
1103 Backwards compatible hint for "advise with fio workload type". This
1104 uses **FADV_RANDOM** for a random workload, and **FADV_SEQUENTIAL**
1105 for a sequential workload.
1108 Advise using **FADV_SEQUENTIAL**.
1111 Advise using **FADV_RANDOM**.
1113 .. option:: write_hint=str
1115 Use :manpage:`fcntl(2)` to advise the kernel what life time to expect
1116 from a write. Only supported on Linux, as of version 4.13. Accepted
1120 No particular life time associated with this file.
1123 Data written to this file has a short life time.
1126 Data written to this file has a medium life time.
1129 Data written to this file has a long life time.
1132 Data written to this file has a very long life time.
1134 The values are all relative to each other, and no absolute meaning
1135 should be associated with them.
1137 .. option:: offset=int
1139 Start I/O at the provided offset in the file, given as either a fixed size in
1140 bytes or a percentage. If a percentage is given, the generated offset will be
1141 aligned to the minimum ``blocksize`` or to the value of ``offset_align`` if
1142 provided. Data before the given offset will not be touched. This
1143 effectively caps the file size at `real_size - offset`. Can be combined with
1144 :option:`size` to constrain the start and end range of the I/O workload.
1145 A percentage can be specified by a number between 1 and 100 followed by '%',
1146 for example, ``offset=20%`` to specify 20%.
1148 .. option:: offset_align=int
1150 If set to non-zero value, the byte offset generated by a percentage ``offset``
1151 is aligned upwards to this value. Defaults to 0 meaning that a percentage
1152 offset is aligned to the minimum block size.
1154 .. option:: offset_increment=int
1156 If this is provided, then the real offset becomes `offset + offset_increment
1157 * thread_number`, where the thread number is a counter that starts at 0 and
1158 is incremented for each sub-job (i.e. when :option:`numjobs` option is
1159 specified). This option is useful if there are several jobs which are
1160 intended to operate on a file in parallel disjoint segments, with even
1161 spacing between the starting points.
1163 .. option:: number_ios=int
1165 Fio will normally perform I/Os until it has exhausted the size of the region
1166 set by :option:`size`, or if it exhaust the allocated time (or hits an error
1167 condition). With this setting, the range/size can be set independently of
1168 the number of I/Os to perform. When fio reaches this number, it will exit
1169 normally and report status. Note that this does not extend the amount of I/O
1170 that will be done, it will only stop fio if this condition is met before
1171 other end-of-job criteria.
1173 .. option:: fsync=int
1175 If writing to a file, issue an :manpage:`fsync(2)` (or its equivalent) of
1176 the dirty data for every number of blocks given. For example, if you give 32
1177 as a parameter, fio will sync the file after every 32 writes issued. If fio is
1178 using non-buffered I/O, we may not sync the file. The exception is the sg
1179 I/O engine, which synchronizes the disk cache anyway. Defaults to 0, which
1180 means fio does not periodically issue and wait for a sync to complete. Also
1181 see :option:`end_fsync` and :option:`fsync_on_close`.
1183 .. option:: fdatasync=int
1185 Like :option:`fsync` but uses :manpage:`fdatasync(2)` to only sync data and
1186 not metadata blocks. In Windows, FreeBSD, and DragonFlyBSD there is no
1187 :manpage:`fdatasync(2)` so this falls back to using :manpage:`fsync(2)`.
1188 Defaults to 0, which means fio does not periodically issue and wait for a
1189 data-only sync to complete.
1191 .. option:: write_barrier=int
1193 Make every `N-th` write a barrier write.
1195 .. option:: sync_file_range=str:int
1197 Use :manpage:`sync_file_range(2)` for every `int` number of write
1198 operations. Fio will track range of writes that have happened since the last
1199 :manpage:`sync_file_range(2)` call. `str` can currently be one or more of:
1202 SYNC_FILE_RANGE_WAIT_BEFORE
1204 SYNC_FILE_RANGE_WRITE
1206 SYNC_FILE_RANGE_WAIT_AFTER
1208 So if you do ``sync_file_range=wait_before,write:8``, fio would use
1209 ``SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE`` for every 8
1210 writes. Also see the :manpage:`sync_file_range(2)` man page. This option is
1213 .. option:: overwrite=bool
1215 If true, writes to a file will always overwrite existing data. If the file
1216 doesn't already exist, it will be created before the write phase begins. If
1217 the file exists and is large enough for the specified write phase, nothing
1218 will be done. Default: false.
1220 .. option:: end_fsync=bool
1222 If true, :manpage:`fsync(2)` file contents when a write stage has completed.
1225 .. option:: fsync_on_close=bool
1227 If true, fio will :manpage:`fsync(2)` a dirty file on close. This differs
1228 from :option:`end_fsync` in that it will happen on every file close, not
1229 just at the end of the job. Default: false.
1231 .. option:: rwmixread=int
1233 Percentage of a mixed workload that should be reads. Default: 50.
1235 .. option:: rwmixwrite=int
1237 Percentage of a mixed workload that should be writes. If both
1238 :option:`rwmixread` and :option:`rwmixwrite` is given and the values do not
1239 add up to 100%, the latter of the two will be used to override the
1240 first. This may interfere with a given rate setting, if fio is asked to
1241 limit reads or writes to a certain rate. If that is the case, then the
1242 distribution may be skewed. Default: 50.
1244 .. option:: random_distribution=str:float[,str:float][,str:float]
1246 By default, fio will use a completely uniform random distribution when asked
1247 to perform random I/O. Sometimes it is useful to skew the distribution in
1248 specific ways, ensuring that some parts of the data is more hot than others.
1249 fio includes the following distribution models:
1252 Uniform random distribution
1261 Normal (Gaussian) distribution
1264 Zoned random distribution
1267 Zone absolute random distribution
1269 When using a **zipf** or **pareto** distribution, an input value is also
1270 needed to define the access pattern. For **zipf**, this is the `Zipf
1271 theta`. For **pareto**, it's the `Pareto power`. Fio includes a test
1272 program, :command:`fio-genzipf`, that can be used visualize what the given input
1273 values will yield in terms of hit rates. If you wanted to use **zipf** with
1274 a `theta` of 1.2, you would use ``random_distribution=zipf:1.2`` as the
1275 option. If a non-uniform model is used, fio will disable use of the random
1276 map. For the **normal** distribution, a normal (Gaussian) deviation is
1277 supplied as a value between 0 and 100.
1279 For a **zoned** distribution, fio supports specifying percentages of I/O
1280 access that should fall within what range of the file or device. For
1281 example, given a criteria of:
1283 * 60% of accesses should be to the first 10%
1284 * 30% of accesses should be to the next 20%
1285 * 8% of accesses should be to the next 30%
1286 * 2% of accesses should be to the next 40%
1288 we can define that through zoning of the random accesses. For the above
1289 example, the user would do::
1291 random_distribution=zoned:60/10:30/20:8/30:2/40
1293 A **zoned_abs** distribution works exactly like the **zoned**, except
1294 that it takes absolute sizes. For example, let's say you wanted to
1295 define access according to the following criteria:
1297 * 60% of accesses should be to the first 20G
1298 * 30% of accesses should be to the next 100G
1299 * 10% of accesses should be to the next 500G
1301 we can define an absolute zoning distribution with:
1303 random_distribution=zoned_abs=60/20G:30/100G:10/500g
1305 For both **zoned** and **zoned_abs**, fio supports defining up to
1308 Similarly to how :option:`bssplit` works for setting ranges and
1309 percentages of block sizes. Like :option:`bssplit`, it's possible to
1310 specify separate zones for reads, writes, and trims. If just one set
1311 is given, it'll apply to all of them. This goes for both **zoned**
1312 **zoned_abs** distributions.
1314 .. option:: percentage_random=int[,int][,int]
1316 For a random workload, set how big a percentage should be random. This
1317 defaults to 100%, in which case the workload is fully random. It can be set
1318 from anywhere from 0 to 100. Setting it to 0 would make the workload fully
1319 sequential. Any setting in between will result in a random mix of sequential
1320 and random I/O, at the given percentages. Comma-separated values may be
1321 specified for reads, writes, and trims as described in :option:`blocksize`.
1323 .. option:: norandommap
1325 Normally fio will cover every block of the file when doing random I/O. If
1326 this option is given, fio will just get a new random offset without looking
1327 at past I/O history. This means that some blocks may not be read or written,
1328 and that some blocks may be read/written more than once. If this option is
1329 used with :option:`verify` and multiple blocksizes (via :option:`bsrange`),
1330 only intact blocks are verified, i.e., partially-overwritten blocks are
1333 .. option:: softrandommap=bool
1335 See :option:`norandommap`. If fio runs with the random block map enabled and
1336 it fails to allocate the map, if this option is set it will continue without
1337 a random block map. As coverage will not be as complete as with random maps,
1338 this option is disabled by default.
1340 .. option:: random_generator=str
1342 Fio supports the following engines for generating I/O offsets for random I/O:
1345 Strong 2^88 cycle random number generator.
1347 Linear feedback shift register generator.
1349 Strong 64-bit 2^258 cycle random number generator.
1351 **tausworthe** is a strong random number generator, but it requires tracking
1352 on the side if we want to ensure that blocks are only read or written
1353 once. **lfsr** guarantees that we never generate the same offset twice, and
1354 it's also less computationally expensive. It's not a true random generator,
1355 however, though for I/O purposes it's typically good enough. **lfsr** only
1356 works with single block sizes, not with workloads that use multiple block
1357 sizes. If used with such a workload, fio may read or write some blocks
1358 multiple times. The default value is **tausworthe**, unless the required
1359 space exceeds 2^32 blocks. If it does, then **tausworthe64** is
1360 selected automatically.
1366 .. option:: blocksize=int[,int][,int], bs=int[,int][,int]
1368 The block size in bytes used for I/O units. Default: 4096. A single value
1369 applies to reads, writes, and trims. Comma-separated values may be
1370 specified for reads, writes, and trims. A value not terminated in a comma
1371 applies to subsequent types.
1376 means 256k for reads, writes and trims.
1379 means 8k for reads, 32k for writes and trims.
1382 means 8k for reads, 32k for writes, and default for trims.
1385 means default for reads, 8k for writes and trims.
1388 means default for reads, 8k for writes, and default for trims.
1390 .. option:: blocksize_range=irange[,irange][,irange], bsrange=irange[,irange][,irange]
1392 A range of block sizes in bytes for I/O units. The issued I/O unit will
1393 always be a multiple of the minimum size, unless
1394 :option:`blocksize_unaligned` is set.
1396 Comma-separated ranges may be specified for reads, writes, and trims as
1397 described in :option:`blocksize`.
1399 Example: ``bsrange=1k-4k,2k-8k``.
1401 .. option:: bssplit=str[,str][,str]
1403 Sometimes you want even finer grained control of the block sizes
1404 issued, not just an even split between them. This option allows you to
1405 weight various block sizes, so that you are able to define a specific
1406 amount of block sizes issued. The format for this option is::
1408 bssplit=blocksize/percentage:blocksize/percentage
1410 for as many block sizes as needed. So if you want to define a workload
1411 that has 50% 64k blocks, 10% 4k blocks, and 40% 32k blocks, you would
1414 bssplit=4k/10:64k/50:32k/40
1416 Ordering does not matter. If the percentage is left blank, fio will
1417 fill in the remaining values evenly. So a bssplit option like this one::
1419 bssplit=4k/50:1k/:32k/
1421 would have 50% 4k ios, and 25% 1k and 32k ios. The percentages always
1422 add up to 100, if bssplit is given a range that adds up to more, it
1425 Comma-separated values may be specified for reads, writes, and trims as
1426 described in :option:`blocksize`.
1428 If you want a workload that has 50% 2k reads and 50% 4k reads, while
1429 having 90% 4k writes and 10% 8k writes, you would specify::
1431 bssplit=2k/50:4k/50,4k/90,8k/10
1433 Fio supports defining up to 64 different weights for each data
1436 .. option:: blocksize_unaligned, bs_unaligned
1438 If set, fio will issue I/O units with any size within
1439 :option:`blocksize_range`, not just multiples of the minimum size. This
1440 typically won't work with direct I/O, as that normally requires sector
1443 .. option:: bs_is_seq_rand=bool
1445 If this option is set, fio will use the normal read,write blocksize settings
1446 as sequential,random blocksize settings instead. Any random read or write
1447 will use the WRITE blocksize settings, and any sequential read or write will
1448 use the READ blocksize settings.
1450 .. option:: blockalign=int[,int][,int], ba=int[,int][,int]
1452 Boundary to which fio will align random I/O units. Default:
1453 :option:`blocksize`. Minimum alignment is typically 512b for using direct
1454 I/O, though it usually depends on the hardware block size. This option is
1455 mutually exclusive with using a random map for files, so it will turn off
1456 that option. Comma-separated values may be specified for reads, writes, and
1457 trims as described in :option:`blocksize`.
1463 .. option:: zero_buffers
1465 Initialize buffers with all zeros. Default: fill buffers with random data.
1467 .. option:: refill_buffers
1469 If this option is given, fio will refill the I/O buffers on every
1470 submit. Only makes sense if :option:`zero_buffers` isn't specified,
1471 naturally. Defaults to being unset i.e., the buffer is only filled at
1472 init time and the data in it is reused when possible but if any of
1473 :option:`verify`, :option:`buffer_compress_percentage` or
1474 :option:`dedupe_percentage` are enabled then `refill_buffers` is also
1475 automatically enabled.
1477 .. option:: scramble_buffers=bool
1479 If :option:`refill_buffers` is too costly and the target is using data
1480 deduplication, then setting this option will slightly modify the I/O buffer
1481 contents to defeat normal de-dupe attempts. This is not enough to defeat
1482 more clever block compression attempts, but it will stop naive dedupe of
1483 blocks. Default: true.
1485 .. option:: buffer_compress_percentage=int
1487 If this is set, then fio will attempt to provide I/O buffer content
1488 (on WRITEs) that compresses to the specified level. Fio does this by
1489 providing a mix of random data followed by fixed pattern data. The
1490 fixed pattern is either zeros, or the pattern specified by
1491 :option:`buffer_pattern`. If the `buffer_pattern` option is used, it
1492 might skew the compression ratio slightly. Setting
1493 `buffer_compress_percentage` to a value other than 100 will also
1494 enable :option:`refill_buffers` in order to reduce the likelihood that
1495 adjacent blocks are so similar that they over compress when seen
1496 together. See :option:`buffer_compress_chunk` for how to set a finer or
1497 coarser granularity for the random/fixed data region. Defaults to unset
1498 i.e., buffer data will not adhere to any compression level.
1500 .. option:: buffer_compress_chunk=int
1502 This setting allows fio to manage how big the random/fixed data region
1503 is when using :option:`buffer_compress_percentage`. When
1504 `buffer_compress_chunk` is set to some non-zero value smaller than the
1505 block size, fio can repeat the random/fixed region throughout the I/O
1506 buffer at the specified interval (which particularly useful when
1507 bigger block sizes are used for a job). When set to 0, fio will use a
1508 chunk size that matches the block size resulting in a single
1509 random/fixed region within the I/O buffer. Defaults to 512. When the
1510 unit is omitted, the value is interpreted in bytes.
1512 .. option:: buffer_pattern=str
1514 If set, fio will fill the I/O buffers with this pattern or with the contents
1515 of a file. If not set, the contents of I/O buffers are defined by the other
1516 options related to buffer contents. The setting can be any pattern of bytes,
1517 and can be prefixed with 0x for hex values. It may also be a string, where
1518 the string must then be wrapped with ``""``. Or it may also be a filename,
1519 where the filename must be wrapped with ``''`` in which case the file is
1520 opened and read. Note that not all the file contents will be read if that
1521 would cause the buffers to overflow. So, for example::
1523 buffer_pattern='filename'
1527 buffer_pattern="abcd"
1535 buffer_pattern=0xdeadface
1537 Also you can combine everything together in any order::
1539 buffer_pattern=0xdeadface"abcd"-12'filename'
1541 .. option:: dedupe_percentage=int
1543 If set, fio will generate this percentage of identical buffers when
1544 writing. These buffers will be naturally dedupable. The contents of the
1545 buffers depend on what other buffer compression settings have been set. It's
1546 possible to have the individual buffers either fully compressible, or not at
1547 all -- this option only controls the distribution of unique buffers. Setting
1548 this option will also enable :option:`refill_buffers` to prevent every buffer
1551 .. option:: invalidate=bool
1553 Invalidate the buffer/page cache parts of the files to be used prior to
1554 starting I/O if the platform and file type support it. Defaults to true.
1555 This will be ignored if :option:`pre_read` is also specified for the
1558 .. option:: sync=bool
1560 Use synchronous I/O for buffered writes. For the majority of I/O engines,
1561 this means using O_SYNC. Default: false.
1563 .. option:: iomem=str, mem=str
1565 Fio can use various types of memory as the I/O unit buffer. The allowed
1569 Use memory from :manpage:`malloc(3)` as the buffers. Default memory
1573 Use shared memory as the buffers. Allocated through
1574 :manpage:`shmget(2)`.
1577 Same as shm, but use huge pages as backing.
1580 Use :manpage:`mmap(2)` to allocate buffers. May either be anonymous memory, or can
1581 be file backed if a filename is given after the option. The format
1582 is `mem=mmap:/path/to/file`.
1585 Use a memory mapped huge file as the buffer backing. Append filename
1586 after mmaphuge, ala `mem=mmaphuge:/hugetlbfs/file`.
1589 Same as mmap, but use a MMAP_SHARED mapping.
1592 Use GPU memory as the buffers for GPUDirect RDMA benchmark.
1593 The :option:`ioengine` must be `rdma`.
1595 The area allocated is a function of the maximum allowed bs size for the job,
1596 multiplied by the I/O depth given. Note that for **shmhuge** and
1597 **mmaphuge** to work, the system must have free huge pages allocated. This
1598 can normally be checked and set by reading/writing
1599 :file:`/proc/sys/vm/nr_hugepages` on a Linux system. Fio assumes a huge page
1600 is 4MiB in size. So to calculate the number of huge pages you need for a
1601 given job file, add up the I/O depth of all jobs (normally one unless
1602 :option:`iodepth` is used) and multiply by the maximum bs set. Then divide
1603 that number by the huge page size. You can see the size of the huge pages in
1604 :file:`/proc/meminfo`. If no huge pages are allocated by having a non-zero
1605 number in `nr_hugepages`, using **mmaphuge** or **shmhuge** will fail. Also
1606 see :option:`hugepage-size`.
1608 **mmaphuge** also needs to have hugetlbfs mounted and the file location
1609 should point there. So if it's mounted in :file:`/huge`, you would use
1610 `mem=mmaphuge:/huge/somefile`.
1612 .. option:: iomem_align=int, mem_align=int
1614 This indicates the memory alignment of the I/O memory buffers. Note that
1615 the given alignment is applied to the first I/O unit buffer, if using
1616 :option:`iodepth` the alignment of the following buffers are given by the
1617 :option:`bs` used. In other words, if using a :option:`bs` that is a
1618 multiple of the page sized in the system, all buffers will be aligned to
1619 this value. If using a :option:`bs` that is not page aligned, the alignment
1620 of subsequent I/O memory buffers is the sum of the :option:`iomem_align` and
1623 .. option:: hugepage-size=int
1625 Defines the size of a huge page. Must at least be equal to the system
1626 setting, see :file:`/proc/meminfo`. Defaults to 4MiB. Should probably
1627 always be a multiple of megabytes, so using ``hugepage-size=Xm`` is the
1628 preferred way to set this to avoid setting a non-pow-2 bad value.
1630 .. option:: lockmem=int
1632 Pin the specified amount of memory with :manpage:`mlock(2)`. Can be used to
1633 simulate a smaller amount of memory. The amount specified is per worker.
1639 .. option:: size=int
1641 The total size of file I/O for each thread of this job. Fio will run until
1642 this many bytes has been transferred, unless runtime is limited by other options
1643 (such as :option:`runtime`, for instance, or increased/decreased by :option:`io_size`).
1644 Fio will divide this size between the available files determined by options
1645 such as :option:`nrfiles`, :option:`filename`, unless :option:`filesize` is
1646 specified by the job. If the result of division happens to be 0, the size is
1647 set to the physical size of the given files or devices if they exist.
1648 If this option is not specified, fio will use the full size of the given
1649 files or devices. If the files do not exist, size must be given. It is also
1650 possible to give size as a percentage between 1 and 100. If ``size=20%`` is
1651 given, fio will use 20% of the full size of the given files or devices.
1652 Can be combined with :option:`offset` to constrain the start and end range
1653 that I/O will be done within.
1655 .. option:: io_size=int, io_limit=int
1657 Normally fio operates within the region set by :option:`size`, which means
1658 that the :option:`size` option sets both the region and size of I/O to be
1659 performed. Sometimes that is not what you want. With this option, it is
1660 possible to define just the amount of I/O that fio should do. For instance,
1661 if :option:`size` is set to 20GiB and :option:`io_size` is set to 5GiB, fio
1662 will perform I/O within the first 20GiB but exit when 5GiB have been
1663 done. The opposite is also possible -- if :option:`size` is set to 20GiB,
1664 and :option:`io_size` is set to 40GiB, then fio will do 40GiB of I/O within
1665 the 0..20GiB region.
1667 .. option:: filesize=irange(int)
1669 Individual file sizes. May be a range, in which case fio will select sizes
1670 for files at random within the given range and limited to :option:`size` in
1671 total (if that is given). If not given, each created file is the same size.
1672 This option overrides :option:`size` in terms of file size, which means
1673 this value is used as a fixed size or possible range of each file.
1675 .. option:: file_append=bool
1677 Perform I/O after the end of the file. Normally fio will operate within the
1678 size of a file. If this option is set, then fio will append to the file
1679 instead. This has identical behavior to setting :option:`offset` to the size
1680 of a file. This option is ignored on non-regular files.
1682 .. option:: fill_device=bool, fill_fs=bool
1684 Sets size to something really large and waits for ENOSPC (no space left on
1685 device) as the terminating condition. Only makes sense with sequential
1686 write. For a read workload, the mount point will be filled first then I/O
1687 started on the result. This option doesn't make sense if operating on a raw
1688 device node, since the size of that is already known by the file system.
1689 Additionally, writing beyond end-of-device will not return ENOSPC there.
1695 .. option:: ioengine=str
1697 Defines how the job issues I/O to the file. The following types are defined:
1700 Basic :manpage:`read(2)` or :manpage:`write(2)`
1701 I/O. :manpage:`lseek(2)` is used to position the I/O location.
1702 See :option:`fsync` and :option:`fdatasync` for syncing write I/Os.
1705 Basic :manpage:`pread(2)` or :manpage:`pwrite(2)` I/O. Default on
1706 all supported operating systems except for Windows.
1709 Basic :manpage:`readv(2)` or :manpage:`writev(2)` I/O. Will emulate
1710 queuing by coalescing adjacent I/Os into a single submission.
1713 Basic :manpage:`preadv(2)` or :manpage:`pwritev(2)` I/O.
1716 Basic :manpage:`preadv2(2)` or :manpage:`pwritev2(2)` I/O.
1719 Linux native asynchronous I/O. Note that Linux may only support
1720 queued behavior with non-buffered I/O (set ``direct=1`` or
1722 This engine defines engine specific options.
1725 POSIX asynchronous I/O using :manpage:`aio_read(3)` and
1726 :manpage:`aio_write(3)`.
1729 Solaris native asynchronous I/O.
1732 Windows native asynchronous I/O. Default on Windows.
1735 File is memory mapped with :manpage:`mmap(2)` and data copied
1736 to/from using :manpage:`memcpy(3)`.
1739 :manpage:`splice(2)` is used to transfer the data and
1740 :manpage:`vmsplice(2)` to transfer data from user space to the
1744 SCSI generic sg v3 I/O. May either be synchronous using the SG_IO
1745 ioctl, or if the target is an sg character device we use
1746 :manpage:`read(2)` and :manpage:`write(2)` for asynchronous
1747 I/O. Requires :option:`filename` option to specify either block or
1751 Doesn't transfer any data, just pretends to. This is mainly used to
1752 exercise fio itself and for debugging/testing purposes.
1755 Transfer over the network to given ``host:port``. Depending on the
1756 :option:`protocol` used, the :option:`hostname`, :option:`port`,
1757 :option:`listen` and :option:`filename` options are used to specify
1758 what sort of connection to make, while the :option:`protocol` option
1759 determines which protocol will be used. This engine defines engine
1763 Like **net**, but uses :manpage:`splice(2)` and
1764 :manpage:`vmsplice(2)` to map data and send/receive.
1765 This engine defines engine specific options.
1768 Doesn't transfer any data, but burns CPU cycles according to the
1769 :option:`cpuload` and :option:`cpuchunks` options. Setting
1770 :option:`cpuload`\=85 will cause that job to do nothing but burn 85%
1771 of the CPU. In case of SMP machines, use :option:`numjobs`\=<nr_of_cpu>
1772 to get desired CPU usage, as the cpuload only loads a
1773 single CPU at the desired rate. A job never finishes unless there is
1774 at least one non-cpuio job.
1777 The GUASI I/O engine is the Generic Userspace Asyncronous Syscall
1778 Interface approach to async I/O. See
1780 http://www.xmailserver.org/guasi-lib.html
1782 for more info on GUASI.
1785 The RDMA I/O engine supports both RDMA memory semantics
1786 (RDMA_WRITE/RDMA_READ) and channel semantics (Send/Recv) for the
1787 InfiniBand, RoCE and iWARP protocols. This engine defines engine
1791 I/O engine that does regular fallocate to simulate data transfer as
1795 does fallocate(,mode = FALLOC_FL_KEEP_SIZE,).
1798 does fallocate(,mode = 0).
1801 does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE).
1804 I/O engine that sends :manpage:`ftruncate(2)` operations in response
1805 to write (DDIR_WRITE) events. Each ftruncate issued sets the file's
1806 size to the current block offset. :option:`blocksize` is ignored.
1809 I/O engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate
1810 defragment activity in request to DDIR_WRITE event.
1813 I/O engine supporting direct access to Ceph Rados Block Devices
1814 (RBD) via librbd without the need to use the kernel rbd driver. This
1815 ioengine defines engine specific options.
1818 Using GlusterFS libgfapi sync interface to direct access to
1819 GlusterFS volumes without having to go through FUSE. This ioengine
1820 defines engine specific options.
1823 Using GlusterFS libgfapi async interface to direct access to
1824 GlusterFS volumes without having to go through FUSE. This ioengine
1825 defines engine specific options.
1828 Read and write through Hadoop (HDFS). The :option:`filename` option
1829 is used to specify host,port of the hdfs name-node to connect. This
1830 engine interprets offsets a little differently. In HDFS, files once
1831 created cannot be modified so random writes are not possible. To
1832 imitate this the libhdfs engine expects a bunch of small files to be
1833 created over HDFS and will randomly pick a file from them
1834 based on the offset generated by fio backend (see the example
1835 job file to create such files, use ``rw=write`` option). Please
1836 note, it may be necessary to set environment variables to work
1837 with HDFS/libhdfs properly. Each job uses its own connection to
1841 Read, write and erase an MTD character device (e.g.,
1842 :file:`/dev/mtd0`). Discards are treated as erases. Depending on the
1843 underlying device type, the I/O may have to go in a certain pattern,
1844 e.g., on NAND, writing sequentially to erase blocks and discarding
1845 before overwriting. The `trimwrite` mode works well for this
1849 Read and write using filesystem DAX to a file on a filesystem
1850 mounted with DAX on a persistent memory device through the NVML
1854 Read and write using device DAX to a persistent memory device (e.g.,
1855 /dev/dax0.0) through the NVML libpmem library.
1858 Prefix to specify loading an external I/O engine object file. Append
1859 the engine filename, e.g. ``ioengine=external:/tmp/foo.o`` to load
1860 ioengine :file:`foo.o` in :file:`/tmp`. The path can be either
1861 absolute or relative. See :file:`engines/skeleton_external.c` for
1862 details of writing an external I/O engine.
1865 Simply create the files and do no I/O to them. You still need to
1866 set `filesize` so that all the accounting still occurs, but no
1867 actual I/O will be done other than creating the file.
1870 Read and write using mmap I/O to a file on a filesystem
1871 mounted with DAX on a persistent memory device through the NVML
1874 I/O engine specific parameters
1875 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1877 In addition, there are some parameters which are only valid when a specific
1878 :option:`ioengine` is in use. These are used identically to normal parameters,
1879 with the caveat that when used on the command line, they must come after the
1880 :option:`ioengine` that defines them is selected.
1882 .. option:: userspace_reap : [libaio]
1884 Normally, with the libaio engine in use, fio will use the
1885 :manpage:`io_getevents(2)` system call to reap newly returned events. With
1886 this flag turned on, the AIO ring will be read directly from user-space to
1887 reap events. The reaping mode is only enabled when polling for a minimum of
1888 0 events (e.g. when :option:`iodepth_batch_complete` `=0`).
1890 .. option:: hipri : [pvsync2]
1892 Set RWF_HIPRI on I/O, indicating to the kernel that it's of higher priority
1895 .. option:: hipri_percentage : [pvsync2]
1897 When hipri is set this determines the probability of a pvsync2 I/O being high
1898 priority. The default is 100%.
1900 .. option:: cpuload=int : [cpuio]
1902 Attempt to use the specified percentage of CPU cycles. This is a mandatory
1903 option when using cpuio I/O engine.
1905 .. option:: cpuchunks=int : [cpuio]
1907 Split the load into cycles of the given time. In microseconds.
1909 .. option:: exit_on_io_done=bool : [cpuio]
1911 Detect when I/O threads are done, then exit.
1913 .. option:: namenode=str : [libhdfs]
1915 The hostname or IP address of a HDFS cluster namenode to contact.
1917 .. option:: port=int
1921 The listening port of the HFDS cluster namenode.
1925 The TCP or UDP port to bind to or connect to. If this is used with
1926 :option:`numjobs` to spawn multiple instances of the same job type, then
1927 this will be the starting port number since fio will use a range of
1932 The port to use for RDMA-CM communication. This should be the same value
1933 on the client and the server side.
1935 .. option:: hostname=str : [netsplice] [net] [rdma]
1937 The hostname or IP address to use for TCP, UDP or RDMA-CM based I/O. If the job
1938 is a TCP listener or UDP reader, the hostname is not used and must be omitted
1939 unless it is a valid UDP multicast address.
1941 .. option:: interface=str : [netsplice] [net]
1943 The IP address of the network interface used to send or receive UDP
1946 .. option:: ttl=int : [netsplice] [net]
1948 Time-to-live value for outgoing UDP multicast packets. Default: 1.
1950 .. option:: nodelay=bool : [netsplice] [net]
1952 Set TCP_NODELAY on TCP connections.
1954 .. option:: protocol=str, proto=str : [netsplice] [net]
1956 The network protocol to use. Accepted values are:
1959 Transmission control protocol.
1961 Transmission control protocol V6.
1963 User datagram protocol.
1965 User datagram protocol V6.
1969 When the protocol is TCP or UDP, the port must also be given, as well as the
1970 hostname if the job is a TCP listener or UDP reader. For unix sockets, the
1971 normal :option:`filename` option should be used and the port is invalid.
1973 .. option:: listen : [netsplice] [net]
1975 For TCP network connections, tell fio to listen for incoming connections
1976 rather than initiating an outgoing connection. The :option:`hostname` must
1977 be omitted if this option is used.
1979 .. option:: pingpong : [netsplice] [net]
1981 Normally a network writer will just continue writing data, and a network
1982 reader will just consume packages. If ``pingpong=1`` is set, a writer will
1983 send its normal payload to the reader, then wait for the reader to send the
1984 same payload back. This allows fio to measure network latencies. The
1985 submission and completion latencies then measure local time spent sending or
1986 receiving, and the completion latency measures how long it took for the
1987 other end to receive and send back. For UDP multicast traffic
1988 ``pingpong=1`` should only be set for a single reader when multiple readers
1989 are listening to the same address.
1991 .. option:: window_size : [netsplice] [net]
1993 Set the desired socket buffer size for the connection.
1995 .. option:: mss : [netsplice] [net]
1997 Set the TCP maximum segment size (TCP_MAXSEG).
1999 .. option:: donorname=str : [e4defrag]
2001 File will be used as a block donor (swap extents between files).
2003 .. option:: inplace=int : [e4defrag]
2005 Configure donor file blocks allocation strategy:
2008 Default. Preallocate donor's file on init.
2010 Allocate space immediately inside defragment event, and free right
2013 .. option:: clustername=str : [rbd]
2015 Specifies the name of the Ceph cluster.
2017 .. option:: rbdname=str : [rbd]
2019 Specifies the name of the RBD.
2021 .. option:: pool=str : [rbd]
2023 Specifies the name of the Ceph pool containing RBD.
2025 .. option:: clientname=str : [rbd]
2027 Specifies the username (without the 'client.' prefix) used to access the
2028 Ceph cluster. If the *clustername* is specified, the *clientname* shall be
2029 the full *type.id* string. If no type. prefix is given, fio will add
2030 'client.' by default.
2032 .. option:: skip_bad=bool : [mtd]
2034 Skip operations against known bad blocks.
2036 .. option:: hdfsdirectory : [libhdfs]
2038 libhdfs will create chunk in this HDFS directory.
2040 .. option:: chunk_size : [libhdfs]
2042 The size of the chunk to use for each file.
2044 .. option:: verb=str : [rdma]
2046 The RDMA verb to use on this side of the RDMA ioengine connection. Valid
2047 values are write, read, send and recv. These correspond to the equivalent
2048 RDMA verbs (e.g. write = rdma_write etc.). Note that this only needs to be
2049 specified on the client side of the connection. See the examples folder.
2051 .. option:: bindname=str : [rdma]
2053 The name to use to bind the local RDMA-CM connection to a local RDMA device.
2054 This could be a hostname or an IPv4 or IPv6 address. On the server side this
2055 will be passed into the rdma_bind_addr() function and on the client site it
2056 will be used in the rdma_resolve_add() function. This can be useful when
2057 multiple paths exist between the client and the server or in certain loopback
2063 .. option:: iodepth=int
2065 Number of I/O units to keep in flight against the file. Note that
2066 increasing *iodepth* beyond 1 will not affect synchronous ioengines (except
2067 for small degrees when :option:`verify_async` is in use). Even async
2068 engines may impose OS restrictions causing the desired depth not to be
2069 achieved. This may happen on Linux when using libaio and not setting
2070 :option:`direct`\=1, since buffered I/O is not async on that OS. Keep an
2071 eye on the I/O depth distribution in the fio output to verify that the
2072 achieved depth is as expected. Default: 1.
2074 .. option:: iodepth_batch_submit=int, iodepth_batch=int
2076 This defines how many pieces of I/O to submit at once. It defaults to 1
2077 which means that we submit each I/O as soon as it is available, but can be
2078 raised to submit bigger batches of I/O at the time. If it is set to 0 the
2079 :option:`iodepth` value will be used.
2081 .. option:: iodepth_batch_complete_min=int, iodepth_batch_complete=int
2083 This defines how many pieces of I/O to retrieve at once. It defaults to 1
2084 which means that we'll ask for a minimum of 1 I/O in the retrieval process
2085 from the kernel. The I/O retrieval will go on until we hit the limit set by
2086 :option:`iodepth_low`. If this variable is set to 0, then fio will always
2087 check for completed events before queuing more I/O. This helps reduce I/O
2088 latency, at the cost of more retrieval system calls.
2090 .. option:: iodepth_batch_complete_max=int
2092 This defines maximum pieces of I/O to retrieve at once. This variable should
2093 be used along with :option:`iodepth_batch_complete_min`\=int variable,
2094 specifying the range of min and max amount of I/O which should be
2095 retrieved. By default it is equal to the :option:`iodepth_batch_complete_min`
2100 iodepth_batch_complete_min=1
2101 iodepth_batch_complete_max=<iodepth>
2103 which means that we will retrieve at least 1 I/O and up to the whole
2104 submitted queue depth. If none of I/O has been completed yet, we will wait.
2108 iodepth_batch_complete_min=0
2109 iodepth_batch_complete_max=<iodepth>
2111 which means that we can retrieve up to the whole submitted queue depth, but
2112 if none of I/O has been completed yet, we will NOT wait and immediately exit
2113 the system call. In this example we simply do polling.
2115 .. option:: iodepth_low=int
2117 The low water mark indicating when to start filling the queue
2118 again. Defaults to the same as :option:`iodepth`, meaning that fio will
2119 attempt to keep the queue full at all times. If :option:`iodepth` is set to
2120 e.g. 16 and *iodepth_low* is set to 4, then after fio has filled the queue of
2121 16 requests, it will let the depth drain down to 4 before starting to fill
2124 .. option:: serialize_overlap=bool
2126 Serialize in-flight I/Os that might otherwise cause or suffer from data races.
2127 When two or more I/Os are submitted simultaneously, there is no guarantee that
2128 the I/Os will be processed or completed in the submitted order. Further, if
2129 two or more of those I/Os are writes, any overlapping region between them can
2130 become indeterminate/undefined on certain storage. These issues can cause
2131 verification to fail erratically when at least one of the racing I/Os is
2132 changing data and the overlapping region has a non-zero size. Setting
2133 ``serialize_overlap`` tells fio to avoid provoking this behavior by explicitly
2134 serializing in-flight I/Os that have a non-zero overlap. Note that setting
2135 this option can reduce both performance and the :option:`iodepth` achieved.
2136 Additionally this option does not work when :option:`io_submit_mode` is set to
2137 offload. Default: false.
2139 .. option:: io_submit_mode=str
2141 This option controls how fio submits the I/O to the I/O engine. The default
2142 is `inline`, which means that the fio job threads submit and reap I/O
2143 directly. If set to `offload`, the job threads will offload I/O submission
2144 to a dedicated pool of I/O threads. This requires some coordination and thus
2145 has a bit of extra overhead, especially for lower queue depth I/O where it
2146 can increase latencies. The benefit is that fio can manage submission rates
2147 independently of the device completion rates. This avoids skewed latency
2148 reporting if I/O gets backed up on the device side (the coordinated omission
2155 .. option:: thinktime=time
2157 Stall the job for the specified period of time after an I/O has completed before issuing the
2158 next. May be used to simulate processing being done by an application.
2159 When the unit is omitted, the value is interpreted in microseconds. See
2160 :option:`thinktime_blocks` and :option:`thinktime_spin`.
2162 .. option:: thinktime_spin=time
2164 Only valid if :option:`thinktime` is set - pretend to spend CPU time doing
2165 something with the data received, before falling back to sleeping for the
2166 rest of the period specified by :option:`thinktime`. When the unit is
2167 omitted, the value is interpreted in microseconds.
2169 .. option:: thinktime_blocks=int
2171 Only valid if :option:`thinktime` is set - control how many blocks to issue,
2172 before waiting :option:`thinktime` usecs. If not set, defaults to 1 which will make
2173 fio wait :option:`thinktime` usecs after every block. This effectively makes any
2174 queue depth setting redundant, since no more than 1 I/O will be queued
2175 before we have to complete it and do our :option:`thinktime`. In other words, this
2176 setting effectively caps the queue depth if the latter is larger.
2178 .. option:: rate=int[,int][,int]
2180 Cap the bandwidth used by this job. The number is in bytes/sec, the normal
2181 suffix rules apply. Comma-separated values may be specified for reads,
2182 writes, and trims as described in :option:`blocksize`.
2184 For example, using `rate=1m,500k` would limit reads to 1MiB/sec and writes to
2185 500KiB/sec. Capping only reads or writes can be done with `rate=,500k` or
2186 `rate=500k,` where the former will only limit writes (to 500KiB/sec) and the
2187 latter will only limit reads.
2189 .. option:: rate_min=int[,int][,int]
2191 Tell fio to do whatever it can to maintain at least this bandwidth. Failing
2192 to meet this requirement will cause the job to exit. Comma-separated values
2193 may be specified for reads, writes, and trims as described in
2194 :option:`blocksize`.
2196 .. option:: rate_iops=int[,int][,int]
2198 Cap the bandwidth to this number of IOPS. Basically the same as
2199 :option:`rate`, just specified independently of bandwidth. If the job is
2200 given a block size range instead of a fixed value, the smallest block size
2201 is used as the metric. Comma-separated values may be specified for reads,
2202 writes, and trims as described in :option:`blocksize`.
2204 .. option:: rate_iops_min=int[,int][,int]
2206 If fio doesn't meet this rate of I/O, it will cause the job to exit.
2207 Comma-separated values may be specified for reads, writes, and trims as
2208 described in :option:`blocksize`.
2210 .. option:: rate_process=str
2212 This option controls how fio manages rated I/O submissions. The default is
2213 `linear`, which submits I/O in a linear fashion with fixed delays between
2214 I/Os that gets adjusted based on I/O completion rates. If this is set to
2215 `poisson`, fio will submit I/O based on a more real world random request
2216 flow, known as the Poisson process
2217 (https://en.wikipedia.org/wiki/Poisson_point_process). The lambda will be
2218 10^6 / IOPS for the given workload.
2220 .. option:: rate_ignore_thinktime=bool
2222 By default, fio will attempt to catch up to the specified rate setting,
2223 if any kind of thinktime setting was used. If this option is set, then
2224 fio will ignore the thinktime and continue doing IO at the specified
2225 rate, instead of entering a catch-up mode after thinktime is done.
2231 .. option:: latency_target=time
2233 If set, fio will attempt to find the max performance point that the given
2234 workload will run at while maintaining a latency below this target. When
2235 the unit is omitted, the value is interpreted in microseconds. See
2236 :option:`latency_window` and :option:`latency_percentile`.
2238 .. option:: latency_window=time
2240 Used with :option:`latency_target` to specify the sample window that the job
2241 is run at varying queue depths to test the performance. When the unit is
2242 omitted, the value is interpreted in microseconds.
2244 .. option:: latency_percentile=float
2246 The percentage of I/Os that must fall within the criteria specified by
2247 :option:`latency_target` and :option:`latency_window`. If not set, this
2248 defaults to 100.0, meaning that all I/Os must be equal or below to the value
2249 set by :option:`latency_target`.
2251 .. option:: max_latency=time
2253 If set, fio will exit the job with an ETIMEDOUT error if it exceeds this
2254 maximum latency. When the unit is omitted, the value is interpreted in
2257 .. option:: rate_cycle=int
2259 Average bandwidth for :option:`rate` and :option:`rate_min` over this number
2260 of milliseconds. Defaults to 1000.
2266 .. option:: write_iolog=str
2268 Write the issued I/O patterns to the specified file. See
2269 :option:`read_iolog`. Specify a separate file for each job, otherwise the
2270 iologs will be interspersed and the file may be corrupt.
2272 .. option:: read_iolog=str
2274 Open an iolog with the specified filename and replay the I/O patterns it
2275 contains. This can be used to store a workload and replay it sometime
2276 later. The iolog given may also be a blktrace binary file, which allows fio
2277 to replay a workload captured by :command:`blktrace`. See
2278 :manpage:`blktrace(8)` for how to capture such logging data. For blktrace
2279 replay, the file needs to be turned into a blkparse binary data file first
2280 (``blkparse <device> -o /dev/null -d file_for_fio.bin``).
2282 .. option:: replay_no_stall=bool
2284 When replaying I/O with :option:`read_iolog` the default behavior is to
2285 attempt to respect the timestamps within the log and replay them with the
2286 appropriate delay between IOPS. By setting this variable fio will not
2287 respect the timestamps and attempt to replay them as fast as possible while
2288 still respecting ordering. The result is the same I/O pattern to a given
2289 device, but different timings.
2291 .. option:: replay_redirect=str
2293 While replaying I/O patterns using :option:`read_iolog` the default behavior
2294 is to replay the IOPS onto the major/minor device that each IOP was recorded
2295 from. This is sometimes undesirable because on a different machine those
2296 major/minor numbers can map to a different device. Changing hardware on the
2297 same system can also result in a different major/minor mapping.
2298 ``replay_redirect`` causes all I/Os to be replayed onto the single specified
2299 device regardless of the device it was recorded
2300 from. i.e. :option:`replay_redirect`\= :file:`/dev/sdc` would cause all I/O
2301 in the blktrace or iolog to be replayed onto :file:`/dev/sdc`. This means
2302 multiple devices will be replayed onto a single device, if the trace
2303 contains multiple devices. If you want multiple devices to be replayed
2304 concurrently to multiple redirected devices you must blkparse your trace
2305 into separate traces and replay them with independent fio invocations.
2306 Unfortunately this also breaks the strict time ordering between multiple
2309 .. option:: replay_align=int
2311 Force alignment of I/O offsets and lengths in a trace to this power of 2
2314 .. option:: replay_scale=int
2316 Scale sector offsets down by this factor when replaying traces.
2319 Threads, processes and job synchronization
2320 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2324 Fio defaults to creating jobs by using fork, however if this option is
2325 given, fio will create jobs by using POSIX Threads' function
2326 :manpage:`pthread_create(3)` to create threads instead.
2328 .. option:: wait_for=str
2330 If set, the current job won't be started until all workers of the specified
2331 waitee job are done.
2333 ``wait_for`` operates on the job name basis, so there are a few
2334 limitations. First, the waitee must be defined prior to the waiter job
2335 (meaning no forward references). Second, if a job is being referenced as a
2336 waitee, it must have a unique name (no duplicate waitees).
2338 .. option:: nice=int
2340 Run the job with the given nice value. See man :manpage:`nice(2)`.
2342 On Windows, values less than -15 set the process class to "High"; -1 through
2343 -15 set "Above Normal"; 1 through 15 "Below Normal"; and above 15 "Idle"
2346 .. option:: prio=int
2348 Set the I/O priority value of this job. Linux limits us to a positive value
2349 between 0 and 7, with 0 being the highest. See man
2350 :manpage:`ionice(1)`. Refer to an appropriate manpage for other operating
2351 systems since meaning of priority may differ.
2353 .. option:: prioclass=int
2355 Set the I/O priority class. See man :manpage:`ionice(1)`.
2357 .. option:: cpumask=int
2359 Set the CPU affinity of this job. The parameter given is a bit mask of
2360 allowed CPUs the job may run on. So if you want the allowed CPUs to be 1
2361 and 5, you would pass the decimal value of (1 << 1 | 1 << 5), or 34. See man
2362 :manpage:`sched_setaffinity(2)`. This may not work on all supported
2363 operating systems or kernel versions. This option doesn't work well for a
2364 higher CPU count than what you can store in an integer mask, so it can only
2365 control cpus 1-32. For boxes with larger CPU counts, use
2366 :option:`cpus_allowed`.
2368 .. option:: cpus_allowed=str
2370 Controls the same options as :option:`cpumask`, but accepts a textual
2371 specification of the permitted CPUs instead. So to use CPUs 1 and 5 you
2372 would specify ``cpus_allowed=1,5``. This option also allows a range of CPUs
2373 to be specified -- say you wanted a binding to CPUs 1, 5, and 8 to 15, you
2374 would set ``cpus_allowed=1,5,8-15``.
2376 .. option:: cpus_allowed_policy=str
2378 Set the policy of how fio distributes the CPUs specified by
2379 :option:`cpus_allowed` or :option:`cpumask`. Two policies are supported:
2382 All jobs will share the CPU set specified.
2384 Each job will get a unique CPU from the CPU set.
2386 **shared** is the default behavior, if the option isn't specified. If
2387 **split** is specified, then fio will will assign one cpu per job. If not
2388 enough CPUs are given for the jobs listed, then fio will roundrobin the CPUs
2391 .. option:: numa_cpu_nodes=str
2393 Set this job running on specified NUMA nodes' CPUs. The arguments allow
2394 comma delimited list of cpu numbers, A-B ranges, or `all`. Note, to enable
2395 NUMA options support, fio must be built on a system with libnuma-dev(el)
2398 .. option:: numa_mem_policy=str
2400 Set this job's memory policy and corresponding NUMA nodes. Format of the
2405 ``mode`` is one of the following memory poicies: ``default``, ``prefer``,
2406 ``bind``, ``interleave`` or ``local``. For ``default`` and ``local`` memory
2407 policies, no node needs to be specified. For ``prefer``, only one node is
2408 allowed. For ``bind`` and ``interleave`` the ``nodelist`` may be as
2409 follows: a comma delimited list of numbers, A-B ranges, or `all`.
2411 .. option:: cgroup=str
2413 Add job to this control group. If it doesn't exist, it will be created. The
2414 system must have a mounted cgroup blkio mount point for this to work. If
2415 your system doesn't have it mounted, you can do so with::
2417 # mount -t cgroup -o blkio none /cgroup
2419 .. option:: cgroup_weight=int
2421 Set the weight of the cgroup to this value. See the documentation that comes
2422 with the kernel, allowed values are in the range of 100..1000.
2424 .. option:: cgroup_nodelete=bool
2426 Normally fio will delete the cgroups it has created after the job
2427 completion. To override this behavior and to leave cgroups around after the
2428 job completion, set ``cgroup_nodelete=1``. This can be useful if one wants
2429 to inspect various cgroup files after job completion. Default: false.
2431 .. option:: flow_id=int
2433 The ID of the flow. If not specified, it defaults to being a global
2434 flow. See :option:`flow`.
2436 .. option:: flow=int
2438 Weight in token-based flow control. If this value is used, then there is a
2439 'flow counter' which is used to regulate the proportion of activity between
2440 two or more jobs. Fio attempts to keep this flow counter near zero. The
2441 ``flow`` parameter stands for how much should be added or subtracted to the
2442 flow counter on each iteration of the main I/O loop. That is, if one job has
2443 ``flow=8`` and another job has ``flow=-1``, then there will be a roughly 1:8
2444 ratio in how much one runs vs the other.
2446 .. option:: flow_watermark=int
2448 The maximum value that the absolute value of the flow counter is allowed to
2449 reach before the job must wait for a lower value of the counter.
2451 .. option:: flow_sleep=int
2453 The period of time, in microseconds, to wait after the flow watermark has
2454 been exceeded before retrying operations.
2456 .. option:: stonewall, wait_for_previous
2458 Wait for preceding jobs in the job file to exit, before starting this
2459 one. Can be used to insert serialization points in the job file. A stone
2460 wall also implies starting a new reporting group, see
2461 :option:`group_reporting`.
2465 By default, fio will continue running all other jobs when one job finishes
2466 but sometimes this is not the desired action. Setting ``exitall`` will
2467 instead make fio terminate all other jobs when one job finishes.
2469 .. option:: exec_prerun=str
2471 Before running this job, issue the command specified through
2472 :manpage:`system(3)`. Output is redirected in a file called
2473 :file:`jobname.prerun.txt`.
2475 .. option:: exec_postrun=str
2477 After the job completes, issue the command specified though
2478 :manpage:`system(3)`. Output is redirected in a file called
2479 :file:`jobname.postrun.txt`.
2483 Instead of running as the invoking user, set the user ID to this value
2484 before the thread/process does any work.
2488 Set group ID, see :option:`uid`.
2494 .. option:: verify_only
2496 Do not perform specified workload, only verify data still matches previous
2497 invocation of this workload. This option allows one to check data multiple
2498 times at a later date without overwriting it. This option makes sense only
2499 for workloads that write data, and does not support workloads with the
2500 :option:`time_based` option set.
2502 .. option:: do_verify=bool
2504 Run the verify phase after a write phase. Only valid if :option:`verify` is
2507 .. option:: verify=str
2509 If writing to a file, fio can verify the file contents after each iteration
2510 of the job. Each verification method also implies verification of special
2511 header, which is written to the beginning of each block. This header also
2512 includes meta information, like offset of the block, block number, timestamp
2513 when block was written, etc. :option:`verify` can be combined with
2514 :option:`verify_pattern` option. The allowed values are:
2517 Use an md5 sum of the data area and store it in the header of
2521 Use an experimental crc64 sum of the data area and store it in the
2522 header of each block.
2525 Use a crc32c sum of the data area and store it in the header of
2526 each block. This will automatically use hardware acceleration
2527 (e.g. SSE4.2 on an x86 or CRC crypto extensions on ARM64) but will
2528 fall back to software crc32c if none is found. Generally the
2529 fatest checksum fio supports when hardware accelerated.
2535 Use a crc32 sum of the data area and store it in the header of each
2539 Use a crc16 sum of the data area and store it in the header of each
2543 Use a crc7 sum of the data area and store it in the header of each
2547 Use xxhash as the checksum function. Generally the fastest software
2548 checksum that fio supports.
2551 Use sha512 as the checksum function.
2554 Use sha256 as the checksum function.
2557 Use optimized sha1 as the checksum function.
2560 Use optimized sha3-224 as the checksum function.
2563 Use optimized sha3-256 as the checksum function.
2566 Use optimized sha3-384 as the checksum function.
2569 Use optimized sha3-512 as the checksum function.
2572 This option is deprecated, since now meta information is included in
2573 generic verification header and meta verification happens by
2574 default. For detailed information see the description of the
2575 :option:`verify` setting. This option is kept because of
2576 compatibility's sake with old configurations. Do not use it.
2579 Verify a strict pattern. Normally fio includes a header with some
2580 basic information and checksumming, but if this option is set, only
2581 the specific pattern set with :option:`verify_pattern` is verified.
2584 Only pretend to verify. Useful for testing internals with
2585 :option:`ioengine`\=null, not for much else.
2587 This option can be used for repeated burn-in tests of a system to make sure
2588 that the written data is also correctly read back. If the data direction
2589 given is a read or random read, fio will assume that it should verify a
2590 previously written file. If the data direction includes any form of write,
2591 the verify will be of the newly written data.
2593 .. option:: verifysort=bool
2595 If true, fio will sort written verify blocks when it deems it faster to read
2596 them back in a sorted manner. This is often the case when overwriting an
2597 existing file, since the blocks are already laid out in the file system. You
2598 can ignore this option unless doing huge amounts of really fast I/O where
2599 the red-black tree sorting CPU time becomes significant. Default: true.
2601 .. option:: verifysort_nr=int
2603 Pre-load and sort verify blocks for a read workload.
2605 .. option:: verify_offset=int
2607 Swap the verification header with data somewhere else in the block before
2608 writing. It is swapped back before verifying.
2610 .. option:: verify_interval=int
2612 Write the verification header at a finer granularity than the
2613 :option:`blocksize`. It will be written for chunks the size of
2614 ``verify_interval``. :option:`blocksize` should divide this evenly.
2616 .. option:: verify_pattern=str
2618 If set, fio will fill the I/O buffers with this pattern. Fio defaults to
2619 filling with totally random bytes, but sometimes it's interesting to fill
2620 with a known pattern for I/O verification purposes. Depending on the width
2621 of the pattern, fio will fill 1/2/3/4 bytes of the buffer at the time (it can
2622 be either a decimal or a hex number). The ``verify_pattern`` if larger than
2623 a 32-bit quantity has to be a hex number that starts with either "0x" or
2624 "0X". Use with :option:`verify`. Also, ``verify_pattern`` supports %o
2625 format, which means that for each block offset will be written and then
2626 verified back, e.g.::
2630 Or use combination of everything::
2632 verify_pattern=0xff%o"abcd"-12
2634 .. option:: verify_fatal=bool
2636 Normally fio will keep checking the entire contents before quitting on a
2637 block verification failure. If this option is set, fio will exit the job on
2638 the first observed failure. Default: false.
2640 .. option:: verify_dump=bool
2642 If set, dump the contents of both the original data block and the data block
2643 we read off disk to files. This allows later analysis to inspect just what
2644 kind of data corruption occurred. Off by default.
2646 .. option:: verify_async=int
2648 Fio will normally verify I/O inline from the submitting thread. This option
2649 takes an integer describing how many async offload threads to create for I/O
2650 verification instead, causing fio to offload the duty of verifying I/O
2651 contents to one or more separate threads. If using this offload option, even
2652 sync I/O engines can benefit from using an :option:`iodepth` setting higher
2653 than 1, as it allows them to have I/O in flight while verifies are running.
2654 Defaults to 0 async threads, i.e. verification is not asynchronous.
2656 .. option:: verify_async_cpus=str
2658 Tell fio to set the given CPU affinity on the async I/O verification
2659 threads. See :option:`cpus_allowed` for the format used.
2661 .. option:: verify_backlog=int
2663 Fio will normally verify the written contents of a job that utilizes verify
2664 once that job has completed. In other words, everything is written then
2665 everything is read back and verified. You may want to verify continually
2666 instead for a variety of reasons. Fio stores the meta data associated with
2667 an I/O block in memory, so for large verify workloads, quite a bit of memory
2668 would be used up holding this meta data. If this option is enabled, fio will
2669 write only N blocks before verifying these blocks.
2671 .. option:: verify_backlog_batch=int
2673 Control how many blocks fio will verify if :option:`verify_backlog` is
2674 set. If not set, will default to the value of :option:`verify_backlog`
2675 (meaning the entire queue is read back and verified). If
2676 ``verify_backlog_batch`` is less than :option:`verify_backlog` then not all
2677 blocks will be verified, if ``verify_backlog_batch`` is larger than
2678 :option:`verify_backlog`, some blocks will be verified more than once.
2680 .. option:: verify_state_save=bool
2682 When a job exits during the write phase of a verify workload, save its
2683 current state. This allows fio to replay up until that point, if the verify
2684 state is loaded for the verify read phase. The format of the filename is,
2687 <type>-<jobname>-<jobindex>-verify.state.
2689 <type> is "local" for a local run, "sock" for a client/server socket
2690 connection, and "ip" (192.168.0.1, for instance) for a networked
2691 client/server connection. Defaults to true.
2693 .. option:: verify_state_load=bool
2695 If a verify termination trigger was used, fio stores the current write state
2696 of each thread. This can be used at verification time so that fio knows how
2697 far it should verify. Without this information, fio will run a full
2698 verification pass, according to the settings in the job file used. Default
2701 .. option:: trim_percentage=int
2703 Number of verify blocks to discard/trim.
2705 .. option:: trim_verify_zero=bool
2707 Verify that trim/discarded blocks are returned as zeros.
2709 .. option:: trim_backlog=int
2711 Trim after this number of blocks are written.
2713 .. option:: trim_backlog_batch=int
2715 Trim this number of I/O blocks.
2717 .. option:: experimental_verify=bool
2719 Enable experimental verification.
2724 .. option:: steadystate=str:float, ss=str:float
2726 Define the criterion and limit for assessing steady state performance. The
2727 first parameter designates the criterion whereas the second parameter sets
2728 the threshold. When the criterion falls below the threshold for the
2729 specified duration, the job will stop. For example, `iops_slope:0.1%` will
2730 direct fio to terminate the job when the least squares regression slope
2731 falls below 0.1% of the mean IOPS. If :option:`group_reporting` is enabled
2732 this will apply to all jobs in the group. Below is the list of available
2733 steady state assessment criteria. All assessments are carried out using only
2734 data from the rolling collection window. Threshold limits can be expressed
2735 as a fixed value or as a percentage of the mean in the collection window.
2738 Collect IOPS data. Stop the job if all individual IOPS measurements
2739 are within the specified limit of the mean IOPS (e.g., ``iops:2``
2740 means that all individual IOPS values must be within 2 of the mean,
2741 whereas ``iops:0.2%`` means that all individual IOPS values must be
2742 within 0.2% of the mean IOPS to terminate the job).
2745 Collect IOPS data and calculate the least squares regression
2746 slope. Stop the job if the slope falls below the specified limit.
2749 Collect bandwidth data. Stop the job if all individual bandwidth
2750 measurements are within the specified limit of the mean bandwidth.
2753 Collect bandwidth data and calculate the least squares regression
2754 slope. Stop the job if the slope falls below the specified limit.
2756 .. option:: steadystate_duration=time, ss_dur=time
2758 A rolling window of this duration will be used to judge whether steady state
2759 has been reached. Data will be collected once per second. The default is 0
2760 which disables steady state detection. When the unit is omitted, the
2761 value is interpreted in seconds.
2763 .. option:: steadystate_ramp_time=time, ss_ramp=time
2765 Allow the job to run for the specified duration before beginning data
2766 collection for checking the steady state job termination criterion. The
2767 default is 0. When the unit is omitted, the value is interpreted in seconds.
2770 Measurements and reporting
2771 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2773 .. option:: per_job_logs=bool
2775 If set, this generates bw/clat/iops log with per file private filenames. If
2776 not set, jobs with identical names will share the log filename. Default:
2779 .. option:: group_reporting
2781 It may sometimes be interesting to display statistics for groups of jobs as
2782 a whole instead of for each individual job. This is especially true if
2783 :option:`numjobs` is used; looking at individual thread/process output
2784 quickly becomes unwieldy. To see the final report per-group instead of
2785 per-job, use :option:`group_reporting`. Jobs in a file will be part of the
2786 same reporting group, unless if separated by a :option:`stonewall`, or by
2787 using :option:`new_group`.
2789 .. option:: new_group
2791 Start a new reporting group. See: :option:`group_reporting`. If not given,
2792 all jobs in a file will be part of the same reporting group, unless
2793 separated by a :option:`stonewall`.
2795 .. option:: stats=bool
2797 By default, fio collects and shows final output results for all jobs
2798 that run. If this option is set to 0, then fio will ignore it in
2799 the final stat output.
2801 .. option:: write_bw_log=str
2803 If given, write a bandwidth log for this job. Can be used to store data of
2804 the bandwidth of the jobs in their lifetime.
2806 If no str argument is given, the default filename of
2807 :file:`jobname_type.x.log` is used. Even when the argument is given, fio
2808 will still append the type of log. So if one specifies::
2812 The actual log name will be :file:`foo_bw.x.log` where `x` is the index
2813 of the job (`1..N`, where `N` is the number of jobs). If
2814 :option:`per_job_logs` is false, then the filename will not include the
2817 The included :command:`fio_generate_plots` script uses :command:`gnuplot` to turn these
2818 text files into nice graphs. See `Log File Formats`_ for how data is
2819 structured within the file.
2821 .. option:: write_lat_log=str
2823 Same as :option:`write_bw_log`, except this option creates I/O
2824 submission (e.g., :file:`name_slat.x.log`), completion (e.g.,
2825 :file:`name_clat.x.log`), and total (e.g., :file:`name_lat.x.log`)
2826 latency files instead. See :option:`write_bw_log` for details about
2827 the filename format and `Log File Formats`_ for how data is structured
2830 .. option:: write_hist_log=str
2832 Same as :option:`write_bw_log` but writes an I/O completion latency
2833 histogram file (e.g., :file:`name_hist.x.log`) instead. Note that this
2834 file will be empty unless :option:`log_hist_msec` has also been set.
2835 See :option:`write_bw_log` for details about the filename format and
2836 `Log File Formats`_ for how data is structured within the file.
2838 .. option:: write_iops_log=str
2840 Same as :option:`write_bw_log`, but writes an IOPS file (e.g.
2841 :file:`name_iops.x.log`) instead. See :option:`write_bw_log` for
2842 details about the filename format and `Log File Formats`_ for how data
2843 is structured within the file.
2845 .. option:: log_avg_msec=int
2847 By default, fio will log an entry in the iops, latency, or bw log for every
2848 I/O that completes. When writing to the disk log, that can quickly grow to a
2849 very large size. Setting this option makes fio average the each log entry
2850 over the specified period of time, reducing the resolution of the log. See
2851 :option:`log_max_value` as well. Defaults to 0, logging all entries.
2852 Also see `Log File Formats`_.
2854 .. option:: log_hist_msec=int
2856 Same as :option:`log_avg_msec`, but logs entries for completion latency
2857 histograms. Computing latency percentiles from averages of intervals using
2858 :option:`log_avg_msec` is inaccurate. Setting this option makes fio log
2859 histogram entries over the specified period of time, reducing log sizes for
2860 high IOPS devices while retaining percentile accuracy. See
2861 :option:`log_hist_coarseness` and :option:`write_hist_log` as well.
2862 Defaults to 0, meaning histogram logging is disabled.
2864 .. option:: log_hist_coarseness=int
2866 Integer ranging from 0 to 6, defining the coarseness of the resolution of
2867 the histogram logs enabled with :option:`log_hist_msec`. For each increment
2868 in coarseness, fio outputs half as many bins. Defaults to 0, for which
2869 histogram logs contain 1216 latency bins. See :option:`write_hist_log`
2870 and `Log File Formats`_.
2872 .. option:: log_max_value=bool
2874 If :option:`log_avg_msec` is set, fio logs the average over that window. If
2875 you instead want to log the maximum value, set this option to 1. Defaults to
2876 0, meaning that averaged values are logged.
2878 .. option:: log_offset=bool
2880 If this is set, the iolog options will include the byte offset for the I/O
2881 entry as well as the other data values. Defaults to 0 meaning that
2882 offsets are not present in logs. Also see `Log File Formats`_.
2884 .. option:: log_compression=int
2886 If this is set, fio will compress the I/O logs as it goes, to keep the
2887 memory footprint lower. When a log reaches the specified size, that chunk is
2888 removed and compressed in the background. Given that I/O logs are fairly
2889 highly compressible, this yields a nice memory savings for longer runs. The
2890 downside is that the compression will consume some background CPU cycles, so
2891 it may impact the run. This, however, is also true if the logging ends up
2892 consuming most of the system memory. So pick your poison. The I/O logs are
2893 saved normally at the end of a run, by decompressing the chunks and storing
2894 them in the specified log file. This feature depends on the availability of
2897 .. option:: log_compression_cpus=str
2899 Define the set of CPUs that are allowed to handle online log compression for
2900 the I/O jobs. This can provide better isolation between performance
2901 sensitive jobs, and background compression work.
2903 .. option:: log_store_compressed=bool
2905 If set, fio will store the log files in a compressed format. They can be
2906 decompressed with fio, using the :option:`--inflate-log` command line
2907 parameter. The files will be stored with a :file:`.fz` suffix.
2909 .. option:: log_unix_epoch=bool
2911 If set, fio will log Unix timestamps to the log files produced by enabling
2912 write_type_log for each log type, instead of the default zero-based
2915 .. option:: block_error_percentiles=bool
2917 If set, record errors in trim block-sized units from writes and trims and
2918 output a histogram of how many trims it took to get to errors, and what kind
2919 of error was encountered.
2921 .. option:: bwavgtime=int
2923 Average the calculated bandwidth over the given time. Value is specified in
2924 milliseconds. If the job also does bandwidth logging through
2925 :option:`write_bw_log`, then the minimum of this option and
2926 :option:`log_avg_msec` will be used. Default: 500ms.
2928 .. option:: iopsavgtime=int
2930 Average the calculated IOPS over the given time. Value is specified in
2931 milliseconds. If the job also does IOPS logging through
2932 :option:`write_iops_log`, then the minimum of this option and
2933 :option:`log_avg_msec` will be used. Default: 500ms.
2935 .. option:: disk_util=bool
2937 Generate disk utilization statistics, if the platform supports it.
2940 .. option:: disable_lat=bool
2942 Disable measurements of total latency numbers. Useful only for cutting back
2943 the number of calls to :manpage:`gettimeofday(2)`, as that does impact
2944 performance at really high IOPS rates. Note that to really get rid of a
2945 large amount of these calls, this option must be used with
2946 :option:`disable_slat` and :option:`disable_bw_measurement` as well.
2948 .. option:: disable_clat=bool
2950 Disable measurements of completion latency numbers. See
2951 :option:`disable_lat`.
2953 .. option:: disable_slat=bool
2955 Disable measurements of submission latency numbers. See
2956 :option:`disable_lat`.
2958 .. option:: disable_bw_measurement=bool, disable_bw=bool
2960 Disable measurements of throughput/bandwidth numbers. See
2961 :option:`disable_lat`.
2963 .. option:: clat_percentiles=bool
2965 Enable the reporting of percentiles of completion latencies. This
2966 option is mutually exclusive with :option:`lat_percentiles`.
2968 .. option:: lat_percentiles=bool
2970 Enable the reporting of percentiles of I/O latencies. This is similar
2971 to :option:`clat_percentiles`, except that this includes the
2972 submission latency. This option is mutually exclusive with
2973 :option:`clat_percentiles`.
2975 .. option:: percentile_list=float_list
2977 Overwrite the default list of percentiles for completion latencies and
2978 the block error histogram. Each number is a floating number in the
2979 range (0,100], and the maximum length of the list is 20. Use ``:`` to
2980 separate the numbers, and list the numbers in ascending order. For
2981 example, ``--percentile_list=99.5:99.9`` will cause fio to report the
2982 values of completion latency below which 99.5% and 99.9% of the observed
2983 latencies fell, respectively.
2985 .. option:: significant_figures=int
2987 If using :option:`--output-format` of `normal`, set the significant
2988 figures to this value. Higher values will yield more precise IOPS and
2989 throughput units, while lower values will round. Requires a minimum
2990 value of 1 and a maximum value of 10. Defaults to 4.
2996 .. option:: exitall_on_error
2998 When one job finishes in error, terminate the rest. The default is to wait
2999 for each job to finish.
3001 .. option:: continue_on_error=str
3003 Normally fio will exit the job on the first observed failure. If this option
3004 is set, fio will continue the job when there is a 'non-fatal error' (EIO or
3005 EILSEQ) until the runtime is exceeded or the I/O size specified is
3006 completed. If this option is used, there are two more stats that are
3007 appended, the total error count and the first error. The error field given
3008 in the stats is the first error that was hit during the run.
3010 The allowed values are:
3013 Exit on any I/O or verify errors.
3016 Continue on read errors, exit on all others.
3019 Continue on write errors, exit on all others.
3022 Continue on any I/O error, exit on all others.
3025 Continue on verify errors, exit on all others.
3028 Continue on all errors.
3031 Backward-compatible alias for 'none'.
3034 Backward-compatible alias for 'all'.
3036 .. option:: ignore_error=str
3038 Sometimes you want to ignore some errors during test in that case you can
3039 specify error list for each error type, instead of only being able to
3040 ignore the default 'non-fatal error' using :option:`continue_on_error`.
3041 ``ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST`` errors for
3042 given error type is separated with ':'. Error may be symbol ('ENOSPC',
3043 'ENOMEM') or integer. Example::
3045 ignore_error=EAGAIN,ENOSPC:122
3047 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from
3048 WRITE. This option works by overriding :option:`continue_on_error` with
3049 the list of errors for each error type if any.
3051 .. option:: error_dump=bool
3053 If set dump every error even if it is non fatal, true by default. If
3054 disabled only fatal error will be dumped.
3056 Running predefined workloads
3057 ----------------------------
3059 Fio includes predefined profiles that mimic the I/O workloads generated by
3062 .. option:: profile=str
3064 The predefined workload to run. Current profiles are:
3067 Threaded I/O bench (tiotest/tiobench) like workload.
3070 Aerospike Certification Tool (ACT) like workload.
3072 To view a profile's additional options use :option:`--cmdhelp` after specifying
3073 the profile. For example::
3075 $ fio --profile=act --cmdhelp
3080 .. option:: device-names=str
3085 .. option:: load=int
3088 ACT load multiplier. Default: 1.
3090 .. option:: test-duration=time
3093 How long the entire test takes to run. When the unit is omitted, the value
3094 is given in seconds. Default: 24h.
3096 .. option:: threads-per-queue=int
3099 Number of read I/O threads per device. Default: 8.
3101 .. option:: read-req-num-512-blocks=int
3104 Number of 512B blocks to read at the time. Default: 3.
3106 .. option:: large-block-op-kbytes=int
3109 Size of large block ops in KiB (writes). Default: 131072.
3114 Set to run ACT prep phase.
3116 Tiobench profile options
3117 ~~~~~~~~~~~~~~~~~~~~~~~~
3119 .. option:: size=str
3124 .. option:: block=int
3127 Block size in bytes. Default: 4096.
3129 .. option:: numruns=int
3139 .. option:: threads=int
3144 Interpreting the output
3145 -----------------------
3148 Example output was based on the following:
3149 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --time_based \
3150 --rate=1256k --bs=14K --name=quick --runtime=1s --name=mixed \
3151 --runtime=2m --rw=rw
3153 Fio spits out a lot of output. While running, fio will display the status of the
3154 jobs created. An example of that would be::
3156 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]
3158 The characters inside the first set of square brackets denote the current status of
3159 each thread. The first character is the first job defined in the job file, and so
3160 forth. The possible values (in typical life cycle order) are:
3162 +------+-----+-----------------------------------------------------------+
3164 +======+=====+===========================================================+
3165 | P | | Thread setup, but not started. |
3166 +------+-----+-----------------------------------------------------------+
3167 | C | | Thread created. |
3168 +------+-----+-----------------------------------------------------------+
3169 | I | | Thread initialized, waiting or generating necessary data. |
3170 +------+-----+-----------------------------------------------------------+
3171 | | p | Thread running pre-reading file(s). |
3172 +------+-----+-----------------------------------------------------------+
3173 | | / | Thread is in ramp period. |
3174 +------+-----+-----------------------------------------------------------+
3175 | | R | Running, doing sequential reads. |
3176 +------+-----+-----------------------------------------------------------+
3177 | | r | Running, doing random reads. |
3178 +------+-----+-----------------------------------------------------------+
3179 | | W | Running, doing sequential writes. |
3180 +------+-----+-----------------------------------------------------------+
3181 | | w | Running, doing random writes. |
3182 +------+-----+-----------------------------------------------------------+
3183 | | M | Running, doing mixed sequential reads/writes. |
3184 +------+-----+-----------------------------------------------------------+
3185 | | m | Running, doing mixed random reads/writes. |
3186 +------+-----+-----------------------------------------------------------+
3187 | | D | Running, doing sequential trims. |
3188 +------+-----+-----------------------------------------------------------+
3189 | | d | Running, doing random trims. |
3190 +------+-----+-----------------------------------------------------------+
3191 | | F | Running, currently waiting for :manpage:`fsync(2)`. |
3192 +------+-----+-----------------------------------------------------------+
3193 | | V | Running, doing verification of written data. |
3194 +------+-----+-----------------------------------------------------------+
3195 | f | | Thread finishing. |
3196 +------+-----+-----------------------------------------------------------+
3197 | E | | Thread exited, not reaped by main thread yet. |
3198 +------+-----+-----------------------------------------------------------+
3199 | _ | | Thread reaped. |
3200 +------+-----+-----------------------------------------------------------+
3201 | X | | Thread reaped, exited with an error. |
3202 +------+-----+-----------------------------------------------------------+
3203 | K | | Thread reaped, exited due to signal. |
3204 +------+-----+-----------------------------------------------------------+
3207 Example output was based on the following:
3208 TZ=UTC fio --iodepth=8 --ioengine=null --size=100M --runtime=58m \
3209 --time_based --rate=2512k --bs=256K --numjobs=10 \
3210 --name=readers --rw=read --name=writers --rw=write
3212 Fio will condense the thread string as not to take up more space on the command
3213 line than needed. For instance, if you have 10 readers and 10 writers running,
3214 the output would look like this::
3216 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]
3218 Note that the status string is displayed in order, so it's possible to tell which of
3219 the jobs are currently doing what. In the example above this means that jobs 1--10
3220 are readers and 11--20 are writers.
3222 The other values are fairly self explanatory -- number of threads currently
3223 running and doing I/O, the number of currently open files (f=), the estimated
3224 completion percentage, the rate of I/O since last check (read speed listed first,
3225 then write speed and optionally trim speed) in terms of bandwidth and IOPS,
3226 and time to completion for the current running group. It's impossible to estimate
3227 runtime of the following groups (if any).
3230 Example output was based on the following:
3231 TZ=UTC fio --iodepth=16 --ioengine=posixaio --filename=/tmp/fiofile \
3232 --direct=1 --size=100M --time_based --runtime=50s --rate_iops=89 \
3233 --bs=7K --name=Client1 --rw=write
3235 When fio is done (or interrupted by :kbd:`Ctrl-C`), it will show the data for
3236 each thread, group of threads, and disks in that order. For each overall thread (or
3237 group) the output looks like::
3239 Client1: (groupid=0, jobs=1): err= 0: pid=16109: Sat Jun 24 12:07:54 2017
3240 write: IOPS=88, BW=623KiB/s (638kB/s)(30.4MiB/50032msec)
3241 slat (nsec): min=500, max=145500, avg=8318.00, stdev=4781.50
3242 clat (usec): min=170, max=78367, avg=4019.02, stdev=8293.31
3243 lat (usec): min=174, max=78375, avg=4027.34, stdev=8291.79
3244 clat percentiles (usec):
3245 | 1.00th=[ 302], 5.00th=[ 326], 10.00th=[ 343], 20.00th=[ 363],
3246 | 30.00th=[ 392], 40.00th=[ 404], 50.00th=[ 416], 60.00th=[ 445],
3247 | 70.00th=[ 816], 80.00th=[ 6718], 90.00th=[12911], 95.00th=[21627],
3248 | 99.00th=[43779], 99.50th=[51643], 99.90th=[68682], 99.95th=[72877],
3250 bw ( KiB/s): min= 532, max= 686, per=0.10%, avg=622.87, stdev=24.82, samples= 100
3251 iops : min= 76, max= 98, avg=88.98, stdev= 3.54, samples= 100
3252 lat (usec) : 250=0.04%, 500=64.11%, 750=4.81%, 1000=2.79%
3253 lat (msec) : 2=4.16%, 4=1.84%, 10=4.90%, 20=11.33%, 50=5.37%
3254 lat (msec) : 100=0.65%
3255 cpu : usr=0.27%, sys=0.18%, ctx=12072, majf=0, minf=21
3256 IO depths : 1=85.0%, 2=13.1%, 4=1.8%, 8=0.1%, 16=0.0%, 32=0.0%, >=64=0.0%
3257 submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3258 complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3259 issued rwt: total=0,4450,0, short=0,0,0, dropped=0,0,0
3260 latency : target=0, window=0, percentile=100.00%, depth=8
3262 The job name (or first job's name when using :option:`group_reporting`) is printed,
3263 along with the group id, count of jobs being aggregated, last error id seen (which
3264 is 0 when there are no errors), pid/tid of that thread and the time the job/group
3265 completed. Below are the I/O statistics for each data direction performed (showing
3266 writes in the example above). In the order listed, they denote:
3269 The string before the colon shows the I/O direction the statistics
3270 are for. **IOPS** is the average I/Os performed per second. **BW**
3271 is the average bandwidth rate shown as: value in power of 2 format
3272 (value in power of 10 format). The last two values show: (**total
3273 I/O performed** in power of 2 format / **runtime** of that thread).
3276 Submission latency (**min** being the minimum, **max** being the
3277 maximum, **avg** being the average, **stdev** being the standard
3278 deviation). This is the time it took to submit the I/O. For
3279 sync I/O this row is not displayed as the slat is really the
3280 completion latency (since queue/complete is one operation there).
3281 This value can be in nanoseconds, microseconds or milliseconds ---
3282 fio will choose the most appropriate base and print that (in the
3283 example above nanoseconds was the best scale). Note: in :option:`--minimal` mode
3284 latencies are always expressed in microseconds.
3287 Completion latency. Same names as slat, this denotes the time from
3288 submission to completion of the I/O pieces. For sync I/O, clat will
3289 usually be equal (or very close) to 0, as the time from submit to
3290 complete is basically just CPU time (I/O has already been done, see slat
3294 Total latency. Same names as slat and clat, this denotes the time from
3295 when fio created the I/O unit to completion of the I/O operation.
3298 Bandwidth statistics based on samples. Same names as the xlat stats,
3299 but also includes the number of samples taken (**samples**) and an
3300 approximate percentage of total aggregate bandwidth this thread
3301 received in its group (**per**). This last value is only really
3302 useful if the threads in this group are on the same disk, since they
3303 are then competing for disk access.
3306 IOPS statistics based on samples. Same names as bw.
3308 **lat (nsec/usec/msec)**
3309 The distribution of I/O completion latencies. This is the time from when
3310 I/O leaves fio and when it gets completed. Unlike the separate
3311 read/write/trim sections above, the data here and in the remaining
3312 sections apply to all I/Os for the reporting group. 250=0.04% means that
3313 0.04% of the I/Os completed in under 250us. 500=64.11% means that 64.11%
3314 of the I/Os required 250 to 499us for completion.
3317 CPU usage. User and system time, along with the number of context
3318 switches this thread went through, usage of system and user time, and
3319 finally the number of major and minor page faults. The CPU utilization
3320 numbers are averages for the jobs in that reporting group, while the
3321 context and fault counters are summed.
3324 The distribution of I/O depths over the job lifetime. The numbers are
3325 divided into powers of 2 and each entry covers depths from that value
3326 up to those that are lower than the next entry -- e.g., 16= covers
3327 depths from 16 to 31. Note that the range covered by a depth
3328 distribution entry can be different to the range covered by the
3329 equivalent submit/complete distribution entry.
3332 How many pieces of I/O were submitting in a single submit call. Each
3333 entry denotes that amount and below, until the previous entry -- e.g.,
3334 16=100% means that we submitted anywhere between 9 to 16 I/Os per submit
3335 call. Note that the range covered by a submit distribution entry can
3336 be different to the range covered by the equivalent depth distribution
3340 Like the above submit number, but for completions instead.
3343 The number of read/write/trim requests issued, and how many of them were
3347 These values are for :option:`latency_target` and related options. When
3348 these options are engaged, this section describes the I/O depth required
3349 to meet the specified latency target.
3352 Example output was based on the following:
3353 TZ=UTC fio --ioengine=null --iodepth=2 --size=100M --numjobs=2 \
3354 --rate_process=poisson --io_limit=32M --name=read --bs=128k \
3355 --rate=11M --name=write --rw=write --bs=2k --rate=700k
3357 After each client has been listed, the group statistics are printed. They
3358 will look like this::
3360 Run status group 0 (all jobs):
3361 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
3362 WRITE: bw=1231KiB/s (1261kB/s), 616KiB/s-621KiB/s (630kB/s-636kB/s), io=64.0MiB (67.1MB), run=52747-53223msec
3364 For each data direction it prints:
3367 Aggregate bandwidth of threads in this group followed by the
3368 minimum and maximum bandwidth of all the threads in this group.
3369 Values outside of brackets are power-of-2 format and those
3370 within are the equivalent value in a power-of-10 format.
3372 Aggregate I/O performed of all threads in this group. The
3373 format is the same as bw.
3375 The smallest and longest runtimes of the threads in this group.
3377 And finally, the disk statistics are printed. This is Linux specific. They will look like this::
3379 Disk stats (read/write):
3380 sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00%
3382 Each value is printed for both reads and writes, with reads first. The
3386 Number of I/Os performed by all groups.
3388 Number of merges performed by the I/O scheduler.
3390 Number of ticks we kept the disk busy.
3392 Total time spent in the disk queue.
3394 The disk utilization. A value of 100% means we kept the disk
3395 busy constantly, 50% would be a disk idling half of the time.
3397 It is also possible to get fio to dump the current output while it is running,
3398 without terminating the job. To do that, send fio the **USR1** signal. You can
3399 also get regularly timed dumps by using the :option:`--status-interval`
3400 parameter, or by creating a file in :file:`/tmp` named
3401 :file:`fio-dump-status`. If fio sees this file, it will unlink it and dump the
3402 current output status.
3408 For scripted usage where you typically want to generate tables or graphs of the
3409 results, fio can output the results in a semicolon separated format. The format
3410 is one long line of values, such as::
3412 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%
3413 A description of this job goes here.
3415 The job description (if provided) follows on a second line.
3417 To enable terse output, use the :option:`--minimal` or
3418 :option:`--output-format`\=terse command line options. The
3419 first value is the version of the terse output format. If the output has to be
3420 changed for some reason, this number will be incremented by 1 to signify that
3423 Split up, the format is as follows (comments in brackets denote when a
3424 field was introduced or whether it's specific to some terse version):
3428 terse version, fio version [v3], jobname, groupid, error
3432 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
3433 Submission latency: min, max, mean, stdev (usec)
3434 Completion latency: min, max, mean, stdev (usec)
3435 Completion latency percentiles: 20 fields (see below)
3436 Total latency: min, max, mean, stdev (usec)
3437 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
3438 IOPS [v5]: min, max, mean, stdev, number of samples
3444 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
3445 Submission latency: min, max, mean, stdev (usec)
3446 Completion latency: min, max, mean, stdev (usec)
3447 Completion latency percentiles: 20 fields (see below)
3448 Total latency: min, max, mean, stdev (usec)
3449 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
3450 IOPS [v5]: min, max, mean, stdev, number of samples
3452 TRIM status [all but version 3]:
3454 Fields are similar to READ/WRITE status.
3458 user, system, context switches, major faults, minor faults
3462 <=1, 2, 4, 8, 16, 32, >=64
3464 I/O latencies microseconds::
3466 <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
3468 I/O latencies milliseconds::
3470 <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
3472 Disk utilization [v3]::
3474 disk name, read ios, write ios, read merges, write merges, read ticks, write ticks,
3475 time spent in queue, disk utilization percentage
3477 Additional Info (dependent on continue_on_error, default off)::
3479 total # errors, first error code
3481 Additional Info (dependent on description being set)::
3485 Completion latency percentiles can be a grouping of up to 20 sets, so for the
3486 terse output fio writes all of them. Each field will look like this::
3490 which is the Xth percentile, and the `usec` latency associated with it.
3492 For `Disk utilization`, all disks used by fio are shown. So for each disk there
3493 will be a disk utilization section.
3495 Below is a single line containing short names for each of the fields in the
3496 minimal output v3, separated by semicolons::
3498 terse_version_3;fio_version;jobname;groupid;error;read_kb;read_bandwidth;read_iops;read_runtime_ms;read_slat_min;read_slat_max;read_slat_mean;read_slat_dev;read_clat_min;read_clat_max;read_clat_mean;read_clat_dev;read_clat_pct01;read_clat_pct02;read_clat_pct03;read_clat_pct04;read_clat_pct05;read_clat_pct06;read_clat_pct07;read_clat_pct08;read_clat_pct09;read_clat_pct10;read_clat_pct11;read_clat_pct12;read_clat_pct13;read_clat_pct14;read_clat_pct15;read_clat_pct16;read_clat_pct17;read_clat_pct18;read_clat_pct19;read_clat_pct20;read_tlat_min;read_lat_max;read_lat_mean;read_lat_dev;read_bw_min;read_bw_max;read_bw_agg_pct;read_bw_mean;read_bw_dev;write_kb;write_bandwidth;write_iops;write_runtime_ms;write_slat_min;write_slat_max;write_slat_mean;write_slat_dev;write_clat_min;write_clat_max;write_clat_mean;write_clat_dev;write_clat_pct01;write_clat_pct02;write_clat_pct03;write_clat_pct04;write_clat_pct05;write_clat_pct06;write_clat_pct07;write_clat_pct08;write_clat_pct09;write_clat_pct10;write_clat_pct11;write_clat_pct12;write_clat_pct13;write_clat_pct14;write_clat_pct15;write_clat_pct16;write_clat_pct17;write_clat_pct18;write_clat_pct19;write_clat_pct20;write_tlat_min;write_lat_max;write_lat_mean;write_lat_dev;write_bw_min;write_bw_max;write_bw_agg_pct;write_bw_mean;write_bw_dev;cpu_user;cpu_sys;cpu_csw;cpu_mjf;cpu_minf;iodepth_1;iodepth_2;iodepth_4;iodepth_8;iodepth_16;iodepth_32;iodepth_64;lat_2us;lat_4us;lat_10us;lat_20us;lat_50us;lat_100us;lat_250us;lat_500us;lat_750us;lat_1000us;lat_2ms;lat_4ms;lat_10ms;lat_20ms;lat_50ms;lat_100ms;lat_250ms;lat_500ms;lat_750ms;lat_1000ms;lat_2000ms;lat_over_2000ms;disk_name;disk_read_iops;disk_write_iops;disk_read_merges;disk_write_merges;disk_read_ticks;write_ticks;disk_queue_time;disk_util
3504 The `json` output format is intended to be both human readable and convenient
3505 for automated parsing. For the most part its sections mirror those of the
3506 `normal` output. The `runtime` value is reported in msec and the `bw` value is
3507 reported in 1024 bytes per second units.
3513 The `json+` output format is identical to the `json` output format except that it
3514 adds a full dump of the completion latency bins. Each `bins` object contains a
3515 set of (key, value) pairs where keys are latency durations and values count how
3516 many I/Os had completion latencies of the corresponding duration. For example,
3519 "bins" : { "87552" : 1, "89600" : 1, "94720" : 1, "96768" : 1, "97792" : 1, "99840" : 1, "100864" : 2, "103936" : 6, "104960" : 534, "105984" : 5995, "107008" : 7529, ... }
3521 This data indicates that one I/O required 87,552ns to complete, two I/Os required
3522 100,864ns to complete, and 7529 I/Os required 107,008ns to complete.
3524 Also included with fio is a Python script `fio_jsonplus_clat2csv` that takes
3525 json+ output and generates CSV-formatted latency data suitable for plotting.
3527 The latency durations actually represent the midpoints of latency intervals.
3528 For details refer to :file:`stat.h`.
3534 There are two trace file format that you can encounter. The older (v1) format is
3535 unsupported since version 1.20-rc3 (March 2008). It will still be described
3536 below in case that you get an old trace and want to understand it.
3538 In any case the trace is a simple text file with a single action per line.
3541 Trace file format v1
3542 ~~~~~~~~~~~~~~~~~~~~
3544 Each line represents a single I/O action in the following format::
3548 where `rw=0/1` for read/write, and the `offset` and `length` entries being in bytes.
3550 This format is not supported in fio versions >= 1.20-rc3.
3553 Trace file format v2
3554 ~~~~~~~~~~~~~~~~~~~~
3556 The second version of the trace file format was added in fio version 1.17. It
3557 allows to access more then one file per trace and has a bigger set of possible
3560 The first line of the trace file has to be::
3564 Following this can be lines in two different formats, which are described below.
3566 The file management format::
3570 The `filename` is given as an absolute path. The `action` can be one of these:
3573 Add the given `filename` to the trace.
3575 Open the file with the given `filename`. The `filename` has to have
3576 been added with the **add** action before.
3578 Close the file with the given `filename`. The file has to have been
3582 The file I/O action format::
3584 filename action offset length
3586 The `filename` is given as an absolute path, and has to have been added and
3587 opened before it can be used with this format. The `offset` and `length` are
3588 given in bytes. The `action` can be one of these:
3591 Wait for `offset` microseconds. Everything below 100 is discarded.
3592 The time is relative to the previous `wait` statement.
3594 Read `length` bytes beginning from `offset`.
3596 Write `length` bytes beginning from `offset`.
3598 :manpage:`fsync(2)` the file.
3600 :manpage:`fdatasync(2)` the file.
3602 Trim the given file from the given `offset` for `length` bytes.
3604 CPU idleness profiling
3605 ----------------------
3607 In some cases, we want to understand CPU overhead in a test. For example, we
3608 test patches for the specific goodness of whether they reduce CPU usage.
3609 Fio implements a balloon approach to create a thread per CPU that runs at idle
3610 priority, meaning that it only runs when nobody else needs the cpu.
3611 By measuring the amount of work completed by the thread, idleness of each CPU
3612 can be derived accordingly.
3614 An unit work is defined as touching a full page of unsigned characters. Mean and
3615 standard deviation of time to complete an unit work is reported in "unit work"
3616 section. Options can be chosen to report detailed percpu idleness or overall
3617 system idleness by aggregating percpu stats.
3620 Verification and triggers
3621 -------------------------
3623 Fio is usually run in one of two ways, when data verification is done. The first
3624 is a normal write job of some sort with verify enabled. When the write phase has
3625 completed, fio switches to reads and verifies everything it wrote. The second
3626 model is running just the write phase, and then later on running the same job
3627 (but with reads instead of writes) to repeat the same I/O patterns and verify
3628 the contents. Both of these methods depend on the write phase being completed,
3629 as fio otherwise has no idea how much data was written.
3631 With verification triggers, fio supports dumping the current write state to
3632 local files. Then a subsequent read verify workload can load this state and know
3633 exactly where to stop. This is useful for testing cases where power is cut to a
3634 server in a managed fashion, for instance.
3636 A verification trigger consists of two things:
3638 1) Storing the write state of each job.
3639 2) Executing a trigger command.
3641 The write state is relatively small, on the order of hundreds of bytes to single
3642 kilobytes. It contains information on the number of completions done, the last X
3645 A trigger is invoked either through creation ('touch') of a specified file in
3646 the system, or through a timeout setting. If fio is run with
3647 :option:`--trigger-file`\= :file:`/tmp/trigger-file`, then it will continually
3648 check for the existence of :file:`/tmp/trigger-file`. When it sees this file, it
3649 will fire off the trigger (thus saving state, and executing the trigger
3652 For client/server runs, there's both a local and remote trigger. If fio is
3653 running as a server backend, it will send the job states back to the client for
3654 safe storage, then execute the remote trigger, if specified. If a local trigger
3655 is specified, the server will still send back the write state, but the client
3656 will then execute the trigger.
3658 Verification trigger example
3659 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3661 Let's say we want to run a powercut test on the remote Linux machine 'server'.
3662 Our write workload is in :file:`write-test.fio`. We want to cut power to 'server' at
3663 some point during the run, and we'll run this test from the safety or our local
3664 machine, 'localbox'. On the server, we'll start the fio backend normally::
3666 server# fio --server
3668 and on the client, we'll fire off the workload::
3670 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger-remote="bash -c \"echo b > /proc/sysrq-triger\""
3672 We set :file:`/tmp/my-trigger` as the trigger file, and we tell fio to execute::
3674 echo b > /proc/sysrq-trigger
3676 on the server once it has received the trigger and sent us the write state. This
3677 will work, but it's not **really** cutting power to the server, it's merely
3678 abruptly rebooting it. If we have a remote way of cutting power to the server
3679 through IPMI or similar, we could do that through a local trigger command
3680 instead. Let's assume we have a script that does IPMI reboot of a given hostname,
3681 ipmi-reboot. On localbox, we could then have run fio with a local trigger
3684 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger="ipmi-reboot server"
3686 For this case, fio would wait for the server to send us the write state, then
3687 execute ``ipmi-reboot server`` when that happened.
3689 Loading verify state
3690 ~~~~~~~~~~~~~~~~~~~~
3692 To load stored write state, a read verification job file must contain the
3693 :option:`verify_state_load` option. If that is set, fio will load the previously
3694 stored state. For a local fio run this is done by loading the files directly,
3695 and on a client/server run, the server backend will ask the client to send the
3696 files over and load them from there.
3702 Fio supports a variety of log file formats, for logging latencies, bandwidth,
3703 and IOPS. The logs share a common format, which looks like this:
3705 *time* (`msec`), *value*, *data direction*, *block size* (`bytes`),
3708 *Time* for the log entry is always in milliseconds. The *value* logged depends
3709 on the type of log, it will be one of the following:
3712 Value is latency in nsecs
3718 *Data direction* is one of the following:
3727 The entry's *block size* is always in bytes. The *offset* is the offset, in bytes,
3728 from the start of the file, for that particular I/O. The logging of the offset can be
3729 toggled with :option:`log_offset`.
3731 Fio defaults to logging every individual I/O. When IOPS are logged for individual
3732 I/Os the *value* entry will always be 1. If windowed logging is enabled through
3733 :option:`log_avg_msec`, fio logs the average values over the specified period of time.
3734 If windowed logging is enabled and :option:`log_max_value` is set, then fio logs
3735 maximum values in that window instead of averages. Since *data direction*, *block
3736 size* and *offset* are per-I/O values, if windowed logging is enabled they
3737 aren't applicable and will be 0.
3742 Normally fio is invoked as a stand-alone application on the machine where the
3743 I/O workload should be generated. However, the backend and frontend of fio can
3744 be run separately i.e., the fio server can generate an I/O workload on the "Device
3745 Under Test" while being controlled by a client on another machine.
3747 Start the server on the machine which has access to the storage DUT::
3751 where `args` defines what fio listens to. The arguments are of the form
3752 ``type,hostname`` or ``IP,port``. *type* is either ``ip`` (or ip4) for TCP/IP
3753 v4, ``ip6`` for TCP/IP v6, or ``sock`` for a local unix domain socket.
3754 *hostname* is either a hostname or IP address, and *port* is the port to listen
3755 to (only valid for TCP/IP, not a local socket). Some examples:
3759 Start a fio server, listening on all interfaces on the default port (8765).
3761 2) ``fio --server=ip:hostname,4444``
3763 Start a fio server, listening on IP belonging to hostname and on port 4444.
3765 3) ``fio --server=ip6:::1,4444``
3767 Start a fio server, listening on IPv6 localhost ::1 and on port 4444.
3769 4) ``fio --server=,4444``
3771 Start a fio server, listening on all interfaces on port 4444.
3773 5) ``fio --server=1.2.3.4``
3775 Start a fio server, listening on IP 1.2.3.4 on the default port.
3777 6) ``fio --server=sock:/tmp/fio.sock``
3779 Start a fio server, listening on the local socket :file:`/tmp/fio.sock`.
3781 Once a server is running, a "client" can connect to the fio server with::
3783 fio <local-args> --client=<server> <remote-args> <job file(s)>
3785 where `local-args` are arguments for the client where it is running, `server`
3786 is the connect string, and `remote-args` and `job file(s)` are sent to the
3787 server. The `server` string follows the same format as it does on the server
3788 side, to allow IP/hostname/socket and port strings.
3790 Fio can connect to multiple servers this way::
3792 fio --client=<server1> <job file(s)> --client=<server2> <job file(s)>
3794 If the job file is located on the fio server, then you can tell the server to
3795 load a local file as well. This is done by using :option:`--remote-config` ::
3797 fio --client=server --remote-config /path/to/file.fio
3799 Then fio will open this local (to the server) job file instead of being passed
3800 one from the client.
3802 If you have many servers (example: 100 VMs/containers), you can input a pathname
3803 of a file containing host IPs/names as the parameter value for the
3804 :option:`--client` option. For example, here is an example :file:`host.list`
3805 file containing 2 hostnames::
3807 host1.your.dns.domain
3808 host2.your.dns.domain
3810 The fio command would then be::
3812 fio --client=host.list <job file(s)>
3814 In this mode, you cannot input server-specific parameters or job files -- all
3815 servers receive the same job file.
3817 In order to let ``fio --client`` runs use a shared filesystem from multiple
3818 hosts, ``fio --client`` now prepends the IP address of the server to the
3819 filename. For example, if fio is using the directory :file:`/mnt/nfs/fio` and is
3820 writing filename :file:`fileio.tmp`, with a :option:`--client` `hostfile`
3821 containing two hostnames ``h1`` and ``h2`` with IP addresses 192.168.10.120 and
3822 192.168.10.121, then fio will create two files::
3824 /mnt/nfs/fio/192.168.10.120.fileio.tmp
3825 /mnt/nfs/fio/192.168.10.121.fileio.tmp