1 .TH fio 1 "October 2013" "User Manual"
3 fio \- flexible I/O tester
6 [\fIoptions\fR] [\fIjobfile\fR]...
9 is a tool that will spawn a number of threads or processes doing a
10 particular type of I/O action as specified by the user.
11 The typical use of fio is to write a job file matching the I/O load
12 one wants to simulate.
15 .BI \-\-debug \fR=\fPtype
16 Enable verbose tracing of various fio actions. May be `all' for all types
17 or individual types separated by a comma (eg \-\-debug=io,file). `help' will
18 list all available tracing options.
20 .BI \-\-output \fR=\fPfilename
21 Write output to \fIfilename\fR.
23 .BI \-\-runtime \fR=\fPruntime
24 Limit run time to \fIruntime\fR seconds.
27 Generate per-job latency logs.
30 Generate per-job bandwidth logs.
33 Print statistics in a terse, semicolon-delimited format.
36 Print statistics in selected mode AND terse, semicolon-delimited format.
39 Display version information and exit.
41 .BI \-\-terse\-version \fR=\fPversion
42 Set terse version output format (Current version 3, or older version 2).
45 Display usage information and exit.
48 Perform test and validation of internal CPU clock
50 .BI \-\-crctest[\fR=\fPtest]
51 Test the speed of the builtin checksumming functions. If no argument is given,
52 all of them are tested. Or a comma separated list can be passed, in which
53 case the given ones are tested.
55 .BI \-\-cmdhelp \fR=\fPcommand
56 Print help information for \fIcommand\fR. May be `all' for all commands.
58 .BI \-\-enghelp \fR=\fPioengine[,command]
59 List all commands defined by \fIioengine\fR, or print help for \fIcommand\fR defined by \fIioengine\fR.
61 .BI \-\-showcmd \fR=\fPjobfile
62 Convert \fIjobfile\fR to a set of command-line options.
64 .BI \-\-eta \fR=\fPwhen
65 Specifies when real-time ETA estimate should be printed. \fIwhen\fR may
66 be one of `always', `never' or `auto'.
68 .BI \-\-eta\-newline \fR=\fPtime
69 Force an ETA newline for every `time` period passed.
71 .BI \-\-status\-interval \fR=\fPtime
72 Report full output status every `time` period passed.
75 Turn on safety read-only checks, preventing any attempted write.
77 .BI \-\-section \fR=\fPsec
78 Only run section \fIsec\fR from job file. Multiple of these options can be given, adding more sections to run.
80 .BI \-\-alloc\-size \fR=\fPkb
81 Set the internal smalloc pool size to \fIkb\fP kilobytes.
83 .BI \-\-warnings\-fatal
84 All fio parser warnings are fatal, causing fio to exit with an error.
86 .BI \-\-max\-jobs \fR=\fPnr
87 Set the maximum allowed number of jobs (threads/processes) to support.
89 .BI \-\-server \fR=\fPargs
90 Start a backend server, with \fIargs\fP specifying what to listen to. See client/server section.
92 .BI \-\-daemonize \fR=\fPpidfile
93 Background a fio server, writing the pid to the given pid file.
95 .BI \-\-client \fR=\fPhost
96 Instead of running the jobs locally, send and run them on the given host.
98 .BI \-\-idle\-prof \fR=\fPoption
99 Report cpu idleness on a system or percpu basis (\fIoption\fP=system,percpu) or run unit work calibration only (\fIoption\fP=calibrate).
100 .SH "JOB FILE FORMAT"
101 Job files are in `ini' format. They consist of one or more
102 job definitions, which begin with a job name in square brackets and
103 extend to the next job name. The job name can be any ASCII string
104 except `global', which has a special meaning. Following the job name is
105 a sequence of zero or more parameters, one per line, that define the
106 behavior of the job. Any line starting with a `;' or `#' character is
107 considered a comment and ignored.
109 If \fIjobfile\fR is specified as `-', the job file will be read from
112 The global section contains default parameters for jobs specified in the
113 job file. A job is only affected by global sections residing above it,
114 and there may be any number of global sections. Specific job definitions
115 may override any parameter set in global sections.
118 Some parameters may take arguments of a specific type. The types used are:
121 String: a sequence of alphanumeric characters.
124 SI integer: a whole number, possibly containing a suffix denoting the base unit
125 of the value. Accepted suffixes are `k', 'M', 'G', 'T', and 'P', denoting
126 kilo (1024), mega (1024^2), giga (1024^3), tera (1024^4), and peta (1024^5)
127 respectively. If prefixed with '0x', the value is assumed to be base 16
128 (hexadecimal). A suffix may include a trailing 'b', for instance 'kb' is
129 identical to 'k'. You can specify a base 10 value by using 'KiB', 'MiB','GiB',
130 etc. This is useful for disk drives where values are often given in base 10
131 values. Specifying '30GiB' will get you 30*1000^3 bytes.
132 When specifying times the default suffix meaning changes, still denoting the
133 base unit of the value, but accepted suffixes are 'D' (days), 'H' (hours), 'M'
134 (minutes), 'S' Seconds, 'ms' (or msec) milli seconds, 'us' (or 'usec') micro
135 seconds. Time values without a unit specify seconds.
136 The suffixes are not case sensitive.
139 Boolean: a true or false value. `0' denotes false, `1' denotes true.
142 Integer range: a range of integers specified in the format
143 \fIlower\fR:\fIupper\fR or \fIlower\fR\-\fIupper\fR. \fIlower\fR and
144 \fIupper\fR may contain a suffix as described above. If an option allows two
145 sets of ranges, they are separated with a `,' or `/' character. For example:
149 List of floating numbers: A list of floating numbers, separated by
154 May be used to override the job name. On the command line, this parameter
155 has the special purpose of signalling the start of a new job.
157 .BI description \fR=\fPstr
158 Human-readable description of the job. It is printed when the job is run, but
159 otherwise has no special purpose.
161 .BI directory \fR=\fPstr
162 Prefix filenames with this directory. Used to place files in a location other
164 You can specify a number of directories by separating the names with a ':'
165 character. These directories will be assigned equally distributed to job clones
166 creates with \fInumjobs\fR as long as they are using generated filenames.
167 If specific \fIfilename(s)\fR are set fio will use the first listed directory,
168 and thereby matching the \fIfilename\fR semantic which generates a file each
169 clone if not specified, but let all clones use the same if set. See
170 \fIfilename\fR for considerations regarding escaping certain characters on
173 .BI filename \fR=\fPstr
175 normally makes up a file name based on the job name, thread number, and file
176 number. If you want to share files between threads in a job or several jobs,
177 specify a \fIfilename\fR for each of them to override the default.
178 If the I/O engine is file-based, you can specify
179 a number of files by separating the names with a `:' character. `\-' is a
180 reserved name, meaning stdin or stdout, depending on the read/write direction
181 set. On Windows, disk devices are accessed as \\.\PhysicalDrive0 for the first
182 device, \\.\PhysicalDrive1 for the second etc. Note: Windows and FreeBSD
183 prevent write access to areas of the disk containing in-use data
184 (e.g. filesystems). If the wanted filename does need to include a colon, then
185 escape that with a '\' character. For instance, if the filename is
186 "/dev/dsk/foo@3,0:c", then you would use filename="/dev/dsk/foo@3,0\:c".
188 .BI filename_format \fR=\fPstr
189 If sharing multiple files between jobs, it is usually necessary to have
190 fio generate the exact names that you want. By default, fio will name a file
191 based on the default file format specification of
192 \fBjobname.jobnumber.filenumber\fP. With this option, that can be
193 customized. Fio will recognize and replace the following keywords in this
199 The name of the worker thread or process.
202 The incremental number of the worker thread or process.
205 The incremental number of the file for that worker thread or process.
208 To have dependent jobs share a set of files, this option can be set to
209 have fio generate filenames that are shared between the two. For instance,
210 if \fBtestfiles.$filenum\fR is specified, file number 4 for any job will
211 be named \fBtestfiles.4\fR. The default of \fB$jobname.$jobnum.$filenum\fR
212 will be used if no other format specifier is given.
216 .BI lockfile \fR=\fPstr
217 Fio defaults to not locking any files before it does IO to them. If a file or
218 file descriptor is shared, fio can serialize IO to that file to make the end
219 result consistent. This is usual for emulating real workloads that share files.
225 No locking. This is the default.
228 Only one thread or process may do IO at the time, excluding all others.
231 Read-write locking on the file. Many readers may access the file at the same
232 time, but writes get exclusive access.
236 .BI opendir \fR=\fPstr
237 Recursively open any files below directory \fIstr\fR.
239 .BI readwrite \fR=\fPstr "\fR,\fP rw" \fR=\fPstr
240 Type of I/O pattern. Accepted values are:
251 Sequential trim (Linux block devices only).
260 Random trim (Linux block devices only).
263 Mixed sequential reads and writes.
266 Mixed random reads and writes.
269 For mixed I/O, the default split is 50/50. For certain types of io the result
270 may still be skewed a bit, since the speed may be different. It is possible to
271 specify a number of IO's to do before getting a new offset, this is done by
272 appending a `:\fI<nr>\fR to the end of the string given. For a random read, it
273 would look like \fBrw=randread:8\fR for passing in an offset modifier with a
274 value of 8. If the postfix is used with a sequential IO pattern, then the value
275 specified will be added to the generated offset for each IO. For instance,
276 using \fBrw=write:4k\fR will skip 4k for every write. It turns sequential IO
277 into sequential IO with holes. See the \fBrw_sequencer\fR option.
280 .BI rw_sequencer \fR=\fPstr
281 If an offset modifier is given by appending a number to the \fBrw=<str>\fR line,
282 then this option controls how that number modifies the IO offset being
283 generated. Accepted values are:
288 Generate sequential offset
291 Generate the same offset
294 \fBsequential\fR is only useful for random IO, where fio would normally
295 generate a new random offset for every IO. If you append eg 8 to randread, you
296 would get a new random offset for every 8 IO's. The result would be a seek for
297 only every 8 IO's, instead of for every IO. Use \fBrw=randread:8\fR to specify
298 that. As sequential IO is already sequential, setting \fBsequential\fR for that
299 would not result in any differences. \fBidentical\fR behaves in a similar
300 fashion, except it sends the same offset 8 number of times before generating a
305 .BI kb_base \fR=\fPint
306 The base unit for a kilobyte. The defacto base is 2^10, 1024. Storage
307 manufacturers like to use 10^3 or 1000 as a base ten unit instead, for obvious
308 reasons. Allowed values are 1024 or 1000, with 1024 being the default.
310 .BI unified_rw_reporting \fR=\fPbool
311 Fio normally reports statistics on a per data direction basis, meaning that
312 read, write, and trim are accounted and reported separately. If this option is
313 set, the fio will sum the results and report them as "mixed" instead.
315 .BI randrepeat \fR=\fPbool
316 Seed the random number generator used for random I/O patterns in a predictable
317 way so the pattern is repeatable across runs. Default: true.
319 .BI allrandrepeat \fR=\fPbool
320 Seed all random number generators in a predictable way so results are
321 repeatable across runs. Default: false.
323 .BI randseed \fR=\fPint
324 Seed the random number generators based on this seed value, to be able to
325 control what sequence of output is being generated. If not set, the random
326 sequence depends on the \fBrandrepeat\fR setting.
328 .BI use_os_rand \fR=\fPbool
329 Fio can either use the random generator supplied by the OS to generator random
330 offsets, or it can use it's own internal generator (based on Tausworthe).
331 Default is to use the internal generator, which is often of better quality and
332 faster. Default: false.
334 .BI fallocate \fR=\fPstr
335 Whether pre-allocation is performed when laying down files. Accepted values
341 Do not pre-allocate space.
344 Pre-allocate via \fBposix_fallocate\fR\|(3).
347 Pre-allocate via \fBfallocate\fR\|(2) with FALLOC_FL_KEEP_SIZE set.
350 Backward-compatible alias for 'none'.
353 Backward-compatible alias for 'posix'.
356 May not be available on all supported platforms. 'keep' is only
357 available on Linux. If using ZFS on Solaris this must be set to 'none'
358 because ZFS doesn't support it. Default: 'posix'.
361 .BI fadvise_hint \fR=\fPbool
362 Use of \fBposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
363 are likely to be issued. Default: true.
366 Total size of I/O for this job. \fBfio\fR will run until this many bytes have
367 been transferred, unless limited by other options (\fBruntime\fR, for instance).
368 Unless \fBnrfiles\fR and \fBfilesize\fR options are given, this amount will be
369 divided between the available files for the job. If not set, fio will use the
370 full size of the given files or devices. If the files do not exist, size
371 must be given. It is also possible to give size as a percentage between 1 and
372 100. If size=20% is given, fio will use 20% of the full size of the given files
375 .BI fill_device \fR=\fPbool "\fR,\fB fill_fs" \fR=\fPbool
376 Sets size to something really large and waits for ENOSPC (no space left on
377 device) as the terminating condition. Only makes sense with sequential write.
378 For a read workload, the mount point will be filled first then IO started on
379 the result. This option doesn't make sense if operating on a raw device node,
380 since the size of that is already known by the file system. Additionally,
381 writing beyond end-of-device will not return ENOSPC there.
383 .BI filesize \fR=\fPirange
384 Individual file sizes. May be a range, in which case \fBfio\fR will select sizes
385 for files at random within the given range, limited to \fBsize\fR in total (if
386 that is given). If \fBfilesize\fR is not specified, each created file is the
389 .BI blocksize \fR=\fPint[,int] "\fR,\fB bs" \fR=\fPint[,int]
390 Block size for I/O units. Default: 4k. Values for reads, writes, and trims
391 can be specified separately in the format \fIread\fR,\fIwrite\fR,\fItrim\fR
392 either of which may be empty to leave that value at its default. If a trailing
393 comma isn't given, the remainder will inherit the last value set.
395 .BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange]
396 Specify a range of I/O block sizes. The issued I/O unit will always be a
397 multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies
398 to both reads and writes if only one range is given, but can be specified
399 separately with a comma separating the values. Example: bsrange=1k-4k,2k-8k.
400 Also (see \fBblocksize\fR).
402 .BI bssplit \fR=\fPstr
403 This option allows even finer grained control of the block sizes issued,
404 not just even splits between them. With this option, you can weight various
405 block sizes for exact control of the issued IO for a job that has mixed
406 block sizes. The format of the option is bssplit=blocksize/percentage,
407 optionally adding as many definitions as needed separated by a colon.
408 Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k
409 blocks and 40% 32k blocks. \fBbssplit\fR also supports giving separate
410 splits to reads and writes. The format is identical to what the
411 \fBbs\fR option accepts, the read and write parts are separated with a
414 .B blocksize_unaligned\fR,\fP bs_unaligned
415 If set, any size in \fBblocksize_range\fR may be used. This typically won't
416 work with direct I/O, as that normally requires sector alignment.
418 .BI blockalign \fR=\fPint[,int] "\fR,\fB ba" \fR=\fPint[,int]
419 At what boundary to align random IO offsets. Defaults to the same as 'blocksize'
420 the minimum blocksize given. Minimum alignment is typically 512b
421 for using direct IO, though it usually depends on the hardware block size.
422 This option is mutually exclusive with using a random map for files, so it
423 will turn off that option.
425 .BI bs_is_seq_rand \fR=\fPbool
426 If this option is set, fio will use the normal read,write blocksize settings as
427 sequential,random instead. Any random read or write will use the WRITE
428 blocksize settings, and any sequential read or write will use the READ
432 Initialise buffers with all zeros. Default: fill buffers with random data.
435 If this option is given, fio will refill the IO buffers on every submit. The
436 default is to only fill it at init time and reuse that data. Only makes sense
437 if zero_buffers isn't specified, naturally. If data verification is enabled,
438 refill_buffers is also automatically enabled.
440 .BI scramble_buffers \fR=\fPbool
441 If \fBrefill_buffers\fR is too costly and the target is using data
442 deduplication, then setting this option will slightly modify the IO buffer
443 contents to defeat normal de-dupe attempts. This is not enough to defeat
444 more clever block compression attempts, but it will stop naive dedupe
445 of blocks. Default: true.
447 .BI buffer_compress_percentage \fR=\fPint
448 If this is set, then fio will attempt to provide IO buffer content (on WRITEs)
449 that compress to the specified level. Fio does this by providing a mix of
450 random data and zeroes. Note that this is per block size unit, for file/disk
451 wide compression level that matches this setting, you'll also want to set
452 \fBrefill_buffers\fR.
454 .BI buffer_compress_chunk \fR=\fPint
455 See \fBbuffer_compress_percentage\fR. This setting allows fio to manage how
456 big the ranges of random data and zeroed data is. Without this set, fio will
457 provide \fBbuffer_compress_percentage\fR of blocksize random data, followed by
458 the remaining zeroed. With this set to some chunk size smaller than the block
459 size, fio can alternate random and zeroed data throughout the IO buffer.
461 .BI buffer_pattern \fR=\fPstr
462 If set, fio will fill the io buffers with this pattern. If not set, the contents
463 of io buffers is defined by the other options related to buffer contents. The
464 setting can be any pattern of bytes, and can be prefixed with 0x for hex
467 .BI nrfiles \fR=\fPint
468 Number of files to use for this job. Default: 1.
470 .BI openfiles \fR=\fPint
471 Number of files to keep open at the same time. Default: \fBnrfiles\fR.
473 .BI file_service_type \fR=\fPstr
474 Defines how files to service are selected. The following types are defined:
479 Choose a file at random.
482 Round robin over open files (default).
485 Do each file in the set sequentially.
488 The number of I/Os to issue before switching a new file can be specified by
489 appending `:\fIint\fR' to the service type.
492 .BI ioengine \fR=\fPstr
493 Defines how the job issues I/O. The following types are defined:
498 Basic \fBread\fR\|(2) or \fBwrite\fR\|(2) I/O. \fBfseek\fR\|(2) is used to
499 position the I/O location.
502 Basic \fBpread\fR\|(2) or \fBpwrite\fR\|(2) I/O.
505 Basic \fBreadv\fR\|(2) or \fBwritev\fR\|(2) I/O. Will emulate queuing by
506 coalescing adjacent IOs into a single submission.
509 Basic \fBpreadv\fR\|(2) or \fBpwritev\fR\|(2) I/O.
512 Linux native asynchronous I/O. This ioengine defines engine specific options.
515 POSIX asynchronous I/O using \fBaio_read\fR\|(3) and \fBaio_write\fR\|(3).
518 Solaris native asynchronous I/O.
521 Windows native asynchronous I/O.
524 File is memory mapped with \fBmmap\fR\|(2) and data copied using
528 \fBsplice\fR\|(2) is used to transfer the data and \fBvmsplice\fR\|(2) to
529 transfer data from user-space to the kernel.
532 Use the syslet system calls to make regular read/write asynchronous.
535 SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if
536 the target is an sg character device, we use \fBread\fR\|(2) and
537 \fBwrite\fR\|(2) for asynchronous I/O.
540 Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR
541 itself and for debugging and testing purposes.
544 Transfer over the network. The protocol to be used can be defined with the
545 \fBprotocol\fR parameter. Depending on the protocol, \fBfilename\fR,
546 \fBhostname\fR, \fBport\fR, or \fBlisten\fR must be specified.
547 This ioengine defines engine specific options.
550 Like \fBnet\fR, but uses \fBsplice\fR\|(2) and \fBvmsplice\fR\|(2) to map data
551 and send/receive. This ioengine defines engine specific options.
554 Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and
555 \fBcpucycles\fR parameters.
558 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface
559 approach to asynchronous I/O.
561 See <http://www.xmailserver.org/guasi\-lib.html>.
564 The RDMA I/O engine supports both RDMA memory semantics (RDMA_WRITE/RDMA_READ)
565 and channel semantics (Send/Recv) for the InfiniBand, RoCE and iWARP protocols.
568 Loads an external I/O engine object file. Append the engine filename as
572 IO engine that does regular linux native fallocate call to simulate data
573 transfer as fio ioengine
575 DDIR_READ does fallocate(,mode = FALLOC_FL_KEEP_SIZE,)
577 DIR_WRITE does fallocate(,mode = 0)
579 DDIR_TRIM does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE)
582 IO engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate defragment activity
583 request to DDIR_WRITE event
586 IO engine supporting direct access to Ceph Rados Block Devices (RBD) via librbd
587 without the need to use the kernel rbd driver. This ioengine defines engine specific
593 .BI iodepth \fR=\fPint
594 Number of I/O units to keep in flight against the file. Note that increasing
595 iodepth beyond 1 will not affect synchronous ioengines (except for small
596 degress when verify_async is in use). Even async engines my impose OS
597 restrictions causing the desired depth not to be achieved. This may happen on
598 Linux when using libaio and not setting \fBdirect\fR=1, since buffered IO is
599 not async on that OS. Keep an eye on the IO depth distribution in the
600 fio output to verify that the achieved depth is as expected. Default: 1.
602 .BI iodepth_batch \fR=\fPint
603 Number of I/Os to submit at once. Default: \fBiodepth\fR.
605 .BI iodepth_batch_complete \fR=\fPint
606 This defines how many pieces of IO to retrieve at once. It defaults to 1 which
607 means that we'll ask for a minimum of 1 IO in the retrieval process from the
608 kernel. The IO retrieval will go on until we hit the limit set by
609 \fBiodepth_low\fR. If this variable is set to 0, then fio will always check for
610 completed events before queuing more IO. This helps reduce IO latency, at the
611 cost of more retrieval system calls.
613 .BI iodepth_low \fR=\fPint
614 Low watermark indicating when to start filling the queue again. Default:
617 .BI direct \fR=\fPbool
618 If true, use non-buffered I/O (usually O_DIRECT). Default: false.
620 .BI atomic \fR=\fPbool
621 If value is true, attempt to use atomic direct IO. Atomic writes are guaranteed
622 to be stable once acknowledged by the operating system. Only Linux supports
625 .BI buffered \fR=\fPbool
626 If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter.
629 .BI offset \fR=\fPint
630 Offset in the file to start I/O. Data before the offset will not be touched.
632 .BI offset_increment \fR=\fPint
633 If this is provided, then the real offset becomes the
634 offset + offset_increment * thread_number, where the thread number is a counter
635 that starts at 0 and is incremented for each job. This option is useful if
636 there are several jobs which are intended to operate on a file in parallel in
637 disjoint segments, with even spacing between the starting points.
639 .BI number_ios \fR=\fPint
640 Fio will normally perform IOs until it has exhausted the size of the region
641 set by \fBsize\fR, or if it exhaust the allocated time (or hits an error
642 condition). With this setting, the range/size can be set independently of
643 the number of IOs to perform. When fio reaches this number, it will exit
644 normally and report status.
647 How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If
648 0, don't sync. Default: 0.
650 .BI fdatasync \fR=\fPint
651 Like \fBfsync\fR, but uses \fBfdatasync\fR\|(2) instead to only sync the
652 data parts of the file. Default: 0.
654 .BI write_barrier \fR=\fPint
655 Make every Nth write a barrier write.
657 .BI sync_file_range \fR=\fPstr:int
658 Use \fBsync_file_range\fR\|(2) for every \fRval\fP number of write operations. Fio will
659 track range of writes that have happened since the last \fBsync_file_range\fR\|(2) call.
660 \fRstr\fP can currently be one or more of:
664 SYNC_FILE_RANGE_WAIT_BEFORE
667 SYNC_FILE_RANGE_WRITE
670 SYNC_FILE_RANGE_WRITE
674 So if you do sync_file_range=wait_before,write:8, fio would use
675 \fBSYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE\fP for every 8 writes.
676 Also see the \fBsync_file_range\fR\|(2) man page. This option is Linux specific.
678 .BI overwrite \fR=\fPbool
679 If writing, setup the file first and do overwrites. Default: false.
681 .BI end_fsync \fR=\fPbool
682 Sync file contents when a write stage has completed. Default: false.
684 .BI fsync_on_close \fR=\fPbool
685 If true, sync file contents on close. This differs from \fBend_fsync\fR in that
686 it will happen on every close, not just at the end of the job. Default: false.
688 .BI rwmixread \fR=\fPint
689 Percentage of a mixed workload that should be reads. Default: 50.
691 .BI rwmixwrite \fR=\fPint
692 Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and
693 \fBrwmixwrite\fR are given and do not sum to 100%, the latter of the two
694 overrides the first. This may interfere with a given rate setting, if fio is
695 asked to limit reads or writes to a certain rate. If that is the case, then
696 the distribution may be skewed. Default: 50.
698 .BI random_distribution \fR=\fPstr:float
699 By default, fio will use a completely uniform random distribution when asked
700 to perform random IO. Sometimes it is useful to skew the distribution in
701 specific ways, ensuring that some parts of the data is more hot than others.
702 Fio includes the following distribution models:
706 Uniform random distribution
716 When using a zipf or pareto distribution, an input value is also needed to
717 define the access pattern. For zipf, this is the zipf theta. For pareto,
718 it's the pareto power. Fio includes a test program, genzipf, that can be
719 used visualize what the given input values will yield in terms of hit rates.
720 If you wanted to use zipf with a theta of 1.2, you would use
721 random_distribution=zipf:1.2 as the option. If a non-uniform model is used,
722 fio will disable use of the random map.
724 .BI percentage_random \fR=\fPint
725 For a random workload, set how big a percentage should be random. This defaults
726 to 100%, in which case the workload is fully random. It can be set from
727 anywhere from 0 to 100. Setting it to 0 would make the workload fully
728 sequential. It is possible to set different values for reads, writes, and
729 trim. To do so, simply use a comma separated list. See \fBblocksize\fR.
732 Normally \fBfio\fR will cover every block of the file when doing random I/O. If
733 this parameter is given, a new offset will be chosen without looking at past
734 I/O history. This parameter is mutually exclusive with \fBverify\fR.
736 .BI softrandommap \fR=\fPbool
737 See \fBnorandommap\fR. If fio runs with the random block map enabled and it
738 fails to allocate the map, if this option is set it will continue without a
739 random block map. As coverage will not be as complete as with random maps, this
740 option is disabled by default.
742 .BI random_generator \fR=\fPstr
743 Fio supports the following engines for generating IO offsets for random IO:
747 Strong 2^88 cycle random number generator
750 Linear feedback shift register generator
754 Tausworthe is a strong random number generator, but it requires tracking on the
755 side if we want to ensure that blocks are only read or written once. LFSR
756 guarantees that we never generate the same offset twice, and it's also less
757 computationally expensive. It's not a true random generator, however, though
758 for IO purposes it's typically good enough. LFSR only works with single block
759 sizes, not with workloads that use multiple block sizes. If used with such a
760 workload, fio may read or write some blocks multiple times.
763 Run job with given nice value. See \fBnice\fR\|(2).
766 Set I/O priority value of this job between 0 (highest) and 7 (lowest). See
769 .BI prioclass \fR=\fPint
770 Set I/O priority class. See \fBionice\fR\|(1).
772 .BI thinktime \fR=\fPint
773 Stall job for given number of microseconds between issuing I/Os.
775 .BI thinktime_spin \fR=\fPint
776 Pretend to spend CPU time for given number of microseconds, sleeping the rest
777 of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set.
779 .BI thinktime_blocks \fR=\fPint
780 Only valid if thinktime is set - control how many blocks to issue, before
781 waiting \fBthinktime\fR microseconds. If not set, defaults to 1 which will
782 make fio wait \fBthinktime\fR microseconds after every block. This
783 effectively makes any queue depth setting redundant, since no more than 1 IO
784 will be queued before we have to complete it and do our thinktime. In other
785 words, this setting effectively caps the queue depth if the latter is larger.
789 Cap bandwidth used by this job. The number is in bytes/sec, the normal postfix
790 rules apply. You can use \fBrate\fR=500k to limit reads and writes to 500k each,
791 or you can specify read and writes separately. Using \fBrate\fR=1m,500k would
792 limit reads to 1MB/sec and writes to 500KB/sec. Capping only reads or writes
793 can be done with \fBrate\fR=,500k or \fBrate\fR=500k,. The former will only
794 limit writes (to 500KB/sec), the latter will only limit reads.
796 .BI ratemin \fR=\fPint
797 Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth.
798 Failing to meet this requirement will cause the job to exit. The same format
799 as \fBrate\fR is used for read vs write separation.
801 .BI rate_iops \fR=\fPint
802 Cap the bandwidth to this number of IOPS. Basically the same as rate, just
803 specified independently of bandwidth. The same format as \fBrate\fR is used for
804 read vs write separation. If \fBblocksize\fR is a range, the smallest block
805 size is used as the metric.
807 .BI rate_iops_min \fR=\fPint
808 If this rate of I/O is not met, the job will exit. The same format as \fBrate\fR
809 is used for read vs write separation.
811 .BI ratecycle \fR=\fPint
812 Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of
813 milliseconds. Default: 1000ms.
815 .BI latency_target \fR=\fPint
816 If set, fio will attempt to find the max performance point that the given
817 workload will run at while maintaining a latency below this target. The
818 values is given in microseconds. See \fBlatency_window\fR and
819 \fBlatency_percentile\fR.
821 .BI latency_window \fR=\fPint
822 Used with \fBlatency_target\fR to specify the sample window that the job
823 is run at varying queue depths to test the performance. The value is given
826 .BI latency_percentile \fR=\fPfloat
827 The percentage of IOs that must fall within the criteria specified by
828 \fBlatency_target\fR and \fBlatency_window\fR. If not set, this defaults
829 to 100.0, meaning that all IOs must be equal or below to the value set
830 by \fBlatency_target\fR.
832 .BI max_latency \fR=\fPint
833 If set, fio will exit the job if it exceeds this maximum latency. It will exit
836 .BI cpumask \fR=\fPint
837 Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job
838 may run on. See \fBsched_setaffinity\fR\|(2).
840 .BI cpus_allowed \fR=\fPstr
841 Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers.
843 .BI cpus_allowed_policy \fR=\fPstr
844 Set the policy of how fio distributes the CPUs specified by \fBcpus_allowed\fR
845 or \fBcpumask\fR. Two policies are supported:
850 All jobs will share the CPU set specified.
853 Each job will get a unique CPU from the CPU set.
856 \fBshared\fR is the default behaviour, if the option isn't specified. If
857 \fBsplit\fR is specified, then fio will assign one cpu per job. If not enough
858 CPUs are given for the jobs listed, then fio will roundrobin the CPUs in
863 .BI numa_cpu_nodes \fR=\fPstr
864 Set this job running on specified NUMA nodes' CPUs. The arguments allow
865 comma delimited list of cpu numbers, A-B ranges, or 'all'.
867 .BI numa_mem_policy \fR=\fPstr
868 Set this job's memory policy and corresponding NUMA nodes. Format of
872 .B <mode>[:<nodelist>]
875 is one of the following memory policy:
877 .B default, prefer, bind, interleave, local
880 For \fBdefault\fR and \fBlocal\fR memory policy, no \fBnodelist\fR is
881 needed to be specified. For \fBprefer\fR, only one node is
882 allowed. For \fBbind\fR and \fBinterleave\fR, \fBnodelist\fR allows
883 comma delimited list of numbers, A-B ranges, or 'all'.
885 .BI startdelay \fR=\fPirange
886 Delay start of job for the specified number of seconds. Supports all time
887 suffixes to allow specification of hours, minutes, seconds and
888 milliseconds - seconds are the default if a unit is ommited.
889 Can be given as a range which causes each thread to choose randomly out of the
892 .BI runtime \fR=\fPint
893 Terminate processing after the specified number of seconds.
896 If given, run for the specified \fBruntime\fR duration even if the files are
897 completely read or written. The same workload will be repeated as many times
898 as \fBruntime\fR allows.
900 .BI ramp_time \fR=\fPint
901 If set, fio will run the specified workload for this amount of time before
902 logging any performance numbers. Useful for letting performance settle before
903 logging results, thus minimizing the runtime required for stable results. Note
904 that the \fBramp_time\fR is considered lead in time for a job, thus it will
905 increase the total runtime if a special timeout or runtime is specified.
907 .BI invalidate \fR=\fPbool
908 Invalidate buffer-cache for the file prior to starting I/O. Default: true.
911 Use synchronous I/O for buffered writes. For the majority of I/O engines,
912 this means using O_SYNC. Default: false.
914 .BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr
915 Allocation method for I/O unit buffer. Allowed values are:
920 Allocate memory with \fBmalloc\fR\|(3).
923 Use shared memory buffers allocated through \fBshmget\fR\|(2).
926 Same as \fBshm\fR, but use huge pages as backing.
929 Use \fBmmap\fR\|(2) for allocation. Uses anonymous memory unless a filename
930 is given after the option in the format `:\fIfile\fR'.
933 Same as \fBmmap\fR, but use huge files as backing.
936 The amount of memory allocated is the maximum allowed \fBblocksize\fR for the
937 job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work,
938 the system must have free huge pages allocated. \fBmmaphuge\fR also needs to
939 have hugetlbfs mounted, and \fIfile\fR must point there. At least on Linux,
940 huge pages must be manually allocated. See \fB/proc/sys/vm/nr_hugehages\fR
941 and the documentation for that. Normally you just need to echo an appropriate
942 number, eg echoing 8 will ensure that the OS has 8 huge pages ready for
946 .BI iomem_align \fR=\fPint "\fR,\fP mem_align" \fR=\fPint
947 This indicates the memory alignment of the IO memory buffers. Note that the
948 given alignment is applied to the first IO unit buffer, if using \fBiodepth\fR
949 the alignment of the following buffers are given by the \fBbs\fR used. In
950 other words, if using a \fBbs\fR that is a multiple of the page sized in the
951 system, all buffers will be aligned to this value. If using a \fBbs\fR that
952 is not page aligned, the alignment of subsequent IO memory buffers is the
953 sum of the \fBiomem_align\fR and \fBbs\fR used.
955 .BI hugepage\-size \fR=\fPint
956 Defines the size of a huge page. Must be at least equal to the system setting.
957 Should be a multiple of 1MB. Default: 4MB.
960 Terminate all jobs when one finishes. Default: wait for each job to finish.
962 .BI bwavgtime \fR=\fPint
963 Average bandwidth calculations over the given time in milliseconds. Default:
966 .BI iopsavgtime \fR=\fPint
967 Average IOPS calculations over the given time in milliseconds. Default:
970 .BI create_serialize \fR=\fPbool
971 If true, serialize file creation for the jobs. Default: true.
973 .BI create_fsync \fR=\fPbool
974 \fBfsync\fR\|(2) data file after creation. Default: true.
976 .BI create_on_open \fR=\fPbool
977 If true, the files are not created until they are opened for IO by the job.
979 .BI create_only \fR=\fPbool
980 If true, fio will only run the setup phase of the job. If files need to be
981 laid out or updated on disk, only that will be done. The actual job contents
984 .BI pre_read \fR=\fPbool
985 If this is given, files will be pre-read into memory before starting the given
986 IO operation. This will also clear the \fR \fBinvalidate\fR flag, since it is
987 pointless to pre-read and then drop the cache. This will only work for IO
988 engines that are seekable, since they allow you to read the same data
989 multiple times. Thus it will not work on eg network or splice IO.
991 .BI unlink \fR=\fPbool
992 Unlink job files when done. Default: false.
995 Specifies the number of iterations (runs of the same workload) of this job.
998 .BI verify_only \fR=\fPbool
999 Do not perform the specified workload, only verify data still matches previous
1000 invocation of this workload. This option allows one to check data multiple
1001 times at a later date without overwriting it. This option makes sense only for
1002 workloads that write data, and does not support workloads with the
1003 \fBtime_based\fR option set.
1005 .BI do_verify \fR=\fPbool
1006 Run the verify phase after a write phase. Only valid if \fBverify\fR is set.
1009 .BI verify \fR=\fPstr
1010 Method of verifying file contents after each iteration of the job. Allowed
1015 .B md5 crc16 crc32 crc32c crc32c-intel crc64 crc7 sha256 sha512 sha1 xxhash
1016 Store appropriate checksum in the header of each block. crc32c-intel is
1017 hardware accelerated SSE4.2 driven, falls back to regular crc32c if
1018 not supported by the system.
1021 Write extra information about each I/O (timestamp, block number, etc.). The
1022 block number is verified. See \fBverify_pattern\fR as well.
1025 Pretend to verify. Used for testing internals.
1028 This option can be used for repeated burn-in tests of a system to make sure
1029 that the written data is also correctly read back. If the data direction given
1030 is a read or random read, fio will assume that it should verify a previously
1031 written file. If the data direction includes any form of write, the verify will
1032 be of the newly written data.
1035 .BI verifysort \fR=\fPbool
1036 If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to
1037 read them back in a sorted manner. Default: true.
1039 .BI verifysort_nr \fR=\fPint
1040 Pre-load and sort verify blocks for a read workload.
1042 .BI verify_offset \fR=\fPint
1043 Swap the verification header with data somewhere else in the block before
1044 writing. It is swapped back before verifying.
1046 .BI verify_interval \fR=\fPint
1047 Write the verification header for this number of bytes, which should divide
1048 \fBblocksize\fR. Default: \fBblocksize\fR.
1050 .BI verify_pattern \fR=\fPstr
1051 If set, fio will fill the io buffers with this pattern. Fio defaults to filling
1052 with totally random bytes, but sometimes it's interesting to fill with a known
1053 pattern for io verification purposes. Depending on the width of the pattern,
1054 fio will fill 1/2/3/4 bytes of the buffer at the time(it can be either a
1055 decimal or a hex number). The verify_pattern if larger than a 32-bit quantity
1056 has to be a hex number that starts with either "0x" or "0X". Use with
1059 .BI verify_fatal \fR=\fPbool
1060 If true, exit the job on the first observed verification failure. Default:
1063 .BI verify_dump \fR=\fPbool
1064 If set, dump the contents of both the original data block and the data block we
1065 read off disk to files. This allows later analysis to inspect just what kind of
1066 data corruption occurred. Off by default.
1068 .BI verify_async \fR=\fPint
1069 Fio will normally verify IO inline from the submitting thread. This option
1070 takes an integer describing how many async offload threads to create for IO
1071 verification instead, causing fio to offload the duty of verifying IO contents
1072 to one or more separate threads. If using this offload option, even sync IO
1073 engines can benefit from using an \fBiodepth\fR setting higher than 1, as it
1074 allows them to have IO in flight while verifies are running.
1076 .BI verify_async_cpus \fR=\fPstr
1077 Tell fio to set the given CPU affinity on the async IO verification threads.
1078 See \fBcpus_allowed\fP for the format used.
1080 .BI verify_backlog \fR=\fPint
1081 Fio will normally verify the written contents of a job that utilizes verify
1082 once that job has completed. In other words, everything is written then
1083 everything is read back and verified. You may want to verify continually
1084 instead for a variety of reasons. Fio stores the meta data associated with an
1085 IO block in memory, so for large verify workloads, quite a bit of memory would
1086 be used up holding this meta data. If this option is enabled, fio will write
1087 only N blocks before verifying these blocks.
1089 .BI verify_backlog_batch \fR=\fPint
1090 Control how many blocks fio will verify if verify_backlog is set. If not set,
1091 will default to the value of \fBverify_backlog\fR (meaning the entire queue is
1092 read back and verified). If \fBverify_backlog_batch\fR is less than
1093 \fBverify_backlog\fR then not all blocks will be verified, if
1094 \fBverify_backlog_batch\fR is larger than \fBverify_backlog\fR, some blocks
1095 will be verified more than once.
1097 .BI trim_percentage \fR=\fPint
1098 Number of verify blocks to discard/trim.
1100 .BI trim_verify_zero \fR=\fPbool
1101 Verify that trim/discarded blocks are returned as zeroes.
1103 .BI trim_backlog \fR=\fPint
1104 Trim after this number of blocks are written.
1106 .BI trim_backlog_batch \fR=\fPint
1107 Trim this number of IO blocks.
1109 .BI experimental_verify \fR=\fPbool
1110 Enable experimental verification.
1112 .B stonewall "\fR,\fP wait_for_previous"
1113 Wait for preceding jobs in the job file to exit before starting this one.
1114 \fBstonewall\fR implies \fBnew_group\fR.
1117 Start a new reporting group. If not given, all jobs in a file will be part
1118 of the same reporting group, unless separated by a stonewall.
1120 .BI numjobs \fR=\fPint
1121 Number of clones (processes/threads performing the same workload) of this job.
1125 If set, display per-group reports instead of per-job when \fBnumjobs\fR is
1129 Use threads created with \fBpthread_create\fR\|(3) instead of processes created
1130 with \fBfork\fR\|(2).
1132 .BI zonesize \fR=\fPint
1133 Divide file into zones of the specified size in bytes. See \fBzoneskip\fR.
1135 .BI zonerange \fR=\fPint
1136 Give size of an IO zone. See \fBzoneskip\fR.
1138 .BI zoneskip \fR=\fPint
1139 Skip the specified number of bytes when \fBzonesize\fR bytes of data have been
1142 .BI write_iolog \fR=\fPstr
1143 Write the issued I/O patterns to the specified file. Specify a separate file
1144 for each job, otherwise the iologs will be interspersed and the file may be
1147 .BI read_iolog \fR=\fPstr
1148 Replay the I/O patterns contained in the specified file generated by
1149 \fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file.
1151 .BI replay_no_stall \fR=\fPint
1152 While replaying I/O patterns using \fBread_iolog\fR the default behavior
1153 attempts to respect timing information between I/Os. Enabling
1154 \fBreplay_no_stall\fR causes I/Os to be replayed as fast as possible while
1155 still respecting ordering.
1157 .BI replay_redirect \fR=\fPstr
1158 While replaying I/O patterns using \fBread_iolog\fR the default behavior
1159 is to replay the IOPS onto the major/minor device that each IOP was recorded
1160 from. Setting \fBreplay_redirect\fR causes all IOPS to be replayed onto the
1161 single specified device regardless of the device it was recorded from.
1163 .BI write_bw_log \fR=\fPstr
1164 If given, write a bandwidth log of the jobs in this job file. Can be used to
1165 store data of the bandwidth of the jobs in their lifetime. The included
1166 fio_generate_plots script uses gnuplot to turn these text files into nice
1167 graphs. See \fBwrite_lat_log\fR for behaviour of given filename. For this
1168 option, the postfix is _bw.log.
1170 .BI write_lat_log \fR=\fPstr
1171 Same as \fBwrite_bw_log\fR, but writes I/O completion latencies. If no
1172 filename is given with this option, the default filename of "jobname_type.log"
1173 is used. Even if the filename is given, fio will still append the type of log.
1175 .BI write_iops_log \fR=\fPstr
1176 Same as \fBwrite_bw_log\fR, but writes IOPS. If no filename is given with this
1177 option, the default filename of "jobname_type.log" is used. Even if the
1178 filename is given, fio will still append the type of log.
1180 .BI log_avg_msec \fR=\fPint
1181 By default, fio will log an entry in the iops, latency, or bw log for every
1182 IO that completes. When writing to the disk log, that can quickly grow to a
1183 very large size. Setting this option makes fio average the each log entry
1184 over the specified period of time, reducing the resolution of the log.
1187 .BI disable_lat \fR=\fPbool
1188 Disable measurements of total latency numbers. Useful only for cutting
1189 back the number of calls to \fBgettimeofday\fR\|(2), as that does impact performance at
1190 really high IOPS rates. Note that to really get rid of a large amount of these
1191 calls, this option must be used with disable_slat and disable_bw as well.
1193 .BI disable_clat \fR=\fPbool
1194 Disable measurements of completion latency numbers. See \fBdisable_lat\fR.
1196 .BI disable_slat \fR=\fPbool
1197 Disable measurements of submission latency numbers. See \fBdisable_lat\fR.
1199 .BI disable_bw_measurement \fR=\fPbool
1200 Disable measurements of throughput/bandwidth numbers. See \fBdisable_lat\fR.
1202 .BI lockmem \fR=\fPint
1203 Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to
1204 simulate a smaller amount of memory. The amount specified is per worker.
1206 .BI exec_prerun \fR=\fPstr
1207 Before running the job, execute the specified command with \fBsystem\fR\|(3).
1209 Output is redirected in a file called \fBjobname.prerun.txt\fR
1212 .BI exec_postrun \fR=\fPstr
1213 Same as \fBexec_prerun\fR, but the command is executed after the job completes.
1215 Output is redirected in a file called \fBjobname.postrun.txt\fR
1218 .BI ioscheduler \fR=\fPstr
1219 Attempt to switch the device hosting the file to the specified I/O scheduler.
1221 .BI cpuload \fR=\fPint
1222 If the job is a CPU cycle-eater, attempt to use the specified percentage of
1225 .BI cpuchunks \fR=\fPint
1226 If the job is a CPU cycle-eater, split the load into cycles of the
1227 given time in milliseconds.
1229 .BI disk_util \fR=\fPbool
1230 Generate disk utilization statistics if the platform supports it. Default: true.
1232 .BI clocksource \fR=\fPstr
1233 Use the given clocksource as the base of timing. The supported options are:
1237 \fBgettimeofday\fR\|(2)
1240 \fBclock_gettime\fR\|(2)
1243 Internal CPU clock source
1247 \fBcpu\fR is the preferred clocksource if it is reliable, as it is very fast
1248 (and fio is heavy on time calls). Fio will automatically use this clocksource
1249 if it's supported and considered reliable on the system it is running on,
1250 unless another clocksource is specifically set. For x86/x86-64 CPUs, this
1251 means supporting TSC Invariant.
1253 .BI gtod_reduce \fR=\fPbool
1254 Enable all of the \fBgettimeofday\fR\|(2) reducing options (disable_clat, disable_slat,
1255 disable_bw) plus reduce precision of the timeout somewhat to really shrink the
1256 \fBgettimeofday\fR\|(2) call count. With this option enabled, we only do about 0.4% of
1257 the gtod() calls we would have done if all time keeping was enabled.
1259 .BI gtod_cpu \fR=\fPint
1260 Sometimes it's cheaper to dedicate a single thread of execution to just getting
1261 the current time. Fio (and databases, for instance) are very intensive on
1262 \fBgettimeofday\fR\|(2) calls. With this option, you can set one CPU aside for doing
1263 nothing but logging current time to a shared memory location. Then the other
1264 threads/processes that run IO workloads need only copy that segment, instead of
1265 entering the kernel with a \fBgettimeofday\fR\|(2) call. The CPU set aside for doing
1266 these time calls will be excluded from other uses. Fio will manually clear it
1267 from the CPU mask of other jobs.
1269 .BI ignore_error \fR=\fPstr
1270 Sometimes you want to ignore some errors during test in that case you can specify
1271 error list for each error type.
1273 ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST
1275 errors for given error type is separated with ':'.
1276 Error may be symbol ('ENOSPC', 'ENOMEM') or an integer.
1278 Example: ignore_error=EAGAIN,ENOSPC:122 .
1280 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from WRITE.
1282 .BI error_dump \fR=\fPbool
1283 If set dump every error even if it is non fatal, true by default. If disabled
1284 only fatal error will be dumped
1286 .BI profile \fR=\fPstr
1287 Select a specific builtin performance test.
1289 .BI cgroup \fR=\fPstr
1290 Add job to this control group. If it doesn't exist, it will be created.
1291 The system must have a mounted cgroup blkio mount point for this to work. If
1292 your system doesn't have it mounted, you can do so with:
1294 # mount \-t cgroup \-o blkio none /cgroup
1296 .BI cgroup_weight \fR=\fPint
1297 Set the weight of the cgroup to this value. See the documentation that comes
1298 with the kernel, allowed values are in the range of 100..1000.
1300 .BI cgroup_nodelete \fR=\fPbool
1301 Normally fio will delete the cgroups it has created after the job completion.
1302 To override this behavior and to leave cgroups around after the job completion,
1303 set cgroup_nodelete=1. This can be useful if one wants to inspect various
1304 cgroup files after job completion. Default: false
1307 Instead of running as the invoking user, set the user ID to this value before
1308 the thread/process does any work.
1311 Set group ID, see \fBuid\fR.
1313 .BI unit_base \fR=\fPint
1314 Base unit for reporting. Allowed values are:
1318 Use auto-detection (default).
1328 .BI flow_id \fR=\fPint
1329 The ID of the flow. If not specified, it defaults to being a global flow. See
1333 Weight in token-based flow control. If this value is used, then there is a
1334 \fBflow counter\fR which is used to regulate the proportion of activity between
1335 two or more jobs. fio attempts to keep this flow counter near zero. The
1336 \fBflow\fR parameter stands for how much should be added or subtracted to the
1337 flow counter on each iteration of the main I/O loop. That is, if one job has
1338 \fBflow=8\fR and another job has \fBflow=-1\fR, then there will be a roughly
1339 1:8 ratio in how much one runs vs the other.
1341 .BI flow_watermark \fR=\fPint
1342 The maximum value that the absolute value of the flow counter is allowed to
1343 reach before the job must wait for a lower value of the counter.
1345 .BI flow_sleep \fR=\fPint
1346 The period of time, in microseconds, to wait after the flow watermark has been
1347 exceeded before retrying operations
1349 .BI clat_percentiles \fR=\fPbool
1350 Enable the reporting of percentiles of completion latencies.
1352 .BI percentile_list \fR=\fPfloat_list
1353 Overwrite the default list of percentiles for completion
1354 latencies. Each number is a floating number in the range (0,100], and
1355 the maximum length of the list is 20. Use ':' to separate the
1356 numbers. For example, \-\-percentile_list=99.5:99.9 will cause fio to
1357 report the values of completion latency below which 99.5% and 99.9% of
1358 the observed latencies fell, respectively.
1359 .SS "Ioengine Parameters List"
1360 Some parameters are only valid when a specific ioengine is in use. These are
1361 used identically to normal parameters, with the caveat that when used on the
1362 command line, the must come after the ioengine that defines them is selected.
1364 .BI (cpu)cpuload \fR=\fPint
1365 Attempt to use the specified percentage of CPU cycles.
1367 .BI (cpu)cpuchunks \fR=\fPint
1368 Split the load into cycles of the given time. In microseconds.
1370 .BI (libaio)userspace_reap
1371 Normally, with the libaio engine in use, fio will use
1372 the io_getevents system call to reap newly returned events.
1373 With this flag turned on, the AIO ring will be read directly
1374 from user-space to reap events. The reaping mode is only
1375 enabled when polling for a minimum of 0 events (eg when
1376 iodepth_batch_complete=0).
1378 .BI (net,netsplice)hostname \fR=\fPstr
1379 The host name or IP address to use for TCP or UDP based IO.
1380 If the job is a TCP listener or UDP reader, the hostname is not
1381 used and must be omitted unless it is a valid UDP multicast address.
1383 .BI (net,netsplice)port \fR=\fPint
1384 The TCP or UDP port to bind to or connect to.
1386 .BI (net,netsplice)interface \fR=\fPstr
1387 The IP address of the network interface used to send or receive UDP multicast
1390 .BI (net,netsplice)ttl \fR=\fPint
1391 Time-to-live value for outgoing UDP multicast packets. Default: 1
1393 .BI (net,netsplice)nodelay \fR=\fPbool
1394 Set TCP_NODELAY on TCP connections.
1396 .BI (net,netsplice)protocol \fR=\fPstr "\fR,\fP proto" \fR=\fPstr
1397 The network protocol to use. Accepted values are:
1402 Transmission control protocol
1405 Transmission control protocol V6
1408 User datagram protocol
1411 User datagram protocol V6
1417 When the protocol is TCP or UDP, the port must also be given,
1418 as well as the hostname if the job is a TCP listener or UDP
1419 reader. For unix sockets, the normal filename option should be
1420 used and the port is invalid.
1423 .BI (net,netsplice)listen
1424 For TCP network connections, tell fio to listen for incoming
1425 connections rather than initiating an outgoing connection. The
1426 hostname must be omitted if this option is used.
1428 .BI (net, pingpong) \fR=\fPbool
1429 Normally a network writer will just continue writing data, and a network reader
1430 will just consume packages. If pingpong=1 is set, a writer will send its normal
1431 payload to the reader, then wait for the reader to send the same payload back.
1432 This allows fio to measure network latencies. The submission and completion
1433 latencies then measure local time spent sending or receiving, and the
1434 completion latency measures how long it took for the other end to receive and
1435 send back. For UDP multicast traffic pingpong=1 should only be set for a single
1436 reader when multiple readers are listening to the same address.
1438 .BI (e4defrag,donorname) \fR=\fPstr
1439 File will be used as a block donor (swap extents between files)
1441 .BI (e4defrag,inplace) \fR=\fPint
1442 Configure donor file block allocation strategy
1445 Preallocate donor's file on init
1448 allocate space immediately inside defragment event, and free right after event
1451 .BI (rbd)rbdname \fR=\fPstr
1452 Specifies the name of the RBD.
1454 .BI (rbd)pool \fR=\fPstr
1455 Specifies the name of the Ceph pool containing the RBD.
1457 .BI (rbd)clientname \fR=\fPstr
1458 Specifies the username (without the 'client.' prefix) used to access the Ceph cluster.
1460 While running, \fBfio\fR will display the status of the created jobs. For
1464 Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
1467 The characters in the first set of brackets denote the current status of each
1468 threads. The possible values are:
1474 Setup but not started.
1480 Initialized, waiting.
1483 Running, doing sequential reads.
1486 Running, doing random reads.
1489 Running, doing sequential writes.
1492 Running, doing random writes.
1495 Running, doing mixed sequential reads/writes.
1498 Running, doing mixed random reads/writes.
1501 Running, currently waiting for \fBfsync\fR\|(2).
1504 Running, verifying written data.
1507 Exited, not reaped by main thread.
1510 Exited, thread reaped.
1514 The second set of brackets shows the estimated completion percentage of
1515 the current group. The third set shows the read and write I/O rate,
1516 respectively. Finally, the estimated run time of the job is displayed.
1518 When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data
1519 for each thread, each group of threads, and each disk, in that order.
1521 Per-thread statistics first show the threads client number, group-id, and
1522 error code. The remaining figures are as follows:
1526 Number of megabytes of I/O performed.
1529 Average data rate (bandwidth).
1535 Submission latency minimum, maximum, average and standard deviation. This is
1536 the time it took to submit the I/O.
1539 Completion latency minimum, maximum, average and standard deviation. This
1540 is the time between submission and completion.
1543 Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average
1544 and standard deviation.
1547 CPU usage statistics. Includes user and system time, number of context switches
1548 this thread went through and number of major and minor page faults.
1551 Distribution of I/O depths. Each depth includes everything less than (or equal)
1552 to it, but greater than the previous depth.
1555 Number of read/write requests issued, and number of short read/write requests.
1558 Distribution of I/O completion latencies. The numbers follow the same pattern
1562 The group statistics show:
1567 Number of megabytes I/O performed.
1570 Aggregate bandwidth of threads in the group.
1573 Minimum average bandwidth a thread saw.
1576 Maximum average bandwidth a thread saw.
1579 Shortest runtime of threads in the group.
1582 Longest runtime of threads in the group.
1586 Finally, disk statistics are printed with reads first:
1591 Number of I/Os performed by all groups.
1594 Number of merges in the I/O scheduler.
1597 Number of ticks we kept the disk busy.
1600 Total time spent in the disk queue.
1607 It is also possible to get fio to dump the current output while it is
1608 running, without terminating the job. To do that, send fio the \fBUSR1\fR
1611 If the \fB\-\-minimal\fR / \fB\-\-append-terse\fR options are given, the
1612 results will be printed/appended in a semicolon-delimited format suitable for
1614 A job description (if provided) follows on a new line. Note that the first
1615 number in the line is the version number. If the output has to be changed
1616 for some reason, this number will be incremented by 1 to signify that
1617 change. The fields are:
1620 .B terse version, fio version, jobname, groupid, error
1624 .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP
1628 .B min, max, mean, standard deviation
1632 .B min, max, mean, standard deviation
1634 Completion latency percentiles (20 fields):
1636 .B Xth percentile=usec
1640 .B min, max, mean, standard deviation
1644 .B min, max, aggregate percentage of total, mean, standard deviation
1650 .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP
1654 .B min, max, mean, standard deviation
1658 .B min, max, mean, standard deviation
1660 Completion latency percentiles (20 fields):
1662 .B Xth percentile=usec
1666 .B min, max, mean, standard deviation
1670 .B min, max, aggregate percentage of total, mean, standard deviation
1676 .B user, system, context switches, major page faults, minor page faults
1679 IO depth distribution:
1681 .B <=1, 2, 4, 8, 16, 32, >=64
1684 IO latency distribution:
1688 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
1692 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
1696 Disk utilization (1 for each disk used):
1698 .B name, read ios, write ios, read merges, write merges, read ticks, write ticks, read in-queue time, write in-queue time, disk utilization percentage
1701 Error Info (dependent on continue_on_error, default off):
1703 .B total # errors, first error code
1706 .B text description (if provided in config - appears on newline)
1709 Normally you would run fio as a stand-alone application on the machine
1710 where the IO workload should be generated. However, it is also possible to
1711 run the frontend and backend of fio separately. This makes it possible to
1712 have a fio server running on the machine(s) where the IO workload should
1713 be running, while controlling it from another machine.
1715 To start the server, you would do:
1717 \fBfio \-\-server=args\fR
1719 on that machine, where args defines what fio listens to. The arguments
1720 are of the form 'type:hostname or IP:port'. 'type' is either 'ip' (or ip4)
1721 for TCP/IP v4, 'ip6' for TCP/IP v6, or 'sock' for a local unix domain
1722 socket. 'hostname' is either a hostname or IP address, and 'port' is the port to
1723 listen to (only valid for TCP/IP, not a local socket). Some examples:
1727 Start a fio server, listening on all interfaces on the default port (8765).
1729 2) fio \-\-server=ip:hostname,4444
1731 Start a fio server, listening on IP belonging to hostname and on port 4444.
1733 3) fio \-\-server=ip6:::1,4444
1735 Start a fio server, listening on IPv6 localhost ::1 and on port 4444.
1737 4) fio \-\-server=,4444
1739 Start a fio server, listening on all interfaces on port 4444.
1741 5) fio \-\-server=1.2.3.4
1743 Start a fio server, listening on IP 1.2.3.4 on the default port.
1745 6) fio \-\-server=sock:/tmp/fio.sock
1747 Start a fio server, listening on the local socket /tmp/fio.sock.
1749 When a server is running, you can connect to it from a client. The client
1752 fio \-\-local-args \-\-client=server \-\-remote-args <job file(s)>
1754 where \-\-local-args are arguments that are local to the client where it is
1755 running, 'server' is the connect string, and \-\-remote-args and <job file(s)>
1756 are sent to the server. The 'server' string follows the same format as it
1757 does on the server side, to allow IP/hostname/socket and port strings.
1758 You can connect to multiple clients as well, to do that you could run:
1760 fio \-\-client=server2 \-\-client=server2 <job file(s)>
1764 was written by Jens Axboe <jens.axboe@oracle.com>,
1765 now Jens Axboe <jaxboe@fusionio.com>.
1767 This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based
1768 on documentation by Jens Axboe.
1769 .SH "REPORTING BUGS"
1770 Report bugs to the \fBfio\fR mailing list <fio@vger.kernel.org>.
1773 For further documentation see \fBHOWTO\fR and \fBREADME\fR.
1775 Sample jobfiles are available in the \fBexamples\fR directory.