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 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:: --bandwidth-log
109 Generate aggregate bandwidth logs.
111 .. option:: --minimal
113 Print statistics in a terse, semicolon-delimited format.
115 .. option:: --append-terse
117 Print statistics in selected mode AND terse, semicolon-delimited format.
118 **deprecated**, use :option:`--output-format` instead to select multiple
121 .. option:: --output-format=type
123 Set the reporting format to `normal`, `terse`, `json`, or `json+`. Multiple
124 formats can be selected, separated by a comma. `terse` is a CSV based
125 format. `json+` is like `json`, except it adds a full dump of the latency
128 .. option:: --terse-version=type
130 Set terse version output format (default 3, or 2 or 4 or 5).
132 .. option:: --version
134 Print version info 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 be passed, in
148 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 :option:`ioengine`, or print help for `command`
157 defined by :option:`ioengine`. If no :option:`ioengine` is given, list all
160 .. option:: --showcmd=jobfile
162 Turn a job file into 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 When real-time ETA estimate should be printed. May be `always`, `never` or
178 .. option:: --eta-newline=time
180 Force a new line for every `time` period passed. When the unit is omitted,
181 the value is interpreted in seconds.
183 .. option:: --status-interval=time
185 Force full status dump every `time` period passed. When the unit is
186 omitted, the value is interpreted in seconds.
188 .. option:: --section=name
190 Only run specified section in job file. Multiple sections can be specified.
191 The ``--section`` option allows one to combine related jobs into one file.
192 E.g. one job file could define light, moderate, and heavy sections. Tell
193 fio to run only the "heavy" section by giving ``--section=heavy``
194 command line option. One can also specify the "write" operations in one
195 section and "verify" operation in another section. The ``--section`` option
196 only applies to job sections. The reserved *global* section is always
199 .. option:: --alloc-size=kb
201 Set the internal smalloc pool to this size in KiB. The
202 ``--alloc-size`` switch allows one to use a larger pool size for smalloc.
203 If running large jobs with randommap enabled, fio can run out of memory.
204 Smalloc is an internal allocator for shared structures from a fixed size
205 memory pool and can grow to 16 pools. The pool size defaults to 16MiB.
207 NOTE: While running :file:`.fio_smalloc.*` backing store files are visible
210 .. option:: --warnings-fatal
212 All fio parser warnings are fatal, causing fio to exit with an
215 .. option:: --max-jobs=nr
217 Maximum number of threads/processes to support.
219 .. option:: --server=args
221 Start a backend server, with `args` specifying what to listen to.
222 See `Client/Server`_ section.
224 .. option:: --daemonize=pidfile
226 Background a fio server, writing the pid to the given `pidfile` file.
228 .. option:: --client=hostname
230 Instead of running the jobs locally, send and run them on the given host or
231 set of hosts. See `Client/Server`_ section.
233 .. option:: --remote-config=file
235 Tell fio server to load this local file.
237 .. option:: --idle-prof=option
239 Report CPU idleness. *option* is one of the following:
242 Run unit work calibration only and exit.
245 Show aggregate system idleness and unit work.
248 As **system** but also show per CPU idleness.
250 .. option:: --inflate-log=log
252 Inflate and output compressed log.
254 .. option:: --trigger-file=file
256 Execute trigger cmd when file exists.
258 .. option:: --trigger-timeout=t
260 Execute trigger at this time.
262 .. option:: --trigger=cmd
264 Set this command as local trigger.
266 .. option:: --trigger-remote=cmd
268 Set this command as remote trigger.
270 .. option:: --aux-path=path
272 Use this path for fio state generated files.
274 Any parameters following the options will be assumed to be job files, unless
275 they match a job file parameter. Multiple job files can be listed and each job
276 file will be regarded as a separate group. Fio will :option:`stonewall`
277 execution between each group.
283 As previously described, fio accepts one or more job files describing what it is
284 supposed to do. The job file format is the classic ini file, where the names
285 enclosed in [] brackets define the job name. You are free to use any ASCII name
286 you want, except *global* which has special meaning. Following the job name is
287 a sequence of zero or more parameters, one per line, that define the behavior of
288 the job. If the first character in a line is a ';' or a '#', the entire line is
289 discarded as a comment.
291 A *global* section sets defaults for the jobs described in that file. A job may
292 override a *global* section parameter, and a job file may even have several
293 *global* sections if so desired. A job is only affected by a *global* section
296 The :option:`--cmdhelp` option also lists all options. If used with an `option`
297 argument, :option:`--cmdhelp` will detail the given `option`.
299 See the `examples/` directory for inspiration on how to write job files. Note
300 the copyright and license requirements currently apply to `examples/` files.
302 So let's look at a really simple job file that defines two processes, each
303 randomly reading from a 128MiB file:
307 ; -- start job file --
318 As you can see, the job file sections themselves are empty as all the described
319 parameters are shared. As no :option:`filename` option is given, fio makes up a
320 `filename` for each of the jobs as it sees fit. On the command line, this job
321 would look as follows::
323 $ fio --name=global --rw=randread --size=128m --name=job1 --name=job2
326 Let's look at an example that has a number of processes writing randomly to
331 ; -- start job file --
342 Here we have no *global* section, as we only have one job defined anyway. We
343 want to use async I/O here, with a depth of 4 for each file. We also increased
344 the buffer size used to 32KiB and define numjobs to 4 to fork 4 identical
345 jobs. The result is 4 processes each randomly writing to their own 64MiB
346 file. Instead of using the above job file, you could have given the parameters
347 on the command line. For this case, you would specify::
349 $ fio --name=random-writers --ioengine=libaio --iodepth=4 --rw=randwrite --bs=32k --direct=0 --size=64m --numjobs=4
351 When fio is utilized as a basis of any reasonably large test suite, it might be
352 desirable to share a set of standardized settings across multiple job files.
353 Instead of copy/pasting such settings, any section may pull in an external
354 :file:`filename.fio` file with *include filename* directive, as in the following
357 ; -- start job file including.fio --
361 include glob-include.fio
368 include test-include.fio
369 ; -- end job file including.fio --
373 ; -- start job file glob-include.fio --
376 ; -- end job file glob-include.fio --
380 ; -- start job file test-include.fio --
383 ; -- end job file test-include.fio --
385 Settings pulled into a section apply to that section only (except *global*
386 section). Include directives may be nested in that any included file may contain
387 further include directive(s). Include files may not contain [] sections.
390 Environment variables
391 ~~~~~~~~~~~~~~~~~~~~~
393 Fio also supports environment variable expansion in job files. Any sub-string of
394 the form ``${VARNAME}`` as part of an option value (in other words, on the right
395 of the '='), will be expanded to the value of the environment variable called
396 `VARNAME`. If no such environment variable is defined, or `VARNAME` is the
397 empty string, the empty string will be substituted.
399 As an example, let's look at a sample fio invocation and job file::
401 $ SIZE=64m NUMJOBS=4 fio jobfile.fio
405 ; -- start job file --
412 This will expand to the following equivalent job file at runtime:
416 ; -- start job file --
423 Fio ships with a few example job files, you can also look there for inspiration.
428 Additionally, fio has a set of reserved keywords that will be replaced
429 internally with the appropriate value. Those keywords are:
433 The architecture page size of the running system.
437 Megabytes of total memory in the system.
441 Number of online available CPUs.
443 These can be used on the command line or in the job file, and will be
444 automatically substituted with the current system values when the job is
445 run. Simple math is also supported on these keywords, so you can perform actions
450 and get that properly expanded to 8 times the size of memory in the machine.
456 This section describes in details each parameter associated with a job. Some
457 parameters take an option of a given type, such as an integer or a
458 string. Anywhere a numeric value is required, an arithmetic expression may be
459 used, provided it is surrounded by parentheses. Supported operators are:
468 For time values in expressions, units are microseconds by default. This is
469 different than for time values not in expressions (not enclosed in
470 parentheses). The following types are used:
477 String. This is a sequence of alpha characters.
480 Integer with possible time suffix. Without a unit value is interpreted as
481 seconds unless otherwise specified. Accepts a suffix of 'd' for days, 'h' for
482 hours, 'm' for minutes, 's' for seconds, 'ms' (or 'msec') for milliseconds and
483 'us' (or 'usec') for microseconds. For example, use 10m for 10 minutes.
488 Integer. A whole number value, which may contain an integer prefix
489 and an integer suffix:
491 [*integer prefix*] **number** [*integer suffix*]
493 The optional *integer prefix* specifies the number's base. The default
494 is decimal. *0x* specifies hexadecimal.
496 The optional *integer suffix* specifies the number's units, and includes an
497 optional unit prefix and an optional unit. For quantities of data, the
498 default unit is bytes. For quantities of time, the default unit is seconds
499 unless otherwise specified.
501 With :option:`kb_base`\=1000, fio follows international standards for unit
502 prefixes. To specify power-of-10 decimal values defined in the
503 International System of Units (SI):
505 * *Ki* -- means kilo (K) or 1000
506 * *Mi* -- means mega (M) or 1000**2
507 * *Gi* -- means giga (G) or 1000**3
508 * *Ti* -- means tera (T) or 1000**4
509 * *Pi* -- means peta (P) or 1000**5
511 To specify power-of-2 binary values defined in IEC 80000-13:
513 * *k* -- means kibi (Ki) or 1024
514 * *M* -- means mebi (Mi) or 1024**2
515 * *G* -- means gibi (Gi) or 1024**3
516 * *T* -- means tebi (Ti) or 1024**4
517 * *P* -- means pebi (Pi) or 1024**5
519 With :option:`kb_base`\=1024 (the default), the unit prefixes are opposite
520 from those specified in the SI and IEC 80000-13 standards to provide
521 compatibility with old scripts. For example, 4k means 4096.
523 For quantities of data, an optional unit of 'B' may be included
524 (e.g., 'kB' is the same as 'k').
526 The *integer suffix* is not case sensitive (e.g., m/mi mean mebi/mega,
527 not milli). 'b' and 'B' both mean byte, not bit.
529 Examples with :option:`kb_base`\=1000:
531 * *4 KiB*: 4096, 4096b, 4096B, 4ki, 4kib, 4kiB, 4Ki, 4KiB
532 * *1 MiB*: 1048576, 1mi, 1024ki
533 * *1 MB*: 1000000, 1m, 1000k
534 * *1 TiB*: 1099511627776, 1ti, 1024gi, 1048576mi
535 * *1 TB*: 1000000000, 1t, 1000m, 1000000k
537 Examples with :option:`kb_base`\=1024 (default):
539 * *4 KiB*: 4096, 4096b, 4096B, 4k, 4kb, 4kB, 4K, 4KB
540 * *1 MiB*: 1048576, 1m, 1024k
541 * *1 MB*: 1000000, 1mi, 1000ki
542 * *1 TiB*: 1099511627776, 1t, 1024g, 1048576m
543 * *1 TB*: 1000000000, 1ti, 1000mi, 1000000ki
545 To specify times (units are not case sensitive):
549 * *M* -- means minutes
550 * *s* -- or sec means seconds (default)
551 * *ms* -- or *msec* means milliseconds
552 * *us* -- or *usec* means microseconds
554 If the option accepts an upper and lower range, use a colon ':' or
555 minus '-' to separate such values. See :ref:`irange <irange>`.
556 If the lower value specified happens to be larger than the upper value
557 the two values are swapped.
562 Boolean. Usually parsed as an integer, however only defined for
563 true and false (1 and 0).
568 Integer range with suffix. Allows value range to be given, such as
569 1024-4096. A colon may also be used as the separator, e.g. 1k:4k. If the
570 option allows two sets of ranges, they can be specified with a ',' or '/'
571 delimiter: 1k-4k/8k-32k. Also see :ref:`int <int>`.
574 A list of floating point numbers, separated by a ':' character.
580 .. option:: kb_base=int
582 Select the interpretation of unit prefixes in input parameters.
585 Inputs comply with IEC 80000-13 and the International
586 System of Units (SI). Use:
588 - power-of-2 values with IEC prefixes (e.g., KiB)
589 - power-of-10 values with SI prefixes (e.g., kB)
592 Compatibility mode (default). To avoid breaking old scripts:
594 - power-of-2 values with SI prefixes
595 - power-of-10 values with IEC prefixes
597 See :option:`bs` for more details on input parameters.
599 Outputs always use correct prefixes. Most outputs include both
602 bw=2383.3kB/s (2327.4KiB/s)
604 If only one value is reported, then kb_base selects the one to use:
606 **1000** -- SI prefixes
608 **1024** -- IEC prefixes
610 .. option:: unit_base=int
612 Base unit for reporting. Allowed values are:
615 Use auto-detection (default).
622 With the above in mind, here follows the complete list of fio job parameters.
630 ASCII name of the job. This may be used to override the name printed by fio
631 for this job. Otherwise the job name is used. On the command line this
632 parameter has the special purpose of also signaling the start of a new job.
634 .. option:: description=str
636 Text description of the job. Doesn't do anything except dump this text
637 description when this job is run. It's not parsed.
639 .. option:: loops=int
641 Run the specified number of iterations of this job. Used to repeat the same
642 workload a given number of times. Defaults to 1.
644 .. option:: numjobs=int
646 Create the specified number of clones of this job. Each clone of job
647 is spawned as an independent thread or process. May be used to setup a
648 larger number of threads/processes doing the same thing. Each thread is
649 reported separately; to see statistics for all clones as a whole, use
650 :option:`group_reporting` in conjunction with :option:`new_group`.
651 See :option:`--max-jobs`. Default: 1.
654 Time related parameters
655 ~~~~~~~~~~~~~~~~~~~~~~~
657 .. option:: runtime=time
659 Tell fio to terminate processing after the specified period of time. It
660 can be quite hard to determine for how long a specified job will run, so
661 this parameter is handy to cap the total runtime to a given time. When
662 the unit is omitted, the value is intepreted in seconds.
664 .. option:: time_based
666 If set, fio will run for the duration of the :option:`runtime` specified
667 even if the file(s) are completely read or written. It will simply loop over
668 the same workload as many times as the :option:`runtime` allows.
670 .. option:: startdelay=irange(time)
672 Delay the start of job for the specified amount of time. Can be a single
673 value or a range. When given as a range, each thread will choose a value
674 randomly from within the range. Value is in seconds if a unit is omitted.
676 .. option:: ramp_time=time
678 If set, fio will run the specified workload for this amount of time before
679 logging any performance numbers. Useful for letting performance settle
680 before logging results, thus minimizing the runtime required for stable
681 results. Note that the ``ramp_time`` is considered lead in time for a job,
682 thus it will increase the total runtime if a special timeout or
683 :option:`runtime` is specified. When the unit is omitted, the value is
686 .. option:: clocksource=str
688 Use the given clocksource as the base of timing. The supported options are:
691 :manpage:`gettimeofday(2)`
694 :manpage:`clock_gettime(2)`
697 Internal CPU clock source
699 cpu is the preferred clocksource if it is reliable, as it is very fast (and
700 fio is heavy on time calls). Fio will automatically use this clocksource if
701 it's supported and considered reliable on the system it is running on,
702 unless another clocksource is specifically set. For x86/x86-64 CPUs, this
703 means supporting TSC Invariant.
705 .. option:: gtod_reduce=bool
707 Enable all of the :manpage:`gettimeofday(2)` reducing options
708 (:option:`disable_clat`, :option:`disable_slat`, :option:`disable_bw_measurement`) plus
709 reduce precision of the timeout somewhat to really shrink the
710 :manpage:`gettimeofday(2)` call count. With this option enabled, we only do
711 about 0.4% of the :manpage:`gettimeofday(2)` calls we would have done if all
712 time keeping was enabled.
714 .. option:: gtod_cpu=int
716 Sometimes it's cheaper to dedicate a single thread of execution to just
717 getting the current time. Fio (and databases, for instance) are very
718 intensive on :manpage:`gettimeofday(2)` calls. With this option, you can set
719 one CPU aside for doing nothing but logging current time to a shared memory
720 location. Then the other threads/processes that run I/O workloads need only
721 copy that segment, instead of entering the kernel with a
722 :manpage:`gettimeofday(2)` call. The CPU set aside for doing these time
723 calls will be excluded from other uses. Fio will manually clear it from the
724 CPU mask of other jobs.
730 .. option:: directory=str
732 Prefix filenames with this directory. Used to place files in a different
733 location than :file:`./`. You can specify a number of directories by
734 separating the names with a ':' character. These directories will be
735 assigned equally distributed to job clones created by :option:`numjobs` as
736 long as they are using generated filenames. If specific `filename(s)` are
737 set fio will use the first listed directory, and thereby matching the
738 `filename` semantic which generates a file each clone if not specified, but
739 let all clones use the same if set.
741 See the :option:`filename` option for information on how to escape "``:``" and
742 "``\``" characters within the directory path itself.
744 .. option:: filename=str
746 Fio normally makes up a `filename` based on the job name, thread number, and
747 file number (see :option:`filename_format`). If you want to share files
748 between threads in a job or several
749 jobs with fixed file paths, specify a `filename` for each of them to override
750 the default. If the ioengine is file based, you can specify a number of files
751 by separating the names with a ':' colon. So if you wanted a job to open
752 :file:`/dev/sda` and :file:`/dev/sdb` as the two working files, you would use
753 ``filename=/dev/sda:/dev/sdb``. This also means that whenever this option is
754 specified, :option:`nrfiles` is ignored. The size of regular files specified
755 by this option will be :option:`size` divided by number of files unless an
756 explicit size is specified by :option:`filesize`.
758 Each colon and backslash in the wanted path must be escaped with a ``\``
759 character. For instance, if the path is :file:`/dev/dsk/foo@3,0:c` then you
760 would use ``filename=/dev/dsk/foo@3,0\:c`` and if the path is
761 :file:`F:\\filename` then you would use ``filename=F\:\\filename``.
763 On Windows, disk devices are accessed as :file:`\\\\.\\PhysicalDrive0` for
764 the first device, :file:`\\\\.\\PhysicalDrive1` for the second etc.
765 Note: Windows and FreeBSD prevent write access to areas
766 of the disk containing in-use data (e.g. filesystems).
768 The filename "`-`" is a reserved name, meaning *stdin* or *stdout*. Which
769 of the two depends on the read/write direction set.
771 .. option:: filename_format=str
773 If sharing multiple files between jobs, it is usually necessary to have fio
774 generate the exact names that you want. By default, fio will name a file
775 based on the default file format specification of
776 :file:`jobname.jobnumber.filenumber`. With this option, that can be
777 customized. Fio will recognize and replace the following keywords in this
781 The name of the worker thread or process.
783 The incremental number of the worker thread or process.
785 The incremental number of the file for that worker thread or
788 To have dependent jobs share a set of files, this option can be set to have
789 fio generate filenames that are shared between the two. For instance, if
790 :file:`testfiles.$filenum` is specified, file number 4 for any job will be
791 named :file:`testfiles.4`. The default of :file:`$jobname.$jobnum.$filenum`
792 will be used if no other format specifier is given.
794 .. option:: unique_filename=bool
796 To avoid collisions between networked clients, fio defaults to prefixing any
797 generated filenames (with a directory specified) with the source of the
798 client connecting. To disable this behavior, set this option to 0.
800 .. option:: opendir=str
802 Recursively open any files below directory `str`.
804 .. option:: lockfile=str
806 Fio defaults to not locking any files before it does I/O to them. If a file
807 or file descriptor is shared, fio can serialize I/O to that file to make the
808 end result consistent. This is usual for emulating real workloads that share
809 files. The lock modes are:
812 No locking. The default.
814 Only one thread or process may do I/O at a time, excluding all
817 Read-write locking on the file. Many readers may
818 access the file at the same time, but writes get exclusive access.
820 .. option:: nrfiles=int
822 Number of files to use for this job. Defaults to 1. The size of files
823 will be :option:`size` divided by this unless explicit size is specified by
824 :option:`filesize`. Files are created for each thread separately, and each
825 file will have a file number within its name by default, as explained in
826 :option:`filename` section.
829 .. option:: openfiles=int
831 Number of files to keep open at the same time. Defaults to the same as
832 :option:`nrfiles`, can be set smaller to limit the number simultaneous
835 .. option:: file_service_type=str
837 Defines how fio decides which file from a job to service next. The following
841 Choose a file at random.
844 Round robin over opened files. This is the default.
847 Finish one file before moving on to the next. Multiple files can
848 still be open depending on 'openfiles'.
851 Use a *Zipf* distribution to decide what file to access.
854 Use a *Pareto* distribution to decide what file to access.
857 Use a *Gaussian* (normal) distribution to decide what file to
860 For *random*, *roundrobin*, and *sequential*, a postfix can be appended to
861 tell fio how many I/Os to issue before switching to a new file. For example,
862 specifying ``file_service_type=random:8`` would cause fio to issue
863 8 I/Os before selecting a new file at random. For the non-uniform
864 distributions, a floating point postfix can be given to influence how the
865 distribution is skewed. See :option:`random_distribution` for a description
866 of how that would work.
868 .. option:: ioscheduler=str
870 Attempt to switch the device hosting the file to the specified I/O scheduler
873 .. option:: create_serialize=bool
875 If true, serialize the file creation for the jobs. This may be handy to
876 avoid interleaving of data files, which may greatly depend on the filesystem
877 used and even the number of processors in the system. Default: true.
879 .. option:: create_fsync=bool
881 :manpage:`fsync(2)` the data file after creation. This is the default.
883 .. option:: create_on_open=bool
885 If true, don't pre-create files but allow the job's open() to create a file
886 when it's time to do I/O. Default: false -- pre-create all necessary files
889 .. option:: create_only=bool
891 If true, fio will only run the setup phase of the job. If files need to be
892 laid out or updated on disk, only that will be done -- the actual job contents
893 are not executed. Default: false.
895 .. option:: allow_file_create=bool
897 If true, fio is permitted to create files as part of its workload. If this
898 option is false, then fio will error out if
899 the files it needs to use don't already exist. Default: true.
901 .. option:: allow_mounted_write=bool
903 If this isn't set, fio will abort jobs that are destructive (e.g. that write)
904 to what appears to be a mounted device or partition. This should help catch
905 creating inadvertently destructive tests, not realizing that the test will
906 destroy data on the mounted file system. Note that some platforms don't allow
907 writing against a mounted device regardless of this option. Default: false.
909 .. option:: pre_read=bool
911 If this is given, files will be pre-read into memory before starting the
912 given I/O operation. This will also clear the :option:`invalidate` flag,
913 since it is pointless to pre-read and then drop the cache. This will only
914 work for I/O engines that are seek-able, since they allow you to read the
915 same data multiple times. Thus it will not work on non-seekable I/O engines
916 (e.g. network, splice). Default: false.
918 .. option:: unlink=bool
920 Unlink the job files when done. Not the default, as repeated runs of that
921 job would then waste time recreating the file set again and again. Default:
924 .. option:: unlink_each_loop=bool
926 Unlink job files after each iteration or loop. Default: false.
928 .. option:: zonesize=int
930 Divide a file into zones of the specified size. See :option:`zoneskip`.
932 .. option:: zonerange=int
934 Give size of an I/O zone. See :option:`zoneskip`.
936 .. option:: zoneskip=int
938 Skip the specified number of bytes when :option:`zonesize` data has been
939 read. The two zone options can be used to only do I/O on zones of a file.
945 .. option:: direct=bool
947 If value is true, use non-buffered I/O. This is usually O_DIRECT. Note that
948 ZFS on Solaris doesn't support direct I/O. On Windows the synchronous
949 ioengines don't support direct I/O. Default: false.
951 .. option:: atomic=bool
953 If value is true, attempt to use atomic direct I/O. Atomic writes are
954 guaranteed to be stable once acknowledged by the operating system. Only
955 Linux supports O_ATOMIC right now.
957 .. option:: buffered=bool
959 If value is true, use buffered I/O. This is the opposite of the
960 :option:`direct` option. Defaults to true.
962 .. option:: readwrite=str, rw=str
964 Type of I/O pattern. Accepted values are:
971 Sequential trims (Linux block devices only).
977 Random trims (Linux block devices only).
979 Sequential mixed reads and writes.
981 Random mixed reads and writes.
983 Sequential trim+write sequences. Blocks will be trimmed first,
984 then the same blocks will be written to.
986 Fio defaults to read if the option is not specified. For the mixed I/O
987 types, the default is to split them 50/50. For certain types of I/O the
988 result may still be skewed a bit, since the speed may be different.
990 It is possible to specify the number of I/Os to do before getting a new
991 offset by appending ``:<nr>`` to the end of the string given. For a
992 random read, it would look like ``rw=randread:8`` for passing in an offset
993 modifier with a value of 8. If the suffix is used with a sequential I/O
994 pattern, then the *<nr>* value specified will be **added** to the generated
995 offset for each I/O turning sequential I/O into sequential I/O with holes.
996 For instance, using ``rw=write:4k`` will skip 4k for every write. Also see
997 the :option:`rw_sequencer` option.
999 .. option:: rw_sequencer=str
1001 If an offset modifier is given by appending a number to the ``rw=<str>``
1002 line, then this option controls how that number modifies the I/O offset
1003 being generated. Accepted values are:
1006 Generate sequential offset.
1008 Generate the same offset.
1010 ``sequential`` is only useful for random I/O, where fio would normally
1011 generate a new random offset for every I/O. If you append e.g. 8 to randread,
1012 you would get a new random offset for every 8 I/O's. The result would be a
1013 seek for only every 8 I/O's, instead of for every I/O. Use ``rw=randread:8``
1014 to specify that. As sequential I/O is already sequential, setting
1015 ``sequential`` for that would not result in any differences. ``identical``
1016 behaves in a similar fashion, except it sends the same offset 8 number of
1017 times before generating a new offset.
1019 .. option:: unified_rw_reporting=bool
1021 Fio normally reports statistics on a per data direction basis, meaning that
1022 reads, writes, and trims are accounted and reported separately. If this
1023 option is set fio sums the results and report them as "mixed" instead.
1025 .. option:: randrepeat=bool
1027 Seed the random number generator used for random I/O patterns in a
1028 predictable way so the pattern is repeatable across runs. Default: true.
1030 .. option:: allrandrepeat=bool
1032 Seed all random number generators in a predictable way so results are
1033 repeatable across runs. Default: false.
1035 .. option:: randseed=int
1037 Seed the random number generators based on this seed value, to be able to
1038 control what sequence of output is being generated. If not set, the random
1039 sequence depends on the :option:`randrepeat` setting.
1041 .. option:: fallocate=str
1043 Whether pre-allocation is performed when laying down files.
1044 Accepted values are:
1047 Do not pre-allocate space.
1050 Pre-allocate via :manpage:`posix_fallocate(3)`.
1053 Pre-allocate via :manpage:`fallocate(2)` with
1054 FALLOC_FL_KEEP_SIZE set.
1057 Backward-compatible alias for **none**.
1060 Backward-compatible alias for **posix**.
1062 May not be available on all supported platforms. **keep** is only available
1063 on Linux. If using ZFS on Solaris this must be set to **none** because ZFS
1064 doesn't support it. Default: **posix**.
1066 .. option:: fadvise_hint=str
1068 Use :manpage:`posix_fadvise(2)` to advise the kernel on what I/O patterns
1069 are likely to be issued. Accepted values are:
1072 Backwards-compatible hint for "no hint".
1075 Backwards compatible hint for "advise with fio workload type". This
1076 uses **FADV_RANDOM** for a random workload, and **FADV_SEQUENTIAL**
1077 for a sequential workload.
1080 Advise using **FADV_SEQUENTIAL**.
1083 Advise using **FADV_RANDOM**.
1085 .. option:: fadvise_stream=int
1087 Use :manpage:`posix_fadvise(2)` to advise the kernel what stream ID the
1088 writes issued belong to. Only supported on Linux. Note, this option may
1089 change going forward.
1091 .. option:: offset=int
1093 Start I/O at the provided offset in the file, given as either a fixed size or
1094 a percentage. If a percentage is given, the next ``blockalign``-ed offset
1095 will be used. Data before the given offset will not be touched. This
1096 effectively caps the file size at `real_size - offset`. Can be combined with
1097 :option:`size` to constrain the start and end range of the I/O workload.
1099 .. option:: offset_increment=int
1101 If this is provided, then the real offset becomes `offset + offset_increment
1102 * thread_number`, where the thread number is a counter that starts at 0 and
1103 is incremented for each sub-job (i.e. when :option:`numjobs` option is
1104 specified). This option is useful if there are several jobs which are
1105 intended to operate on a file in parallel disjoint segments, with even
1106 spacing between the starting points.
1108 .. option:: number_ios=int
1110 Fio will normally perform I/Os until it has exhausted the size of the region
1111 set by :option:`size`, or if it exhaust the allocated time (or hits an error
1112 condition). With this setting, the range/size can be set independently of
1113 the number of I/Os to perform. When fio reaches this number, it will exit
1114 normally and report status. Note that this does not extend the amount of I/O
1115 that will be done, it will only stop fio if this condition is met before
1116 other end-of-job criteria.
1118 .. option:: fsync=int
1120 If writing to a file, issue an :manpage:`fsync(2)` (or its equivalent) of
1121 the dirty data for every number of blocks given. For example, if you give 32
1122 as a parameter, fio will sync the file after every 32 writes issued. If fio is
1123 using non-buffered I/O, we may not sync the file. The exception is the sg
1124 I/O engine, which synchronizes the disk cache anyway. Defaults to 0, which
1125 means fio does not periodically issue and wait for a sync to complete. Also
1126 see :option:`end_fsync` and :option:`fsync_on_close`.
1128 .. option:: fdatasync=int
1130 Like :option:`fsync` but uses :manpage:`fdatasync(2)` to only sync data and
1131 not metadata blocks. In Windows, FreeBSD, and DragonFlyBSD there is no
1132 :manpage:`fdatasync(2)` so this falls back to using :manpage:`fsync(2)`.
1133 Defaults to 0, which means fio does not periodically issue and wait for a
1134 data-only sync to complete.
1136 .. option:: write_barrier=int
1138 Make every `N-th` write a barrier write.
1140 .. option:: sync_file_range=str:val
1142 Use :manpage:`sync_file_range(2)` for every `val` number of write
1143 operations. Fio will track range of writes that have happened since the last
1144 :manpage:`sync_file_range(2)` call. `str` can currently be one or more of:
1147 SYNC_FILE_RANGE_WAIT_BEFORE
1149 SYNC_FILE_RANGE_WRITE
1151 SYNC_FILE_RANGE_WAIT_AFTER
1153 So if you do ``sync_file_range=wait_before,write:8``, fio would use
1154 ``SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE`` for every 8
1155 writes. Also see the :manpage:`sync_file_range(2)` man page. This option is
1158 .. option:: overwrite=bool
1160 If true, writes to a file will always overwrite existing data. If the file
1161 doesn't already exist, it will be created before the write phase begins. If
1162 the file exists and is large enough for the specified write phase, nothing
1163 will be done. Default: false.
1165 .. option:: end_fsync=bool
1167 If true, :manpage:`fsync(2)` file contents when a write stage has completed.
1170 .. option:: fsync_on_close=bool
1172 If true, fio will :manpage:`fsync(2)` a dirty file on close. This differs
1173 from :option:`end_fsync` in that it will happen on every file close, not
1174 just at the end of the job. Default: false.
1176 .. option:: rwmixread=int
1178 Percentage of a mixed workload that should be reads. Default: 50.
1180 .. option:: rwmixwrite=int
1182 Percentage of a mixed workload that should be writes. If both
1183 :option:`rwmixread` and :option:`rwmixwrite` is given and the values do not
1184 add up to 100%, the latter of the two will be used to override the
1185 first. This may interfere with a given rate setting, if fio is asked to
1186 limit reads or writes to a certain rate. If that is the case, then the
1187 distribution may be skewed. Default: 50.
1189 .. option:: random_distribution=str:float[,str:float][,str:float]
1191 By default, fio will use a completely uniform random distribution when asked
1192 to perform random I/O. Sometimes it is useful to skew the distribution in
1193 specific ways, ensuring that some parts of the data is more hot than others.
1194 fio includes the following distribution models:
1197 Uniform random distribution
1206 Normal (Gaussian) distribution
1209 Zoned random distribution
1211 When using a **zipf** or **pareto** distribution, an input value is also
1212 needed to define the access pattern. For **zipf**, this is the `zipf
1213 theta`. For **pareto**, it's the `Pareto power`. Fio includes a test
1214 program, :command:`genzipf`, that can be used visualize what the given input
1215 values will yield in terms of hit rates. If you wanted to use **zipf** with
1216 a `theta` of 1.2, you would use ``random_distribution=zipf:1.2`` as the
1217 option. If a non-uniform model is used, fio will disable use of the random
1218 map. For the **gauss** distribution, a normal deviation is supplied as a
1219 value between 0 and 100.
1221 For a **zoned** distribution, fio supports specifying percentages of I/O
1222 access that should fall within what range of the file or device. For
1223 example, given a criteria of:
1225 * 60% of accesses should be to the first 10%
1226 * 30% of accesses should be to the next 20%
1227 * 8% of accesses should be to to the next 30%
1228 * 2% of accesses should be to the next 40%
1230 we can define that through zoning of the random accesses. For the above
1231 example, the user would do::
1233 random_distribution=zoned:60/10:30/20:8/30:2/40
1235 similarly to how :option:`bssplit` works for setting ranges and percentages
1236 of block sizes. Like :option:`bssplit`, it's possible to specify separate
1237 zones for reads, writes, and trims. If just one set is given, it'll apply to
1240 .. option:: percentage_random=int[,int][,int]
1242 For a random workload, set how big a percentage should be random. This
1243 defaults to 100%, in which case the workload is fully random. It can be set
1244 from anywhere from 0 to 100. Setting it to 0 would make the workload fully
1245 sequential. Any setting in between will result in a random mix of sequential
1246 and random I/O, at the given percentages. Comma-separated values may be
1247 specified for reads, writes, and trims as described in :option:`blocksize`.
1249 .. option:: norandommap
1251 Normally fio will cover every block of the file when doing random I/O. If
1252 this option is given, fio will just get a new random offset without looking
1253 at past I/O history. This means that some blocks may not be read or written,
1254 and that some blocks may be read/written more than once. If this option is
1255 used with :option:`verify` and multiple blocksizes (via :option:`bsrange`),
1256 only intact blocks are verified, i.e., partially-overwritten blocks are
1259 .. option:: softrandommap=bool
1261 See :option:`norandommap`. If fio runs with the random block map enabled and
1262 it fails to allocate the map, if this option is set it will continue without
1263 a random block map. As coverage will not be as complete as with random maps,
1264 this option is disabled by default.
1266 .. option:: random_generator=str
1268 Fio supports the following engines for generating
1269 I/O offsets for random I/O:
1272 Strong 2^88 cycle random number generator
1274 Linear feedback shift register generator
1276 Strong 64-bit 2^258 cycle random number generator
1278 **tausworthe** is a strong random number generator, but it requires tracking
1279 on the side if we want to ensure that blocks are only read or written
1280 once. **LFSR** guarantees that we never generate the same offset twice, and
1281 it's also less computationally expensive. It's not a true random generator,
1282 however, though for I/O purposes it's typically good enough. **LFSR** only
1283 works with single block sizes, not with workloads that use multiple block
1284 sizes. If used with such a workload, fio may read or write some blocks
1285 multiple times. The default value is **tausworthe**, unless the required
1286 space exceeds 2^32 blocks. If it does, then **tausworthe64** is
1287 selected automatically.
1293 .. option:: blocksize=int[,int][,int], bs=int[,int][,int]
1295 The block size in bytes used for I/O units. Default: 4096. A single value
1296 applies to reads, writes, and trims. Comma-separated values may be
1297 specified for reads, writes, and trims. A value not terminated in a comma
1298 applies to subsequent types.
1303 means 256k for reads, writes and trims.
1306 means 8k for reads, 32k for writes and trims.
1309 means 8k for reads, 32k for writes, and default for trims.
1312 means default for reads, 8k for writes and trims.
1315 means default for reads, 8k for writes, and default for trims.
1317 .. option:: blocksize_range=irange[,irange][,irange], bsrange=irange[,irange][,irange]
1319 A range of block sizes in bytes for I/O units. The issued I/O unit will
1320 always be a multiple of the minimum size, unless
1321 :option:`blocksize_unaligned` is set.
1323 Comma-separated ranges may be specified for reads, writes, and trims as
1324 described in :option:`blocksize`.
1326 Example: ``bsrange=1k-4k,2k-8k``.
1328 .. option:: bssplit=str[,str][,str]
1330 Sometimes you want even finer grained control of the block sizes issued, not
1331 just an even split between them. This option allows you to weight various
1332 block sizes, so that you are able to define a specific amount of block sizes
1333 issued. The format for this option is::
1335 bssplit=blocksize/percentage:blocksize/percentage
1337 for as many block sizes as needed. So if you want to define a workload that
1338 has 50% 64k blocks, 10% 4k blocks, and 40% 32k blocks, you would write::
1340 bssplit=4k/10:64k/50:32k/40
1342 Ordering does not matter. If the percentage is left blank, fio will fill in
1343 the remaining values evenly. So a bssplit option like this one::
1345 bssplit=4k/50:1k/:32k/
1347 would have 50% 4k ios, and 25% 1k and 32k ios. The percentages always add up
1348 to 100, if bssplit is given a range that adds up to more, it will error out.
1350 Comma-separated values may be specified for reads, writes, and trims as
1351 described in :option:`blocksize`.
1353 If you want a workload that has 50% 2k reads and 50% 4k reads, while having
1354 90% 4k writes and 10% 8k writes, you would specify::
1356 bssplit=2k/50:4k/50,4k/90,8k/10
1358 .. option:: blocksize_unaligned, bs_unaligned
1360 If set, fio will issue I/O units with any size within
1361 :option:`blocksize_range`, not just multiples of the minimum size. This
1362 typically won't work with direct I/O, as that normally requires sector
1365 .. option:: bs_is_seq_rand
1367 If this option is set, fio will use the normal read,write blocksize settings
1368 as sequential,random blocksize settings instead. Any random read or write
1369 will use the WRITE blocksize settings, and any sequential read or write will
1370 use the READ blocksize settings.
1372 .. option:: blockalign=int[,int][,int], ba=int[,int][,int]
1374 Boundary to which fio will align random I/O units. Default:
1375 :option:`blocksize`. Minimum alignment is typically 512b for using direct
1376 I/O, though it usually depends on the hardware block size. This option is
1377 mutually exclusive with using a random map for files, so it will turn off
1378 that option. Comma-separated values may be specified for reads, writes, and
1379 trims as described in :option:`blocksize`.
1385 .. option:: zero_buffers
1387 Initialize buffers with all zeros. Default: fill buffers with random data.
1389 .. option:: refill_buffers
1391 If this option is given, fio will refill the I/O buffers on every
1392 submit. The default is to only fill it at init time and reuse that
1393 data. Only makes sense if zero_buffers isn't specified, naturally. If data
1394 verification is enabled, `refill_buffers` is also automatically enabled.
1396 .. option:: scramble_buffers=bool
1398 If :option:`refill_buffers` is too costly and the target is using data
1399 deduplication, then setting this option will slightly modify the I/O buffer
1400 contents to defeat normal de-dupe attempts. This is not enough to defeat
1401 more clever block compression attempts, but it will stop naive dedupe of
1402 blocks. Default: true.
1404 .. option:: buffer_compress_percentage=int
1406 If this is set, then fio will attempt to provide I/O buffer content (on
1407 WRITEs) that compresses to the specified level. Fio does this by providing a
1408 mix of random data and a fixed pattern. The fixed pattern is either zeros,
1409 or the pattern specified by :option:`buffer_pattern`. If the pattern option
1410 is used, it might skew the compression ratio slightly. Note that this is per
1411 block size unit, for file/disk wide compression level that matches this
1412 setting, you'll also want to set :option:`refill_buffers`.
1414 .. option:: buffer_compress_chunk=int
1416 See :option:`buffer_compress_percentage`. This setting allows fio to manage
1417 how big the ranges of random data and zeroed data is. Without this set, fio
1418 will provide :option:`buffer_compress_percentage` of blocksize random data,
1419 followed by the remaining zeroed. With this set to some chunk size smaller
1420 than the block size, fio can alternate random and zeroed data throughout the
1423 .. option:: buffer_pattern=str
1425 If set, fio will fill the I/O buffers with this pattern or with the contents
1426 of a file. If not set, the contents of I/O buffers are defined by the other
1427 options related to buffer contents. The setting can be any pattern of bytes,
1428 and can be prefixed with 0x for hex values. It may also be a string, where
1429 the string must then be wrapped with ``""``. Or it may also be a filename,
1430 where the filename must be wrapped with ``''`` in which case the file is
1431 opened and read. Note that not all the file contents will be read if that
1432 would cause the buffers to overflow. So, for example::
1434 buffer_pattern='filename'
1438 buffer_pattern="abcd"
1446 buffer_pattern=0xdeadface
1448 Also you can combine everything together in any order::
1450 buffer_pattern=0xdeadface"abcd"-12'filename'
1452 .. option:: dedupe_percentage=int
1454 If set, fio will generate this percentage of identical buffers when
1455 writing. These buffers will be naturally dedupable. The contents of the
1456 buffers depend on what other buffer compression settings have been set. It's
1457 possible to have the individual buffers either fully compressible, or not at
1458 all. This option only controls the distribution of unique buffers.
1460 .. option:: invalidate=bool
1462 Invalidate the buffer/page cache parts of the files to be used prior to
1463 starting I/O if the platform and file type support it. Defaults to true.
1464 This will be ignored if :option:`pre_read` is also specified for the
1467 .. option:: sync=bool
1469 Use synchronous I/O for buffered writes. For the majority of I/O engines,
1470 this means using O_SYNC. Default: false.
1472 .. option:: iomem=str, mem=str
1474 Fio can use various types of memory as the I/O unit buffer. The allowed
1478 Use memory from :manpage:`malloc(3)` as the buffers. Default memory
1482 Use shared memory as the buffers. Allocated through
1483 :manpage:`shmget(2)`.
1486 Same as shm, but use huge pages as backing.
1489 Use :manpage:`mmap(2)` to allocate buffers. May either be anonymous memory, or can
1490 be file backed if a filename is given after the option. The format
1491 is `mem=mmap:/path/to/file`.
1494 Use a memory mapped huge file as the buffer backing. Append filename
1495 after mmaphuge, ala `mem=mmaphuge:/hugetlbfs/file`.
1498 Same as mmap, but use a MMAP_SHARED mapping.
1501 Use GPU memory as the buffers for GPUDirect RDMA benchmark.
1503 The area allocated is a function of the maximum allowed bs size for the job,
1504 multiplied by the I/O depth given. Note that for **shmhuge** and
1505 **mmaphuge** to work, the system must have free huge pages allocated. This
1506 can normally be checked and set by reading/writing
1507 :file:`/proc/sys/vm/nr_hugepages` on a Linux system. Fio assumes a huge page
1508 is 4MiB in size. So to calculate the number of huge pages you need for a
1509 given job file, add up the I/O depth of all jobs (normally one unless
1510 :option:`iodepth` is used) and multiply by the maximum bs set. Then divide
1511 that number by the huge page size. You can see the size of the huge pages in
1512 :file:`/proc/meminfo`. If no huge pages are allocated by having a non-zero
1513 number in `nr_hugepages`, using **mmaphuge** or **shmhuge** will fail. Also
1514 see :option:`hugepage-size`.
1516 **mmaphuge** also needs to have hugetlbfs mounted and the file location
1517 should point there. So if it's mounted in :file:`/huge`, you would use
1518 `mem=mmaphuge:/huge/somefile`.
1520 .. option:: iomem_align=int
1522 This indicates the memory alignment of the I/O memory buffers. Note that
1523 the given alignment is applied to the first I/O unit buffer, if using
1524 :option:`iodepth` the alignment of the following buffers are given by the
1525 :option:`bs` used. In other words, if using a :option:`bs` that is a
1526 multiple of the page sized in the system, all buffers will be aligned to
1527 this value. If using a :option:`bs` that is not page aligned, the alignment
1528 of subsequent I/O memory buffers is the sum of the :option:`iomem_align` and
1531 .. option:: hugepage-size=int
1533 Defines the size of a huge page. Must at least be equal to the system
1534 setting, see :file:`/proc/meminfo`. Defaults to 4MiB. Should probably
1535 always be a multiple of megabytes, so using ``hugepage-size=Xm`` is the
1536 preferred way to set this to avoid setting a non-pow-2 bad value.
1538 .. option:: lockmem=int
1540 Pin the specified amount of memory with :manpage:`mlock(2)`. Can be used to
1541 simulate a smaller amount of memory. The amount specified is per worker.
1547 .. option:: size=int
1549 The total size of file I/O for each thread of this job. Fio will run until
1550 this many bytes has been transferred, unless runtime is limited by other options
1551 (such as :option:`runtime`, for instance, or increased/decreased by :option:`io_size`).
1552 Fio will divide this size between the available files determined by options
1553 such as :option:`nrfiles`, :option:`filename`, unless :option:`filesize` is
1554 specified by the job. If the result of division happens to be 0, the size is
1555 set to the physical size of the given files or devices if they exist.
1556 If this option is not specified, fio will use the full size of the given
1557 files or devices. If the files do not exist, size must be given. It is also
1558 possible to give size as a percentage between 1 and 100. If ``size=20%`` is
1559 given, fio will use 20% of the full size of the given files or devices.
1560 Can be combined with :option:`offset` to constrain the start and end range
1561 that I/O will be done within.
1563 .. option:: io_size=int, io_limit=int
1565 Normally fio operates within the region set by :option:`size`, which means
1566 that the :option:`size` option sets both the region and size of I/O to be
1567 performed. Sometimes that is not what you want. With this option, it is
1568 possible to define just the amount of I/O that fio should do. For instance,
1569 if :option:`size` is set to 20GiB and :option:`io_size` is set to 5GiB, fio
1570 will perform I/O within the first 20GiB but exit when 5GiB have been
1571 done. The opposite is also possible -- if :option:`size` is set to 20GiB,
1572 and :option:`io_size` is set to 40GiB, then fio will do 40GiB of I/O within
1573 the 0..20GiB region.
1575 .. option:: filesize=irange(int)
1577 Individual file sizes. May be a range, in which case fio will select sizes
1578 for files at random within the given range and limited to :option:`size` in
1579 total (if that is given). If not given, each created file is the same size.
1580 This option overrides :option:`size` in terms of file size, which means
1581 this value is used as a fixed size or possible range of each file.
1583 .. option:: file_append=bool
1585 Perform I/O after the end of the file. Normally fio will operate within the
1586 size of a file. If this option is set, then fio will append to the file
1587 instead. This has identical behavior to setting :option:`offset` to the size
1588 of a file. This option is ignored on non-regular files.
1590 .. option:: fill_device=bool, fill_fs=bool
1592 Sets size to something really large and waits for ENOSPC (no space left on
1593 device) as the terminating condition. Only makes sense with sequential
1594 write. For a read workload, the mount point will be filled first then I/O
1595 started on the result. This option doesn't make sense if operating on a raw
1596 device node, since the size of that is already known by the file system.
1597 Additionally, writing beyond end-of-device will not return ENOSPC there.
1603 .. option:: ioengine=str
1605 Defines how the job issues I/O to the file. The following types are defined:
1608 Basic :manpage:`read(2)` or :manpage:`write(2)`
1609 I/O. :manpage:`lseek(2)` is used to position the I/O location.
1610 See :option:`fsync` and :option:`fdatasync` for syncing write I/Os.
1613 Basic :manpage:`pread(2)` or :manpage:`pwrite(2)` I/O. Default on
1614 all supported operating systems except for Windows.
1617 Basic :manpage:`readv(2)` or :manpage:`writev(2)` I/O. Will emulate
1618 queuing by coalescing adjacent I/Os into a single submission.
1621 Basic :manpage:`preadv(2)` or :manpage:`pwritev(2)` I/O.
1624 Basic :manpage:`preadv2(2)` or :manpage:`pwritev2(2)` I/O.
1627 Linux native asynchronous I/O. Note that Linux may only support
1628 queued behavior with non-buffered I/O (set ``direct=1`` or
1630 This engine defines engine specific options.
1633 POSIX asynchronous I/O using :manpage:`aio_read(3)` and
1634 :manpage:`aio_write(3)`.
1637 Solaris native asynchronous I/O.
1640 Windows native asynchronous I/O. Default on Windows.
1643 File is memory mapped with :manpage:`mmap(2)` and data copied
1644 to/from using :manpage:`memcpy(3)`.
1647 :manpage:`splice(2)` is used to transfer the data and
1648 :manpage:`vmsplice(2)` to transfer data from user space to the
1652 SCSI generic sg v3 I/O. May either be synchronous using the SG_IO
1653 ioctl, or if the target is an sg character device we use
1654 :manpage:`read(2)` and :manpage:`write(2)` for asynchronous
1655 I/O. Requires filename option to specify either block or character
1659 Doesn't transfer any data, just pretends to. This is mainly used to
1660 exercise fio itself and for debugging/testing purposes.
1663 Transfer over the network to given ``host:port``. Depending on the
1664 :option:`protocol` used, the :option:`hostname`, :option:`port`,
1665 :option:`listen` and :option:`filename` options are used to specify
1666 what sort of connection to make, while the :option:`protocol` option
1667 determines which protocol will be used. This engine defines engine
1671 Like **net**, but uses :manpage:`splice(2)` and
1672 :manpage:`vmsplice(2)` to map data and send/receive.
1673 This engine defines engine specific options.
1676 Doesn't transfer any data, but burns CPU cycles according to the
1677 :option:`cpuload` and :option:`cpuchunks` options. Setting
1678 :option:`cpuload`\=85 will cause that job to do nothing but burn 85%
1679 of the CPU. In case of SMP machines, use :option:`numjobs`
1680 =<no_of_cpu> to get desired CPU usage, as the cpuload only loads a
1681 single CPU at the desired rate. A job never finishes unless there is
1682 at least one non-cpuio job.
1685 The GUASI I/O engine is the Generic Userspace Asyncronous Syscall
1686 Interface approach to async I/O. See
1688 http://www.xmailserver.org/guasi-lib.html
1690 for more info on GUASI.
1693 The RDMA I/O engine supports both RDMA memory semantics
1694 (RDMA_WRITE/RDMA_READ) and channel semantics (Send/Recv) for the
1695 InfiniBand, RoCE and iWARP protocols.
1698 I/O engine that does regular fallocate to simulate data transfer as
1702 does fallocate(,mode = FALLOC_FL_KEEP_SIZE,).
1705 does fallocate(,mode = 0).
1708 does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE).
1711 I/O engine that sends :manpage:`ftruncate(2)` operations in response
1712 to write (DDIR_WRITE) events. Each ftruncate issued sets the file's
1713 size to the current block offset. Block size is ignored.
1716 I/O engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate
1717 defragment activity in request to DDIR_WRITE event.
1720 I/O engine supporting direct access to Ceph Rados Block Devices
1721 (RBD) via librbd without the need to use the kernel rbd driver. This
1722 ioengine defines engine specific options.
1725 Using GlusterFS libgfapi sync interface to direct access to
1726 GlusterFS volumes without having to go through FUSE. This ioengine
1727 defines engine specific options.
1730 Using GlusterFS libgfapi async interface to direct access to
1731 GlusterFS volumes without having to go through FUSE. This ioengine
1732 defines engine specific options.
1735 Read and write through Hadoop (HDFS). The :file:`filename` option
1736 is used to specify host,port of the hdfs name-node to connect. This
1737 engine interprets offsets a little differently. In HDFS, files once
1738 created cannot be modified. So random writes are not possible. To
1739 imitate this, libhdfs engine expects bunch of small files to be
1740 created over HDFS, and engine will randomly pick a file out of those
1741 files based on the offset generated by fio backend. (see the example
1742 job file to create such files, use ``rw=write`` option). Please
1743 note, you might want to set necessary environment variables to work
1744 with hdfs/libhdfs properly. Each job uses its own connection to
1748 Read, write and erase an MTD character device (e.g.,
1749 :file:`/dev/mtd0`). Discards are treated as erases. Depending on the
1750 underlying device type, the I/O may have to go in a certain pattern,
1751 e.g., on NAND, writing sequentially to erase blocks and discarding
1752 before overwriting. The writetrim mode works well for this
1756 Read and write using filesystem DAX to a file on a filesystem
1757 mounted with DAX on a persistent memory device through the NVML
1761 Read and write using device DAX to a persistent memory device (e.g.,
1762 /dev/dax0.0) through the NVML libpmem library.
1765 Prefix to specify loading an external I/O engine object file. Append
1766 the engine filename, e.g. ``ioengine=external:/tmp/foo.o`` to load
1767 ioengine :file:`foo.o` in :file:`/tmp`.
1770 I/O engine specific parameters
1771 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1773 In addition, there are some parameters which are only valid when a specific
1774 ioengine is in use. These are used identically to normal parameters, with the
1775 caveat that when used on the command line, they must come after the
1776 :option:`ioengine` that defines them is selected.
1778 .. option:: userspace_reap : [libaio]
1780 Normally, with the libaio engine in use, fio will use the
1781 :manpage:`io_getevents(2)` system call to reap newly returned events. With
1782 this flag turned on, the AIO ring will be read directly from user-space to
1783 reap events. The reaping mode is only enabled when polling for a minimum of
1784 0 events (e.g. when :option:`iodepth_batch_complete` `=0`).
1786 .. option:: hipri : [pvsync2]
1788 Set RWF_HIPRI on I/O, indicating to the kernel that it's of higher priority
1791 .. option:: cpuload=int : [cpuio]
1793 Attempt to use the specified percentage of CPU cycles. This is a mandatory
1794 option when using cpuio I/O engine.
1796 .. option:: cpuchunks=int : [cpuio]
1798 Split the load into cycles of the given time. In microseconds.
1800 .. option:: exit_on_io_done=bool : [cpuio]
1802 Detect when I/O threads are done, then exit.
1804 .. option:: hostname=str : [netsplice] [net]
1806 The hostname or IP address to use for TCP or UDP based I/O. If the job is
1807 a TCP listener or UDP reader, the hostname is not used and must be omitted
1808 unless it is a valid UDP multicast address.
1810 .. option:: namenode=str : [libhdfs]
1812 The hostname or IP address of a HDFS cluster namenode to contact.
1814 .. option:: port=int
1818 The TCP or UDP port to bind to or connect to. If this is used with
1819 :option:`numjobs` to spawn multiple instances of the same job type, then
1820 this will be the starting port number since fio will use a range of
1825 the listening port of the HFDS cluster namenode.
1827 .. option:: interface=str : [netsplice] [net]
1829 The IP address of the network interface used to send or receive UDP
1832 .. option:: ttl=int : [netsplice] [net]
1834 Time-to-live value for outgoing UDP multicast packets. Default: 1.
1836 .. option:: nodelay=bool : [netsplice] [net]
1838 Set TCP_NODELAY on TCP connections.
1840 .. option:: protocol=str : [netsplice] [net]
1842 .. option:: proto=str : [netsplice] [net]
1844 The network protocol to use. Accepted values are:
1847 Transmission control protocol.
1849 Transmission control protocol V6.
1851 User datagram protocol.
1853 User datagram protocol V6.
1857 When the protocol is TCP or UDP, the port must also be given, as well as the
1858 hostname if the job is a TCP listener or UDP reader. For unix sockets, the
1859 normal filename option should be used and the port is invalid.
1861 .. option:: listen : [net]
1863 For TCP network connections, tell fio to listen for incoming connections
1864 rather than initiating an outgoing connection. The :option:`hostname` must
1865 be omitted if this option is used.
1867 .. option:: pingpong : [net]
1869 Normally a network writer will just continue writing data, and a network
1870 reader will just consume packages. If ``pingpong=1`` is set, a writer will
1871 send its normal payload to the reader, then wait for the reader to send the
1872 same payload back. This allows fio to measure network latencies. The
1873 submission and completion latencies then measure local time spent sending or
1874 receiving, and the completion latency measures how long it took for the
1875 other end to receive and send back. For UDP multicast traffic
1876 ``pingpong=1`` should only be set for a single reader when multiple readers
1877 are listening to the same address.
1879 .. option:: window_size : [net]
1881 Set the desired socket buffer size for the connection.
1883 .. option:: mss : [net]
1885 Set the TCP maximum segment size (TCP_MAXSEG).
1887 .. option:: donorname=str : [e4defrag]
1889 File will be used as a block donor (swap extents between files).
1891 .. option:: inplace=int : [e4defrag]
1893 Configure donor file blocks allocation strategy:
1896 Default. Preallocate donor's file on init.
1898 Allocate space immediately inside defragment event, and free right
1901 .. option:: clustername=str : [rbd]
1903 Specifies the name of the Ceph cluster.
1905 .. option:: rbdname=str : [rbd]
1907 Specifies the name of the RBD.
1909 .. option:: pool=str : [rbd]
1911 Specifies the name of the Ceph pool containing RBD.
1913 .. option:: clientname=str : [rbd]
1915 Specifies the username (without the 'client.' prefix) used to access the
1916 Ceph cluster. If the *clustername* is specified, the *clientname* shall be
1917 the full *type.id* string. If no type. prefix is given, fio will add
1918 'client.' by default.
1920 .. option:: skip_bad=bool : [mtd]
1922 Skip operations against known bad blocks.
1924 .. option:: hdfsdirectory : [libhdfs]
1926 libhdfs will create chunk in this HDFS directory.
1928 .. option:: chunk_size : [libhdfs]
1930 the size of the chunk to use for each file.
1936 .. option:: iodepth=int
1938 Number of I/O units to keep in flight against the file. Note that
1939 increasing *iodepth* beyond 1 will not affect synchronous ioengines (except
1940 for small degrees when :option:`verify_async` is in use). Even async
1941 engines may impose OS restrictions causing the desired depth not to be
1942 achieved. This may happen on Linux when using libaio and not setting
1943 :option:`direct`\=1, since buffered I/O is not async on that OS. Keep an
1944 eye on the I/O depth distribution in the fio output to verify that the
1945 achieved depth is as expected. Default: 1.
1947 .. option:: iodepth_batch_submit=int, iodepth_batch=int
1949 This defines how many pieces of I/O to submit at once. It defaults to 1
1950 which means that we submit each I/O as soon as it is available, but can be
1951 raised to submit bigger batches of I/O at the time. If it is set to 0 the
1952 :option:`iodepth` value will be used.
1954 .. option:: iodepth_batch_complete_min=int, iodepth_batch_complete=int
1956 This defines how many pieces of I/O to retrieve at once. It defaults to 1
1957 which means that we'll ask for a minimum of 1 I/O in the retrieval process
1958 from the kernel. The I/O retrieval will go on until we hit the limit set by
1959 :option:`iodepth_low`. If this variable is set to 0, then fio will always
1960 check for completed events before queuing more I/O. This helps reduce I/O
1961 latency, at the cost of more retrieval system calls.
1963 .. option:: iodepth_batch_complete_max=int
1965 This defines maximum pieces of I/O to retrieve at once. This variable should
1966 be used along with :option:`iodepth_batch_complete_min`\=int variable,
1967 specifying the range of min and max amount of I/O which should be
1968 retrieved. By default it is equal to the :option:`iodepth_batch_complete_min`
1973 iodepth_batch_complete_min=1
1974 iodepth_batch_complete_max=<iodepth>
1976 which means that we will retrieve at least 1 I/O and up to the whole
1977 submitted queue depth. If none of I/O has been completed yet, we will wait.
1981 iodepth_batch_complete_min=0
1982 iodepth_batch_complete_max=<iodepth>
1984 which means that we can retrieve up to the whole submitted queue depth, but
1985 if none of I/O has been completed yet, we will NOT wait and immediately exit
1986 the system call. In this example we simply do polling.
1988 .. option:: iodepth_low=int
1990 The low water mark indicating when to start filling the queue
1991 again. Defaults to the same as :option:`iodepth`, meaning that fio will
1992 attempt to keep the queue full at all times. If :option:`iodepth` is set to
1993 e.g. 16 and *iodepth_low* is set to 4, then after fio has filled the queue of
1994 16 requests, it will let the depth drain down to 4 before starting to fill
1997 .. option:: io_submit_mode=str
1999 This option controls how fio submits the I/O to the I/O engine. The default
2000 is `inline`, which means that the fio job threads submit and reap I/O
2001 directly. If set to `offload`, the job threads will offload I/O submission
2002 to a dedicated pool of I/O threads. This requires some coordination and thus
2003 has a bit of extra overhead, especially for lower queue depth I/O where it
2004 can increase latencies. The benefit is that fio can manage submission rates
2005 independently of the device completion rates. This avoids skewed latency
2006 reporting if I/O gets backed up on the device side (the coordinated omission
2013 .. option:: thinktime=time
2015 Stall the job for the specified period of time after an I/O has completed before issuing the
2016 next. May be used to simulate processing being done by an application.
2017 When the unit is omitted, the value is interpreted in microseconds. See
2018 :option:`thinktime_blocks` and :option:`thinktime_spin`.
2020 .. option:: thinktime_spin=time
2022 Only valid if :option:`thinktime` is set - pretend to spend CPU time doing
2023 something with the data received, before falling back to sleeping for the
2024 rest of the period specified by :option:`thinktime`. When the unit is
2025 omitted, the value is interpreted in microseconds.
2027 .. option:: thinktime_blocks=int
2029 Only valid if :option:`thinktime` is set - control how many blocks to issue,
2030 before waiting `thinktime` usecs. If not set, defaults to 1 which will make
2031 fio wait `thinktime` usecs after every block. This effectively makes any
2032 queue depth setting redundant, since no more than 1 I/O will be queued
2033 before we have to complete it and do our thinktime. In other words, this
2034 setting effectively caps the queue depth if the latter is larger.
2036 .. option:: rate=int[,int][,int]
2038 Cap the bandwidth used by this job. The number is in bytes/sec, the normal
2039 suffix rules apply. Comma-separated values may be specified for reads,
2040 writes, and trims as described in :option:`blocksize`.
2042 .. option:: rate_min=int[,int][,int]
2044 Tell fio to do whatever it can to maintain at least this bandwidth. Failing
2045 to meet this requirement will cause the job to exit. Comma-separated values
2046 may be specified for reads, writes, and trims as described in
2047 :option:`blocksize`.
2049 .. option:: rate_iops=int[,int][,int]
2051 Cap the bandwidth to this number of IOPS. Basically the same as
2052 :option:`rate`, just specified independently of bandwidth. If the job is
2053 given a block size range instead of a fixed value, the smallest block size
2054 is used as the metric. Comma-separated values may be specified for reads,
2055 writes, and trims as described in :option:`blocksize`.
2057 .. option:: rate_iops_min=int[,int][,int]
2059 If fio doesn't meet this rate of I/O, it will cause the job to exit.
2060 Comma-separated values may be specified for reads, writes, and trims as
2061 described in :option:`blocksize`.
2063 .. option:: rate_process=str
2065 This option controls how fio manages rated I/O submissions. The default is
2066 `linear`, which submits I/O in a linear fashion with fixed delays between
2067 I/Os that gets adjusted based on I/O completion rates. If this is set to
2068 `poisson`, fio will submit I/O based on a more real world random request
2069 flow, known as the Poisson process
2070 (https://en.wikipedia.org/wiki/Poisson_point_process). The lambda will be
2071 10^6 / IOPS for the given workload.
2077 .. option:: latency_target=time
2079 If set, fio will attempt to find the max performance point that the given
2080 workload will run at while maintaining a latency below this target. When
2081 the unit is omitted, the value is interpreted in microseconds. See
2082 :option:`latency_window` and :option:`latency_percentile`.
2084 .. option:: latency_window=time
2086 Used with :option:`latency_target` to specify the sample window that the job
2087 is run at varying queue depths to test the performance. When the unit is
2088 omitted, the value is interpreted in microseconds.
2090 .. option:: latency_percentile=float
2092 The percentage of I/Os that must fall within the criteria specified by
2093 :option:`latency_target` and :option:`latency_window`. If not set, this
2094 defaults to 100.0, meaning that all I/Os must be equal or below to the value
2095 set by :option:`latency_target`.
2097 .. option:: max_latency=time
2099 If set, fio will exit the job with an ETIMEDOUT error if it exceeds this
2100 maximum latency. When the unit is omitted, the value is interpreted in
2103 .. option:: rate_cycle=int
2105 Average bandwidth for :option:`rate` and :option:`rate_min` over this number
2106 of milliseconds. Defaults to 1000.
2112 .. option:: write_iolog=str
2114 Write the issued I/O patterns to the specified file. See
2115 :option:`read_iolog`. Specify a separate file for each job, otherwise the
2116 iologs will be interspersed and the file may be corrupt.
2118 .. option:: read_iolog=str
2120 Open an iolog with the specified filename and replay the I/O patterns it
2121 contains. This can be used to store a workload and replay it sometime
2122 later. The iolog given may also be a blktrace binary file, which allows fio
2123 to replay a workload captured by :command:`blktrace`. See
2124 :manpage:`blktrace(8)` for how to capture such logging data. For blktrace
2125 replay, the file needs to be turned into a blkparse binary data file first
2126 (``blkparse <device> -o /dev/null -d file_for_fio.bin``).
2128 .. option:: replay_no_stall=int
2130 When replaying I/O with :option:`read_iolog` the default behavior is to
2131 attempt to respect the timestamps within the log and replay them with the
2132 appropriate delay between IOPS. By setting this variable fio will not
2133 respect the timestamps and attempt to replay them as fast as possible while
2134 still respecting ordering. The result is the same I/O pattern to a given
2135 device, but different timings.
2137 .. option:: replay_redirect=str
2139 While replaying I/O patterns using :option:`read_iolog` the default behavior
2140 is to replay the IOPS onto the major/minor device that each IOP was recorded
2141 from. This is sometimes undesirable because on a different machine those
2142 major/minor numbers can map to a different device. Changing hardware on the
2143 same system can also result in a different major/minor mapping.
2144 ``replay_redirect`` causes all I/Os to be replayed onto the single specified
2145 device regardless of the device it was recorded
2146 from. i.e. :option:`replay_redirect`\= :file:`/dev/sdc` would cause all I/O
2147 in the blktrace or iolog to be replayed onto :file:`/dev/sdc`. This means
2148 multiple devices will be replayed onto a single device, if the trace
2149 contains multiple devices. If you want multiple devices to be replayed
2150 concurrently to multiple redirected devices you must blkparse your trace
2151 into separate traces and replay them with independent fio invocations.
2152 Unfortunately this also breaks the strict time ordering between multiple
2155 .. option:: replay_align=int
2157 Force alignment of I/O offsets and lengths in a trace to this power of 2
2160 .. option:: replay_scale=int
2162 Scale sector offsets down by this factor when replaying traces.
2165 Threads, processes and job synchronization
2166 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2170 Fio defaults to creating jobs by using fork, however if this option is
2171 given, fio will create jobs by using POSIX Threads' function
2172 :manpage:`pthread_create(3)` to create threads instead.
2174 .. option:: wait_for=str
2176 If set, the current job won't be started until all workers of the specified
2177 waitee job are done.
2179 ``wait_for`` operates on the job name basis, so there are a few
2180 limitations. First, the waitee must be defined prior to the waiter job
2181 (meaning no forward references). Second, if a job is being referenced as a
2182 waitee, it must have a unique name (no duplicate waitees).
2184 .. option:: nice=int
2186 Run the job with the given nice value. See man :manpage:`nice(2)`.
2188 On Windows, values less than -15 set the process class to "High"; -1 through
2189 -15 set "Above Normal"; 1 through 15 "Below Normal"; and above 15 "Idle"
2192 .. option:: prio=int
2194 Set the I/O priority value of this job. Linux limits us to a positive value
2195 between 0 and 7, with 0 being the highest. See man
2196 :manpage:`ionice(1)`. Refer to an appropriate manpage for other operating
2197 systems since meaning of priority may differ.
2199 .. option:: prioclass=int
2201 Set the I/O priority class. See man :manpage:`ionice(1)`.
2203 .. option:: cpumask=int
2205 Set the CPU affinity of this job. The parameter given is a bit mask of
2206 allowed CPUs the job may run on. So if you want the allowed CPUs to be 1
2207 and 5, you would pass the decimal value of (1 << 1 | 1 << 5), or 34. See man
2208 :manpage:`sched_setaffinity(2)`. This may not work on all supported
2209 operating systems or kernel versions. This option doesn't work well for a
2210 higher CPU count than what you can store in an integer mask, so it can only
2211 control cpus 1-32. For boxes with larger CPU counts, use
2212 :option:`cpus_allowed`.
2214 .. option:: cpus_allowed=str
2216 Controls the same options as :option:`cpumask`, but accepts a textual
2217 specification of the permitted CPUs instead. So to use CPUs 1 and 5 you
2218 would specify ``cpus_allowed=1,5``. This option also allows a range of CPUs
2219 to be specified -- say you wanted a binding to CPUs 1, 5, and 8 to 15, you
2220 would set ``cpus_allowed=1,5,8-15``.
2222 .. option:: cpus_allowed_policy=str
2224 Set the policy of how fio distributes the CPUs specified by
2225 :option:`cpus_allowed` or :option:`cpumask`. Two policies are supported:
2228 All jobs will share the CPU set specified.
2230 Each job will get a unique CPU from the CPU set.
2232 **shared** is the default behavior, if the option isn't specified. If
2233 **split** is specified, then fio will will assign one cpu per job. If not
2234 enough CPUs are given for the jobs listed, then fio will roundrobin the CPUs
2237 .. option:: numa_cpu_nodes=str
2239 Set this job running on specified NUMA nodes' CPUs. The arguments allow
2240 comma delimited list of cpu numbers, A-B ranges, or `all`. Note, to enable
2241 NUMA options support, fio must be built on a system with libnuma-dev(el)
2244 .. option:: numa_mem_policy=str
2246 Set this job's memory policy and corresponding NUMA nodes. Format of the
2251 ``mode`` is one of the following memory poicies: ``default``, ``prefer``,
2252 ``bind``, ``interleave`` or ``local``. For ``default`` and ``local`` memory
2253 policies, no node needs to be specified. For ``prefer``, only one node is
2254 allowed. For ``bind`` and ``interleave`` the ``nodelist`` may be as
2255 follows: a comma delimited list of numbers, A-B ranges, or `all`.
2257 .. option:: cgroup=str
2259 Add job to this control group. If it doesn't exist, it will be created. The
2260 system must have a mounted cgroup blkio mount point for this to work. If
2261 your system doesn't have it mounted, you can do so with::
2263 # mount -t cgroup -o blkio none /cgroup
2265 .. option:: cgroup_weight=int
2267 Set the weight of the cgroup to this value. See the documentation that comes
2268 with the kernel, allowed values are in the range of 100..1000.
2270 .. option:: cgroup_nodelete=bool
2272 Normally fio will delete the cgroups it has created after the job
2273 completion. To override this behavior and to leave cgroups around after the
2274 job completion, set ``cgroup_nodelete=1``. This can be useful if one wants
2275 to inspect various cgroup files after job completion. Default: false.
2277 .. option:: flow_id=int
2279 The ID of the flow. If not specified, it defaults to being a global
2280 flow. See :option:`flow`.
2282 .. option:: flow=int
2284 Weight in token-based flow control. If this value is used, then there is a
2285 'flow counter' which is used to regulate the proportion of activity between
2286 two or more jobs. Fio attempts to keep this flow counter near zero. The
2287 ``flow`` parameter stands for how much should be added or subtracted to the
2288 flow counter on each iteration of the main I/O loop. That is, if one job has
2289 ``flow=8`` and another job has ``flow=-1``, then there will be a roughly 1:8
2290 ratio in how much one runs vs the other.
2292 .. option:: flow_watermark=int
2294 The maximum value that the absolute value of the flow counter is allowed to
2295 reach before the job must wait for a lower value of the counter.
2297 .. option:: flow_sleep=int
2299 The period of time, in microseconds, to wait after the flow watermark has
2300 been exceeded before retrying operations.
2302 .. option:: stonewall, wait_for_previous
2304 Wait for preceding jobs in the job file to exit, before starting this
2305 one. Can be used to insert serialization points in the job file. A stone
2306 wall also implies starting a new reporting group, see
2307 :option:`group_reporting`.
2311 By default, fio will continue running all other jobs when one job finishes
2312 but sometimes this is not the desired action. Setting ``exitall`` will
2313 instead make fio terminate all other jobs when one job finishes.
2315 .. option:: exec_prerun=str
2317 Before running this job, issue the command specified through
2318 :manpage:`system(3)`. Output is redirected in a file called
2319 :file:`jobname.prerun.txt`.
2321 .. option:: exec_postrun=str
2323 After the job completes, issue the command specified though
2324 :manpage:`system(3)`. Output is redirected in a file called
2325 :file:`jobname.postrun.txt`.
2329 Instead of running as the invoking user, set the user ID to this value
2330 before the thread/process does any work.
2334 Set group ID, see :option:`uid`.
2340 .. option:: verify_only
2342 Do not perform specified workload, only verify data still matches previous
2343 invocation of this workload. This option allows one to check data multiple
2344 times at a later date without overwriting it. This option makes sense only
2345 for workloads that write data, and does not support workloads with the
2346 :option:`time_based` option set.
2348 .. option:: do_verify=bool
2350 Run the verify phase after a write phase. Only valid if :option:`verify` is
2353 .. option:: verify=str
2355 If writing to a file, fio can verify the file contents after each iteration
2356 of the job. Each verification method also implies verification of special
2357 header, which is written to the beginning of each block. This header also
2358 includes meta information, like offset of the block, block number, timestamp
2359 when block was written, etc. :option:`verify` can be combined with
2360 :option:`verify_pattern` option. The allowed values are:
2363 Use an md5 sum of the data area and store it in the header of
2367 Use an experimental crc64 sum of the data area and store it in the
2368 header of each block.
2371 Use a crc32c sum of the data area and store it in the header of each
2375 Use hardware assisted crc32c calculation provided on SSE4.2 enabled
2376 processors. Falls back to regular software crc32c, if not supported
2380 Use a crc32 sum of the data area and store it in the header of each
2384 Use a crc16 sum of the data area and store it in the header of each
2388 Use a crc7 sum of the data area and store it in the header of each
2392 Use xxhash as the checksum function. Generally the fastest software
2393 checksum that fio supports.
2396 Use sha512 as the checksum function.
2399 Use sha256 as the checksum function.
2402 Use optimized sha1 as the checksum function.
2405 Use optimized sha3-224 as the checksum function.
2408 Use optimized sha3-256 as the checksum function.
2411 Use optimized sha3-384 as the checksum function.
2414 Use optimized sha3-512 as the checksum function.
2417 This option is deprecated, since now meta information is included in
2418 generic verification header and meta verification happens by
2419 default. For detailed information see the description of the
2420 :option:`verify` setting. This option is kept because of
2421 compatibility's sake with old configurations. Do not use it.
2424 Verify a strict pattern. Normally fio includes a header with some
2425 basic information and checksumming, but if this option is set, only
2426 the specific pattern set with :option:`verify_pattern` is verified.
2429 Only pretend to verify. Useful for testing internals with
2430 :option:`ioengine`\=null, not for much else.
2432 This option can be used for repeated burn-in tests of a system to make sure
2433 that the written data is also correctly read back. If the data direction
2434 given is a read or random read, fio will assume that it should verify a
2435 previously written file. If the data direction includes any form of write,
2436 the verify will be of the newly written data.
2438 .. option:: verifysort=bool
2440 If true, fio will sort written verify blocks when it deems it faster to read
2441 them back in a sorted manner. This is often the case when overwriting an
2442 existing file, since the blocks are already laid out in the file system. You
2443 can ignore this option unless doing huge amounts of really fast I/O where
2444 the red-black tree sorting CPU time becomes significant. Default: true.
2446 .. option:: verifysort_nr=int
2448 Pre-load and sort verify blocks for a read workload.
2450 .. option:: verify_offset=int
2452 Swap the verification header with data somewhere else in the block before
2453 writing. It is swapped back before verifying.
2455 .. option:: verify_interval=int
2457 Write the verification header at a finer granularity than the
2458 :option:`blocksize`. It will be written for chunks the size of
2459 ``verify_interval``. :option:`blocksize` should divide this evenly.
2461 .. option:: verify_pattern=str
2463 If set, fio will fill the I/O buffers with this pattern. Fio defaults to
2464 filling with totally random bytes, but sometimes it's interesting to fill
2465 with a known pattern for I/O verification purposes. Depending on the width
2466 of the pattern, fio will fill 1/2/3/4 bytes of the buffer at the time (it can
2467 be either a decimal or a hex number). The ``verify_pattern`` if larger than
2468 a 32-bit quantity has to be a hex number that starts with either "0x" or
2469 "0X". Use with :option:`verify`. Also, ``verify_pattern`` supports %o
2470 format, which means that for each block offset will be written and then
2471 verified back, e.g.::
2475 Or use combination of everything::
2477 verify_pattern=0xff%o"abcd"-12
2479 .. option:: verify_fatal=bool
2481 Normally fio will keep checking the entire contents before quitting on a
2482 block verification failure. If this option is set, fio will exit the job on
2483 the first observed failure. Default: false.
2485 .. option:: verify_dump=bool
2487 If set, dump the contents of both the original data block and the data block
2488 we read off disk to files. This allows later analysis to inspect just what
2489 kind of data corruption occurred. Off by default.
2491 .. option:: verify_async=int
2493 Fio will normally verify I/O inline from the submitting thread. This option
2494 takes an integer describing how many async offload threads to create for I/O
2495 verification instead, causing fio to offload the duty of verifying I/O
2496 contents to one or more separate threads. If using this offload option, even
2497 sync I/O engines can benefit from using an :option:`iodepth` setting higher
2498 than 1, as it allows them to have I/O in flight while verifies are running.
2499 Defaults to 0 async threads, i.e. verification is not asynchronous.
2501 .. option:: verify_async_cpus=str
2503 Tell fio to set the given CPU affinity on the async I/O verification
2504 threads. See :option:`cpus_allowed` for the format used.
2506 .. option:: verify_backlog=int
2508 Fio will normally verify the written contents of a job that utilizes verify
2509 once that job has completed. In other words, everything is written then
2510 everything is read back and verified. You may want to verify continually
2511 instead for a variety of reasons. Fio stores the meta data associated with
2512 an I/O block in memory, so for large verify workloads, quite a bit of memory
2513 would be used up holding this meta data. If this option is enabled, fio will
2514 write only N blocks before verifying these blocks.
2516 .. option:: verify_backlog_batch=int
2518 Control how many blocks fio will verify if :option:`verify_backlog` is
2519 set. If not set, will default to the value of :option:`verify_backlog`
2520 (meaning the entire queue is read back and verified). If
2521 ``verify_backlog_batch`` is less than :option:`verify_backlog` then not all
2522 blocks will be verified, if ``verify_backlog_batch`` is larger than
2523 :option:`verify_backlog`, some blocks will be verified more than once.
2525 .. option:: verify_state_save=bool
2527 When a job exits during the write phase of a verify workload, save its
2528 current state. This allows fio to replay up until that point, if the verify
2529 state is loaded for the verify read phase. The format of the filename is,
2532 <type>-<jobname>-<jobindex>-verify.state.
2534 <type> is "local" for a local run, "sock" for a client/server socket
2535 connection, and "ip" (192.168.0.1, for instance) for a networked
2536 client/server connection. Defaults to true.
2538 .. option:: verify_state_load=bool
2540 If a verify termination trigger was used, fio stores the current write state
2541 of each thread. This can be used at verification time so that fio knows how
2542 far it should verify. Without this information, fio will run a full
2543 verification pass, according to the settings in the job file used. Default
2546 .. option:: trim_percentage=int
2548 Number of verify blocks to discard/trim.
2550 .. option:: trim_verify_zero=bool
2552 Verify that trim/discarded blocks are returned as zeros.
2554 .. option:: trim_backlog=int
2556 Verify that trim/discarded blocks are returned as zeros.
2558 .. option:: trim_backlog_batch=int
2560 Trim this number of I/O blocks.
2562 .. option:: experimental_verify=bool
2564 Enable experimental verification.
2570 .. option:: steadystate=str:float, ss=str:float
2572 Define the criterion and limit for assessing steady state performance. The
2573 first parameter designates the criterion whereas the second parameter sets
2574 the threshold. When the criterion falls below the threshold for the
2575 specified duration, the job will stop. For example, `iops_slope:0.1%` will
2576 direct fio to terminate the job when the least squares regression slope
2577 falls below 0.1% of the mean IOPS. If :option:`group_reporting` is enabled
2578 this will apply to all jobs in the group. Below is the list of available
2579 steady state assessment criteria. All assessments are carried out using only
2580 data from the rolling collection window. Threshold limits can be expressed
2581 as a fixed value or as a percentage of the mean in the collection window.
2584 Collect IOPS data. Stop the job if all individual IOPS measurements
2585 are within the specified limit of the mean IOPS (e.g., ``iops:2``
2586 means that all individual IOPS values must be within 2 of the mean,
2587 whereas ``iops:0.2%`` means that all individual IOPS values must be
2588 within 0.2% of the mean IOPS to terminate the job).
2591 Collect IOPS data and calculate the least squares regression
2592 slope. Stop the job if the slope falls below the specified limit.
2595 Collect bandwidth data. Stop the job if all individual bandwidth
2596 measurements are within the specified limit of the mean bandwidth.
2599 Collect bandwidth data and calculate the least squares regression
2600 slope. Stop the job if the slope falls below the specified limit.
2602 .. option:: steadystate_duration=time, ss_dur=time
2604 A rolling window of this duration will be used to judge whether steady state
2605 has been reached. Data will be collected once per second. The default is 0
2606 which disables steady state detection. When the unit is omitted, the
2607 value is interpreted in seconds.
2609 .. option:: steadystate_ramp_time=time, ss_ramp=time
2611 Allow the job to run for the specified duration before beginning data
2612 collection for checking the steady state job termination criterion. The
2613 default is 0. When the unit is omitted, the value is interpreted in seconds.
2616 Measurements and reporting
2617 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2619 .. option:: per_job_logs=bool
2621 If set, this generates bw/clat/iops log with per file private filenames. If
2622 not set, jobs with identical names will share the log filename. Default:
2625 .. option:: group_reporting
2627 It may sometimes be interesting to display statistics for groups of jobs as
2628 a whole instead of for each individual job. This is especially true if
2629 :option:`numjobs` is used; looking at individual thread/process output
2630 quickly becomes unwieldy. To see the final report per-group instead of
2631 per-job, use :option:`group_reporting`. Jobs in a file will be part of the
2632 same reporting group, unless if separated by a :option:`stonewall`, or by
2633 using :option:`new_group`.
2635 .. option:: new_group
2637 Start a new reporting group. See: :option:`group_reporting`. If not given,
2638 all jobs in a file will be part of the same reporting group, unless
2639 separated by a :option:`stonewall`.
2643 By default, fio collects and shows final output results for all jobs
2644 that run. If this option is set to 0, then fio will ignore it in
2645 the final stat output.
2647 .. option:: write_bw_log=str
2649 If given, write a bandwidth log for this job. Can be used to store data of
2650 the bandwidth of the jobs in their lifetime. The included
2651 :command:`fio_generate_plots` script uses :command:`gnuplot` to turn these
2652 text files into nice graphs. See :option:`write_lat_log` for behavior of
2653 given filename. For this option, the postfix is :file:`_bw.x.log`, where `x`
2654 is the index of the job (`1..N`, where `N` is the number of jobs). If
2655 :option:`per_job_logs` is false, then the filename will not include the job
2656 index. See `Log File Formats`_.
2658 .. option:: write_lat_log=str
2660 Same as :option:`write_bw_log`, except that this option stores I/O
2661 submission, completion, and total latencies instead. If no filename is given
2662 with this option, the default filename of :file:`jobname_type.log` is
2663 used. Even if the filename is given, fio will still append the type of
2664 log. So if one specifies::
2668 The actual log names will be :file:`foo_slat.x.log`, :file:`foo_clat.x.log`,
2669 and :file:`foo_lat.x.log`, where `x` is the index of the job (1..N, where N
2670 is the number of jobs). This helps :command:`fio_generate_plot` find the
2671 logs automatically. If :option:`per_job_logs` is false, then the filename
2672 will not include the job index. See `Log File Formats`_.
2674 .. option:: write_hist_log=str
2676 Same as :option:`write_lat_log`, but writes I/O completion latency
2677 histograms. If no filename is given with this option, the default filename
2678 of :file:`jobname_clat_hist.x.log` is used, where `x` is the index of the
2679 job (1..N, where `N` is the number of jobs). Even if the filename is given,
2680 fio will still append the type of log. If :option:`per_job_logs` is false,
2681 then the filename will not include the job index. See `Log File Formats`_.
2683 .. option:: write_iops_log=str
2685 Same as :option:`write_bw_log`, but writes IOPS. If no filename is given
2686 with this option, the default filename of :file:`jobname_type.x.log` is
2687 used,where `x` is the index of the job (1..N, where `N` is the number of
2688 jobs). Even if the filename is given, fio will still append the type of
2689 log. If :option:`per_job_logs` is false, then the filename will not include
2690 the job index. See `Log File Formats`_.
2692 .. option:: log_avg_msec=int
2694 By default, fio will log an entry in the iops, latency, or bw log for every
2695 I/O that completes. When writing to the disk log, that can quickly grow to a
2696 very large size. Setting this option makes fio average the each log entry
2697 over the specified period of time, reducing the resolution of the log. See
2698 :option:`log_max_value` as well. Defaults to 0, logging all entries.
2700 .. option:: log_hist_msec=int
2702 Same as :option:`log_avg_msec`, but logs entries for completion latency
2703 histograms. Computing latency percentiles from averages of intervals using
2704 :option:`log_avg_msec` is inaccurate. Setting this option makes fio log
2705 histogram entries over the specified period of time, reducing log sizes for
2706 high IOPS devices while retaining percentile accuracy. See
2707 :option:`log_hist_coarseness` as well. Defaults to 0, meaning histogram
2708 logging is disabled.
2710 .. option:: log_hist_coarseness=int
2712 Integer ranging from 0 to 6, defining the coarseness of the resolution of
2713 the histogram logs enabled with :option:`log_hist_msec`. For each increment
2714 in coarseness, fio outputs half as many bins. Defaults to 0, for which
2715 histogram logs contain 1216 latency bins. See `Log File Formats`_.
2717 .. option:: log_max_value=bool
2719 If :option:`log_avg_msec` is set, fio logs the average over that window. If
2720 you instead want to log the maximum value, set this option to 1. Defaults to
2721 0, meaning that averaged values are logged.
2723 .. option:: log_offset=int
2725 If this is set, the iolog options will include the byte offset for the I/O
2726 entry as well as the other data values.
2728 .. option:: log_compression=int
2730 If this is set, fio will compress the I/O logs as it goes, to keep the
2731 memory footprint lower. When a log reaches the specified size, that chunk is
2732 removed and compressed in the background. Given that I/O logs are fairly
2733 highly compressible, this yields a nice memory savings for longer runs. The
2734 downside is that the compression will consume some background CPU cycles, so
2735 it may impact the run. This, however, is also true if the logging ends up
2736 consuming most of the system memory. So pick your poison. The I/O logs are
2737 saved normally at the end of a run, by decompressing the chunks and storing
2738 them in the specified log file. This feature depends on the availability of
2741 .. option:: log_compression_cpus=str
2743 Define the set of CPUs that are allowed to handle online log compression for
2744 the I/O jobs. This can provide better isolation between performance
2745 sensitive jobs, and background compression work.
2747 .. option:: log_store_compressed=bool
2749 If set, fio will store the log files in a compressed format. They can be
2750 decompressed with fio, using the :option:`--inflate-log` command line
2751 parameter. The files will be stored with a :file:`.fz` suffix.
2753 .. option:: log_unix_epoch=bool
2755 If set, fio will log Unix timestamps to the log files produced by enabling
2756 write_type_log for each log type, instead of the default zero-based
2759 .. option:: block_error_percentiles=bool
2761 If set, record errors in trim block-sized units from writes and trims and
2762 output a histogram of how many trims it took to get to errors, and what kind
2763 of error was encountered.
2765 .. option:: bwavgtime=int
2767 Average the calculated bandwidth over the given time. Value is specified in
2768 milliseconds. If the job also does bandwidth logging through
2769 :option:`write_bw_log`, then the minimum of this option and
2770 :option:`log_avg_msec` will be used. Default: 500ms.
2772 .. option:: iopsavgtime=int
2774 Average the calculated IOPS over the given time. Value is specified in
2775 milliseconds. If the job also does IOPS logging through
2776 :option:`write_iops_log`, then the minimum of this option and
2777 :option:`log_avg_msec` will be used. Default: 500ms.
2779 .. option:: disk_util=bool
2781 Generate disk utilization statistics, if the platform supports it.
2784 .. option:: disable_lat=bool
2786 Disable measurements of total latency numbers. Useful only for cutting back
2787 the number of calls to :manpage:`gettimeofday(2)`, as that does impact
2788 performance at really high IOPS rates. Note that to really get rid of a
2789 large amount of these calls, this option must be used with
2790 :option:`disable_slat` and :option:`disable_bw_measurement` as well.
2792 .. option:: disable_clat=bool
2794 Disable measurements of completion latency numbers. See
2795 :option:`disable_lat`.
2797 .. option:: disable_slat=bool
2799 Disable measurements of submission latency numbers. See
2800 :option:`disable_slat`.
2802 .. option:: disable_bw_measurement=bool, disable_bw=bool
2804 Disable measurements of throughput/bandwidth numbers. See
2805 :option:`disable_lat`.
2807 .. option:: clat_percentiles=bool
2809 Enable the reporting of percentiles of completion latencies.
2811 .. option:: percentile_list=float_list
2813 Overwrite the default list of percentiles for completion latencies and the
2814 block error histogram. Each number is a floating number in the range
2815 (0,100], and the maximum length of the list is 20. Use ``:`` to separate the
2816 numbers, and list the numbers in ascending order. For example,
2817 ``--percentile_list=99.5:99.9`` will cause fio to report the values of
2818 completion latency below which 99.5% and 99.9% of the observed latencies
2825 .. option:: exitall_on_error
2827 When one job finishes in error, terminate the rest. The default is to wait
2828 for each job to finish.
2830 .. option:: continue_on_error=str
2832 Normally fio will exit the job on the first observed failure. If this option
2833 is set, fio will continue the job when there is a 'non-fatal error' (EIO or
2834 EILSEQ) until the runtime is exceeded or the I/O size specified is
2835 completed. If this option is used, there are two more stats that are
2836 appended, the total error count and the first error. The error field given
2837 in the stats is the first error that was hit during the run.
2839 The allowed values are:
2842 Exit on any I/O or verify errors.
2845 Continue on read errors, exit on all others.
2848 Continue on write errors, exit on all others.
2851 Continue on any I/O error, exit on all others.
2854 Continue on verify errors, exit on all others.
2857 Continue on all errors.
2860 Backward-compatible alias for 'none'.
2863 Backward-compatible alias for 'all'.
2865 .. option:: ignore_error=str
2867 Sometimes you want to ignore some errors during test in that case you can
2868 specify error list for each error type, instead of only being able to
2869 ignore the default 'non-fatal error' using :option:`continue_on_error`.
2870 ``ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST`` errors for
2871 given error type is separated with ':'. Error may be symbol ('ENOSPC',
2872 'ENOMEM') or integer. Example::
2874 ignore_error=EAGAIN,ENOSPC:122
2876 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from
2877 WRITE. This option works by overriding :option:`continue_on_error` with
2878 the list of errors for each error type if any.
2880 .. option:: error_dump=bool
2882 If set dump every error even if it is non fatal, true by default. If
2883 disabled only fatal error will be dumped.
2885 Running predefined workloads
2886 ----------------------------
2888 Fio includes predefined profiles that mimic the I/O workloads generated by
2891 .. option:: profile=str
2893 The predefined workload to run. Current profiles are:
2896 Threaded I/O bench (tiotest/tiobench) like workload.
2899 Aerospike Certification Tool (ACT) like workload.
2901 To view a profile's additional options use :option:`--cmdhelp` after specifying
2902 the profile. For example::
2904 $ fio --profile=act --cmdhelp
2909 .. option:: device-names=str
2914 .. option:: load=int
2917 ACT load multiplier. Default: 1.
2919 .. option:: test-duration=time
2922 How long the entire test takes to run. When the unit is omitted, the value
2923 is given in seconds. Default: 24h.
2925 .. option:: threads-per-queue=int
2928 Number of read IO threads per device. Default: 8.
2930 .. option:: read-req-num-512-blocks=int
2933 Number of 512B blocks to read at the time. Default: 3.
2935 .. option:: large-block-op-kbytes=int
2938 Size of large block ops in KiB (writes). Default: 131072.
2943 Set to run ACT prep phase.
2945 Tiobench profile options
2946 ~~~~~~~~~~~~~~~~~~~~~~~~
2948 .. option:: size=str
2953 .. option:: block=int
2956 Block size in bytes. Default: 4096.
2958 .. option:: numruns=int
2968 .. option:: threads=int
2973 Interpreting the output
2974 -----------------------
2976 Fio spits out a lot of output. While running, fio will display the status of the
2977 jobs created. An example of that would be::
2979 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]
2981 The characters inside the square brackets denote the current status of each
2982 thread. The possible values (in typical life cycle order) are:
2984 +------+-----+-----------------------------------------------------------+
2986 +======+=====+===========================================================+
2987 | P | | Thread setup, but not started. |
2988 +------+-----+-----------------------------------------------------------+
2989 | C | | Thread created. |
2990 +------+-----+-----------------------------------------------------------+
2991 | I | | Thread initialized, waiting or generating necessary data. |
2992 +------+-----+-----------------------------------------------------------+
2993 | | p | Thread running pre-reading file(s). |
2994 +------+-----+-----------------------------------------------------------+
2995 | | R | Running, doing sequential reads. |
2996 +------+-----+-----------------------------------------------------------+
2997 | | r | Running, doing random reads. |
2998 +------+-----+-----------------------------------------------------------+
2999 | | W | Running, doing sequential writes. |
3000 +------+-----+-----------------------------------------------------------+
3001 | | w | Running, doing random writes. |
3002 +------+-----+-----------------------------------------------------------+
3003 | | M | Running, doing mixed sequential reads/writes. |
3004 +------+-----+-----------------------------------------------------------+
3005 | | m | Running, doing mixed random reads/writes. |
3006 +------+-----+-----------------------------------------------------------+
3007 | | F | Running, currently waiting for :manpage:`fsync(2)` |
3008 +------+-----+-----------------------------------------------------------+
3009 | | V | Running, doing verification of written data. |
3010 +------+-----+-----------------------------------------------------------+
3011 | E | | Thread exited, not reaped by main thread yet. |
3012 +------+-----+-----------------------------------------------------------+
3013 | _ | | Thread reaped, or |
3014 +------+-----+-----------------------------------------------------------+
3015 | X | | Thread reaped, exited with an error. |
3016 +------+-----+-----------------------------------------------------------+
3017 | K | | Thread reaped, exited due to signal. |
3018 +------+-----+-----------------------------------------------------------+
3020 Fio will condense the thread string as not to take up more space on the command
3021 line as is needed. For instance, if you have 10 readers and 10 writers running,
3022 the output would look like this::
3024 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]
3026 Fio will still maintain the ordering, though. So the above means that jobs 1..10
3027 are readers, and 11..20 are writers.
3029 The other values are fairly self explanatory -- number of threads currently
3030 running and doing I/O, the number of currently open files (f=), the rate of I/O
3031 since last check (read speed listed first, then write speed and optionally trim
3032 speed), and the estimated completion percentage and time for the current
3033 running group. It's impossible to estimate runtime of the following groups (if
3034 any). Note that the string is displayed in order, so it's possible to tell which
3035 of the jobs are currently doing what. The first character is the first job
3036 defined in the job file, and so forth.
3038 When fio is done (or interrupted by :kbd:`ctrl-c`), it will show the data for
3039 each thread, group of threads, and disks in that order. For each data direction,
3040 the output looks like::
3042 Client1 (g=0): err= 0:
3043 write: io= 32MiB, bw= 666KiB/s, iops=89 , runt= 50320msec
3044 slat (msec): min= 0, max= 136, avg= 0.03, stdev= 1.92
3045 clat (msec): min= 0, max= 631, avg=48.50, stdev=86.82
3046 bw (KiB/s) : min= 0, max= 1196, per=51.00%, avg=664.02, stdev=681.68
3047 cpu : usr=1.49%, sys=0.25%, ctx=7969, majf=0, minf=17
3048 IO depths : 1=0.1%, 2=0.3%, 4=0.5%, 8=99.0%, 16=0.0%, 32=0.0%, >32=0.0%
3049 submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3050 complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
3051 issued r/w: total=0/32768, short=0/0
3052 lat (msec): 2=1.6%, 4=0.0%, 10=3.2%, 20=12.8%, 50=38.4%, 100=24.8%,
3053 lat (msec): 250=15.2%, 500=0.0%, 750=0.0%, 1000=0.0%, >=2048=0.0%
3055 The client number is printed, along with the group id and error of that
3056 thread. Below is the I/O statistics, here for writes. In the order listed, they
3060 Number of megabytes I/O performed.
3063 Average bandwidth rate.
3066 Average I/Os performed per second.
3069 The runtime of that thread.
3072 Submission latency (avg being the average, stdev being the standard
3073 deviation). This is the time it took to submit the I/O. For sync I/O,
3074 the slat is really the completion latency, since queue/complete is one
3075 operation there. This value can be in milliseconds or microseconds, fio
3076 will choose the most appropriate base and print that. In the example
3077 above, milliseconds is the best scale. Note: in :option:`--minimal` mode
3078 latencies are always expressed in microseconds.
3081 Completion latency. Same names as slat, this denotes the time from
3082 submission to completion of the I/O pieces. For sync I/O, clat will
3083 usually be equal (or very close) to 0, as the time from submit to
3084 complete is basically just CPU time (I/O has already been done, see slat
3088 Bandwidth. Same names as the xlat stats, but also includes an
3089 approximate percentage of total aggregate bandwidth this thread received
3090 in this group. This last value is only really useful if the threads in
3091 this group are on the same disk, since they are then competing for disk
3095 CPU usage. User and system time, along with the number of context
3096 switches this thread went through, usage of system and user time, and
3097 finally the number of major and minor page faults. The CPU utilization
3098 numbers are averages for the jobs in that reporting group, while the
3099 context and fault counters are summed.
3102 The distribution of I/O depths over the job life time. The numbers are
3103 divided into powers of 2, so for example the 16= entries includes depths
3104 up to that value but higher than the previous entry. In other words, it
3105 covers the range from 16 to 31.
3108 How many pieces of I/O were submitting in a single submit call. Each
3109 entry denotes that amount and below, until the previous entry -- e.g.,
3110 8=100% mean that we submitted anywhere in between 5-8 I/Os per submit
3114 Like the above submit number, but for completions instead.
3117 The number of read/write requests issued, and how many of them were
3121 The distribution of I/O completion latencies. This is the time from when
3122 I/O leaves fio and when it gets completed. The numbers follow the same
3123 pattern as the I/O depths, meaning that 2=1.6% means that 1.6% of the
3124 I/O completed within 2 msecs, 20=12.8% means that 12.8% of the I/O took
3125 more than 10 msecs, but less than (or equal to) 20 msecs.
3127 After each client has been listed, the group statistics are printed. They
3128 will look like this::
3130 Run status group 0 (all jobs):
3131 READ: io=64MB, aggrb=22178, minb=11355, maxb=11814, mint=2840msec, maxt=2955msec
3132 WRITE: io=64MB, aggrb=1302, minb=666, maxb=669, mint=50093msec, maxt=50320msec
3134 For each data direction, it prints:
3137 Number of megabytes I/O performed.
3139 Aggregate bandwidth of threads in this group.
3141 The minimum average bandwidth a thread saw.
3143 The maximum average bandwidth a thread saw.
3145 The smallest runtime of the threads in that group.
3147 The longest runtime of the threads in that group.
3149 And finally, the disk statistics are printed. They will look like this::
3151 Disk stats (read/write):
3152 sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00%
3154 Each value is printed for both reads and writes, with reads first. The
3158 Number of I/Os performed by all groups.
3160 Number of merges I/O the I/O scheduler.
3162 Number of ticks we kept the disk busy.
3164 Total time spent in the disk queue.
3166 The disk utilization. A value of 100% means we kept the disk
3167 busy constantly, 50% would be a disk idling half of the time.
3169 It is also possible to get fio to dump the current output while it is running,
3170 without terminating the job. To do that, send fio the **USR1** signal. You can
3171 also get regularly timed dumps by using the :option:`--status-interval`
3172 parameter, or by creating a file in :file:`/tmp` named
3173 :file:`fio-dump-status`. If fio sees this file, it will unlink it and dump the
3174 current output status.
3180 For scripted usage where you typically want to generate tables or graphs of the
3181 results, fio can output the results in a semicolon separated format. The format
3182 is one long line of values, such as::
3184 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%
3185 A description of this job goes here.
3187 The job description (if provided) follows on a second line.
3189 To enable terse output, use the :option:`--minimal` command line option. The
3190 first value is the version of the terse output format. If the output has to be
3191 changed for some reason, this number will be incremented by 1 to signify that
3194 Split up, the format is as follows (comments in brackets denote when a
3195 field was introduced or whether its specific to some terse version):
3199 terse version, fio version [v3], jobname, groupid, error
3203 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
3204 Submission latency: min, max, mean, stdev (usec)
3205 Completion latency: min, max, mean, stdev (usec)
3206 Completion latency percentiles: 20 fields (see below)
3207 Total latency: min, max, mean, stdev (usec)
3208 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
3209 IOPS [v5]: min, max, mean, stdev, number of samples
3215 Total IO (KiB), bandwidth (KiB/sec), IOPS, runtime (msec)
3216 Submission latency: min, max, mean, stdev (usec)
3217 Completion latency: min, max, mean, stdev (usec)
3218 Completion latency percentiles: 20 fields (see below)
3219 Total latency: min, max, mean, stdev (usec)
3220 Bw (KiB/s): min, max, aggregate percentage of total, mean, stdev, number of samples [v5]
3221 IOPS [v5]: min, max, mean, stdev, number of samples
3223 TRIM status [all but version 3]:
3225 Fields are similar to READ/WRITE status.
3229 user, system, context switches, major faults, minor faults
3233 <=1, 2, 4, 8, 16, 32, >=64
3235 I/O latencies microseconds::
3237 <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
3239 I/O latencies milliseconds::
3241 <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
3243 Disk utilization [v3]::
3245 Disk name, Read ios, write ios,
3246 Read merges, write merges,
3247 Read ticks, write ticks,
3248 Time spent in queue, disk utilization percentage
3250 Additional Info (dependent on continue_on_error, default off)::
3252 total # errors, first error code
3254 Additional Info (dependent on description being set)::
3258 Completion latency percentiles can be a grouping of up to 20 sets, so for the
3259 terse output fio writes all of them. Each field will look like this::
3263 which is the Xth percentile, and the `usec` latency associated with it.
3265 For disk utilization, all disks used by fio are shown. So for each disk there
3266 will be a disk utilization section.
3268 Below is a single line containing short names for each of the fields in the
3269 minimal output v3, separated by semicolons::
3271 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_max;read_clat_min;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_max;write_clat_min;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;pu_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
3277 There are two trace file format that you can encounter. The older (v1) format is
3278 unsupported since version 1.20-rc3 (March 2008). It will still be described
3279 below in case that you get an old trace and want to understand it.
3281 In any case the trace is a simple text file with a single action per line.
3284 Trace file format v1
3285 ~~~~~~~~~~~~~~~~~~~~
3287 Each line represents a single I/O action in the following format::
3291 where `rw=0/1` for read/write, and the offset and length entries being in bytes.
3293 This format is not supported in fio versions >= 1.20-rc3.
3296 Trace file format v2
3297 ~~~~~~~~~~~~~~~~~~~~
3299 The second version of the trace file format was added in fio version 1.17. It
3300 allows to access more then one file per trace and has a bigger set of possible
3303 The first line of the trace file has to be::
3307 Following this can be lines in two different formats, which are described below.
3309 The file management format::
3313 The filename is given as an absolute path. The action can be one of these:
3316 Add the given filename to the trace.
3318 Open the file with the given filename. The filename has to have
3319 been added with the **add** action before.
3321 Close the file with the given filename. The file has to have been
3325 The file I/O action format::
3327 filename action offset length
3329 The `filename` is given as an absolute path, and has to have been added and
3330 opened before it can be used with this format. The `offset` and `length` are
3331 given in bytes. The `action` can be one of these:
3334 Wait for `offset` microseconds. Everything below 100 is discarded.
3335 The time is relative to the previous `wait` statement.
3337 Read `length` bytes beginning from `offset`.
3339 Write `length` bytes beginning from `offset`.
3341 :manpage:`fsync(2)` the file.
3343 :manpage:`fdatasync(2)` the file.
3345 Trim the given file from the given `offset` for `length` bytes.
3347 CPU idleness profiling
3348 ----------------------
3350 In some cases, we want to understand CPU overhead in a test. For example, we
3351 test patches for the specific goodness of whether they reduce CPU usage.
3352 Fio implements a balloon approach to create a thread per CPU that runs at idle
3353 priority, meaning that it only runs when nobody else needs the cpu.
3354 By measuring the amount of work completed by the thread, idleness of each CPU
3355 can be derived accordingly.
3357 An unit work is defined as touching a full page of unsigned characters. Mean and
3358 standard deviation of time to complete an unit work is reported in "unit work"
3359 section. Options can be chosen to report detailed percpu idleness or overall
3360 system idleness by aggregating percpu stats.
3363 Verification and triggers
3364 -------------------------
3366 Fio is usually run in one of two ways, when data verification is done. The first
3367 is a normal write job of some sort with verify enabled. When the write phase has
3368 completed, fio switches to reads and verifies everything it wrote. The second
3369 model is running just the write phase, and then later on running the same job
3370 (but with reads instead of writes) to repeat the same I/O patterns and verify
3371 the contents. Both of these methods depend on the write phase being completed,
3372 as fio otherwise has no idea how much data was written.
3374 With verification triggers, fio supports dumping the current write state to
3375 local files. Then a subsequent read verify workload can load this state and know
3376 exactly where to stop. This is useful for testing cases where power is cut to a
3377 server in a managed fashion, for instance.
3379 A verification trigger consists of two things:
3381 1) Storing the write state of each job.
3382 2) Executing a trigger command.
3384 The write state is relatively small, on the order of hundreds of bytes to single
3385 kilobytes. It contains information on the number of completions done, the last X
3388 A trigger is invoked either through creation ('touch') of a specified file in
3389 the system, or through a timeout setting. If fio is run with
3390 :option:`--trigger-file`\= :file:`/tmp/trigger-file`, then it will continually
3391 check for the existence of :file:`/tmp/trigger-file`. When it sees this file, it
3392 will fire off the trigger (thus saving state, and executing the trigger
3395 For client/server runs, there's both a local and remote trigger. If fio is
3396 running as a server backend, it will send the job states back to the client for
3397 safe storage, then execute the remote trigger, if specified. If a local trigger
3398 is specified, the server will still send back the write state, but the client
3399 will then execute the trigger.
3401 Verification trigger example
3402 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3404 Let's say we want to run a powercut test on the remote machine 'server'. Our
3405 write workload is in :file:`write-test.fio`. We want to cut power to 'server' at
3406 some point during the run, and we'll run this test from the safety or our local
3407 machine, 'localbox'. On the server, we'll start the fio backend normally::
3409 server# fio --server
3411 and on the client, we'll fire off the workload::
3413 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger-remote="bash -c \"echo b > /proc/sysrq-triger\""
3415 We set :file:`/tmp/my-trigger` as the trigger file, and we tell fio to execute::
3417 echo b > /proc/sysrq-trigger
3419 on the server once it has received the trigger and sent us the write state. This
3420 will work, but it's not **really** cutting power to the server, it's merely
3421 abruptly rebooting it. If we have a remote way of cutting power to the server
3422 through IPMI or similar, we could do that through a local trigger command
3423 instead. Let's assume we have a script that does IPMI reboot of a given hostname,
3424 ipmi-reboot. On localbox, we could then have run fio with a local trigger
3427 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger="ipmi-reboot server"
3429 For this case, fio would wait for the server to send us the write state, then
3430 execute ``ipmi-reboot server`` when that happened.
3432 Loading verify state
3433 ~~~~~~~~~~~~~~~~~~~~
3435 To load stored write state, a read verification job file must contain the
3436 :option:`verify_state_load` option. If that is set, fio will load the previously
3437 stored state. For a local fio run this is done by loading the files directly,
3438 and on a client/server run, the server backend will ask the client to send the
3439 files over and load them from there.
3445 Fio supports a variety of log file formats, for logging latencies, bandwidth,
3446 and IOPS. The logs share a common format, which looks like this:
3448 *time* (`msec`), *value*, *data direction*, *offset*
3450 Time for the log entry is always in milliseconds. The *value* logged depends
3451 on the type of log, it will be one of the following:
3454 Value is latency in usecs
3460 *Data direction* is one of the following:
3469 The *offset* is the offset, in bytes, from the start of the file, for that
3470 particular I/O. The logging of the offset can be toggled with
3471 :option:`log_offset`.
3473 If windowed logging is enabled through :option:`log_avg_msec` then fio doesn't
3474 log individual I/Os. Instead of logs the average values over the specified period
3475 of time. Since 'data direction' and 'offset' are per-I/O values, they aren't
3476 applicable if windowed logging is enabled. If windowed logging is enabled and
3477 :option:`log_max_value` is set, then fio logs maximum values in that window
3478 instead of averages.
3484 Normally fio is invoked as a stand-alone application on the machine where the
3485 I/O workload should be generated. However, the frontend and backend of fio can
3486 be run separately. Ie the fio server can generate an I/O workload on the "Device
3487 Under Test" while being controlled from another machine.
3489 Start the server on the machine which has access to the storage DUT::
3493 where args defines what fio listens to. The arguments are of the form
3494 ``type,hostname`` or ``IP,port``. *type* is either ``ip`` (or ip4) for TCP/IP
3495 v4, ``ip6`` for TCP/IP v6, or ``sock`` for a local unix domain socket.
3496 *hostname* is either a hostname or IP address, and *port* is the port to listen
3497 to (only valid for TCP/IP, not a local socket). Some examples:
3501 Start a fio server, listening on all interfaces on the default port (8765).
3503 2) ``fio --server=ip:hostname,4444``
3505 Start a fio server, listening on IP belonging to hostname and on port 4444.
3507 3) ``fio --server=ip6:::1,4444``
3509 Start a fio server, listening on IPv6 localhost ::1 and on port 4444.
3511 4) ``fio --server=,4444``
3513 Start a fio server, listening on all interfaces on port 4444.
3515 5) ``fio --server=1.2.3.4``
3517 Start a fio server, listening on IP 1.2.3.4 on the default port.
3519 6) ``fio --server=sock:/tmp/fio.sock``
3521 Start a fio server, listening on the local socket /tmp/fio.sock.
3523 Once a server is running, a "client" can connect to the fio server with::
3525 fio <local-args> --client=<server> <remote-args> <job file(s)>
3527 where `local-args` are arguments for the client where it is running, `server`
3528 is the connect string, and `remote-args` and `job file(s)` are sent to the
3529 server. The `server` string follows the same format as it does on the server
3530 side, to allow IP/hostname/socket and port strings.
3532 Fio can connect to multiple servers this way::
3534 fio --client=<server1> <job file(s)> --client=<server2> <job file(s)>
3536 If the job file is located on the fio server, then you can tell the server to
3537 load a local file as well. This is done by using :option:`--remote-config` ::
3539 fio --client=server --remote-config /path/to/file.fio
3541 Then fio will open this local (to the server) job file instead of being passed
3542 one from the client.
3544 If you have many servers (example: 100 VMs/containers), you can input a pathname
3545 of a file containing host IPs/names as the parameter value for the
3546 :option:`--client` option. For example, here is an example :file:`host.list`
3547 file containing 2 hostnames::
3549 host1.your.dns.domain
3550 host2.your.dns.domain
3552 The fio command would then be::
3554 fio --client=host.list <job file(s)>
3556 In this mode, you cannot input server-specific parameters or job files -- all
3557 servers receive the same job file.
3559 In order to let ``fio --client`` runs use a shared filesystem from multiple
3560 hosts, ``fio --client`` now prepends the IP address of the server to the
3561 filename. For example, if fio is using the directory :file:`/mnt/nfs/fio` and is
3562 writing filename :file:`fileio.tmp`, with a :option:`--client` `hostfile`
3563 containing two hostnames ``h1`` and ``h2`` with IP addresses 192.168.10.120 and
3564 192.168.10.121, then fio will create two files::
3566 /mnt/nfs/fio/192.168.10.120.fileio.tmp
3567 /mnt/nfs/fio/192.168.10.121.fileio.tmp