1 .TH fio 1 "December 2014" "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 \-\-output-format \fR=\fPformat
24 Set the reporting format to \fInormal\fR, \fIterse\fR, or \fIjosn\fR.
26 .BI \-\-runtime \fR=\fPruntime
27 Limit run time to \fIruntime\fR seconds.
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. This option can be used multiple times to add 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.
119 Anywhere a numeric value is required, an arithmetic expression may be used,
120 provided it is surrounded by parentheses. Supported operators are:
128 .B multiplication (*)
134 .B exponentiation (^)
138 For time values in expressions, units are microseconds by default. This is
139 different than for time values not in expressions (not enclosed in
140 parentheses). The types used are:
143 String: a sequence of alphanumeric characters.
146 SI integer: a whole number, possibly containing a suffix denoting the base unit
147 of the value. Accepted suffixes are `k', 'M', 'G', 'T', and 'P', denoting
148 kilo (1024), mega (1024^2), giga (1024^3), tera (1024^4), and peta (1024^5)
149 respectively. If prefixed with '0x', the value is assumed to be base 16
150 (hexadecimal). A suffix may include a trailing 'b', for instance 'kb' is
151 identical to 'k'. You can specify a base 10 value by using 'KiB', 'MiB','GiB',
152 etc. This is useful for disk drives where values are often given in base 10
153 values. Specifying '30GiB' will get you 30*1000^3 bytes.
154 When specifying times the default suffix meaning changes, still denoting the
155 base unit of the value, but accepted suffixes are 'D' (days), 'H' (hours), 'M'
156 (minutes), 'S' Seconds, 'ms' (or msec) milli seconds, 'us' (or 'usec') micro
157 seconds. Time values without a unit specify seconds.
158 The suffixes are not case sensitive.
161 Boolean: a true or false value. `0' denotes false, `1' denotes true.
164 Integer range: a range of integers specified in the format
165 \fIlower\fR:\fIupper\fR or \fIlower\fR\-\fIupper\fR. \fIlower\fR and
166 \fIupper\fR may contain a suffix as described above. If an option allows two
167 sets of ranges, they are separated with a `,' or `/' character. For example:
171 List of floating numbers: A list of floating numbers, separated by
176 May be used to override the job name. On the command line, this parameter
177 has the special purpose of signalling the start of a new job.
179 .BI description \fR=\fPstr
180 Human-readable description of the job. It is printed when the job is run, but
181 otherwise has no special purpose.
183 .BI directory \fR=\fPstr
184 Prefix filenames with this directory. Used to place files in a location other
186 You can specify a number of directories by separating the names with a ':'
187 character. These directories will be assigned equally distributed to job clones
188 creates with \fInumjobs\fR as long as they are using generated filenames.
189 If specific \fIfilename(s)\fR are set fio will use the first listed directory,
190 and thereby matching the \fIfilename\fR semantic which generates a file each
191 clone if not specified, but let all clones use the same if set. See
192 \fIfilename\fR for considerations regarding escaping certain characters on
195 .BI filename \fR=\fPstr
197 normally makes up a file name based on the job name, thread number, and file
198 number. If you want to share files between threads in a job or several jobs,
199 specify a \fIfilename\fR for each of them to override the default.
200 If the I/O engine is file-based, you can specify
201 a number of files by separating the names with a `:' character. `\-' is a
202 reserved name, meaning stdin or stdout, depending on the read/write direction
203 set. On Windows, disk devices are accessed as \\.\PhysicalDrive0 for the first
204 device, \\.\PhysicalDrive1 for the second etc. Note: Windows and FreeBSD
205 prevent write access to areas of the disk containing in-use data
206 (e.g. filesystems). If the wanted filename does need to include a colon, then
207 escape that with a '\\' character. For instance, if the filename is
208 "/dev/dsk/foo@3,0:c", then you would use filename="/dev/dsk/foo@3,0\\:c".
210 .BI filename_format \fR=\fPstr
211 If sharing multiple files between jobs, it is usually necessary to have
212 fio generate the exact names that you want. By default, fio will name a file
213 based on the default file format specification of
214 \fBjobname.jobnumber.filenumber\fP. With this option, that can be
215 customized. Fio will recognize and replace the following keywords in this
221 The name of the worker thread or process.
224 The incremental number of the worker thread or process.
227 The incremental number of the file for that worker thread or process.
230 To have dependent jobs share a set of files, this option can be set to
231 have fio generate filenames that are shared between the two. For instance,
232 if \fBtestfiles.$filenum\fR is specified, file number 4 for any job will
233 be named \fBtestfiles.4\fR. The default of \fB$jobname.$jobnum.$filenum\fR
234 will be used if no other format specifier is given.
238 .BI lockfile \fR=\fPstr
239 Fio defaults to not locking any files before it does IO to them. If a file or
240 file descriptor is shared, fio can serialize IO to that file to make the end
241 result consistent. This is usual for emulating real workloads that share files.
247 No locking. This is the default.
250 Only one thread or process may do IO at a time, excluding all others.
253 Read-write locking on the file. Many readers may access the file at the same
254 time, but writes get exclusive access.
258 .BI opendir \fR=\fPstr
259 Recursively open any files below directory \fIstr\fR.
261 .BI readwrite \fR=\fPstr "\fR,\fP rw" \fR=\fPstr
262 Type of I/O pattern. Accepted values are:
273 Sequential trim (Linux block devices only).
282 Random trim (Linux block devices only).
285 Mixed sequential reads and writes.
288 Mixed random reads and writes.
291 For mixed I/O, the default split is 50/50. For certain types of io the result
292 may still be skewed a bit, since the speed may be different. It is possible to
293 specify a number of IO's to do before getting a new offset, this is done by
294 appending a `:\fI<nr>\fR to the end of the string given. For a random read, it
295 would look like \fBrw=randread:8\fR for passing in an offset modifier with a
296 value of 8. If the postfix is used with a sequential IO pattern, then the value
297 specified will be added to the generated offset for each IO. For instance,
298 using \fBrw=write:4k\fR will skip 4k for every write. It turns sequential IO
299 into sequential IO with holes. See the \fBrw_sequencer\fR option.
302 .BI rw_sequencer \fR=\fPstr
303 If an offset modifier is given by appending a number to the \fBrw=<str>\fR line,
304 then this option controls how that number modifies the IO offset being
305 generated. Accepted values are:
310 Generate sequential offset
313 Generate the same offset
316 \fBsequential\fR is only useful for random IO, where fio would normally
317 generate a new random offset for every IO. If you append eg 8 to randread, you
318 would get a new random offset for every 8 IO's. The result would be a seek for
319 only every 8 IO's, instead of for every IO. Use \fBrw=randread:8\fR to specify
320 that. As sequential IO is already sequential, setting \fBsequential\fR for that
321 would not result in any differences. \fBidentical\fR behaves in a similar
322 fashion, except it sends the same offset 8 number of times before generating a
327 .BI kb_base \fR=\fPint
328 The base unit for a kilobyte. The defacto base is 2^10, 1024. Storage
329 manufacturers like to use 10^3 or 1000 as a base ten unit instead, for obvious
330 reasons. Allowed values are 1024 or 1000, with 1024 being the default.
332 .BI unified_rw_reporting \fR=\fPbool
333 Fio normally reports statistics on a per data direction basis, meaning that
334 read, write, and trim are accounted and reported separately. If this option is
335 set fio sums the results and reports them as "mixed" instead.
337 .BI randrepeat \fR=\fPbool
338 Seed the random number generator used for random I/O patterns in a predictable
339 way so the pattern is repeatable across runs. Default: true.
341 .BI allrandrepeat \fR=\fPbool
342 Seed all random number generators in a predictable way so results are
343 repeatable across runs. Default: false.
345 .BI randseed \fR=\fPint
346 Seed the random number generators based on this seed value, to be able to
347 control what sequence of output is being generated. If not set, the random
348 sequence depends on the \fBrandrepeat\fR setting.
350 .BI fallocate \fR=\fPstr
351 Whether pre-allocation is performed when laying down files. Accepted values
357 Do not pre-allocate space.
360 Pre-allocate via \fBposix_fallocate\fR\|(3).
363 Pre-allocate via \fBfallocate\fR\|(2) with FALLOC_FL_KEEP_SIZE set.
366 Backward-compatible alias for 'none'.
369 Backward-compatible alias for 'posix'.
372 May not be available on all supported platforms. 'keep' is only
373 available on Linux. If using ZFS on Solaris this must be set to 'none'
374 because ZFS doesn't support it. Default: 'posix'.
377 .BI fadvise_hint \fR=\fPbool
378 Use \fBposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
379 are likely to be issued. Default: true.
382 Total size of I/O for this job. \fBfio\fR will run until this many bytes have
383 been transferred, unless limited by other options (\fBruntime\fR, for instance,
384 or increased/descreased by \fBio_size\fR). Unless \fBnrfiles\fR and
385 \fBfilesize\fR options are given, this amount will be divided between the
386 available files for the job. If not set, fio will use the full size of the
387 given files or devices. If the files do not exist, size must be given. It is
388 also possible to give size as a percentage between 1 and 100. If size=20% is
389 given, fio will use 20% of the full size of the given files or devices.
391 .BI io_size \fR=\fPint "\fR,\fB io_limit \fR=\fPint
392 Normally fio operates within the region set by \fBsize\fR, which means that
393 the \fBsize\fR option sets both the region and size of IO to be performed.
394 Sometimes that is not what you want. With this option, it is possible to
395 define just the amount of IO that fio should do. For instance, if \fBsize\fR
396 is set to 20G and \fBio_limit\fR is set to 5G, fio will perform IO within
397 the first 20G but exit when 5G have been done. The opposite is also
398 possible - if \fBsize\fR is set to 20G, and \fBio_size\fR is set to 40G, then
399 fio will do 40G of IO within the 0..20G region.
401 .BI fill_device \fR=\fPbool "\fR,\fB fill_fs" \fR=\fPbool
402 Sets size to something really large and waits for ENOSPC (no space left on
403 device) as the terminating condition. Only makes sense with sequential write.
404 For a read workload, the mount point will be filled first then IO started on
405 the result. This option doesn't make sense if operating on a raw device node,
406 since the size of that is already known by the file system. Additionally,
407 writing beyond end-of-device will not return ENOSPC there.
409 .BI filesize \fR=\fPirange
410 Individual file sizes. May be a range, in which case \fBfio\fR will select sizes
411 for files at random within the given range, limited to \fBsize\fR in total (if
412 that is given). If \fBfilesize\fR is not specified, each created file is the
415 .BI file_append \fR=\fPbool
416 Perform IO after the end of the file. Normally fio will operate within the
417 size of a file. If this option is set, then fio will append to the file
418 instead. This has identical behavior to setting \fRoffset\fP to the size
419 of a file. This option is ignored on non-regular files.
421 .BI blocksize \fR=\fPint[,int] "\fR,\fB bs" \fR=\fPint[,int]
422 Block size for I/O units. Default: 4k. Values for reads, writes, and trims
423 can be specified separately in the format \fIread\fR,\fIwrite\fR,\fItrim\fR
424 either of which may be empty to leave that value at its default. If a trailing
425 comma isn't given, the remainder will inherit the last value set.
427 .BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange]
428 Specify a range of I/O block sizes. The issued I/O unit will always be a
429 multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies
430 to both reads and writes if only one range is given, but can be specified
431 separately with a comma separating the values. Example: bsrange=1k-4k,2k-8k.
432 Also (see \fBblocksize\fR).
434 .BI bssplit \fR=\fPstr
435 This option allows even finer grained control of the block sizes issued,
436 not just even splits between them. With this option, you can weight various
437 block sizes for exact control of the issued IO for a job that has mixed
438 block sizes. The format of the option is bssplit=blocksize/percentage,
439 optionally adding as many definitions as needed separated by a colon.
440 Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k
441 blocks and 40% 32k blocks. \fBbssplit\fR also supports giving separate
442 splits to reads and writes. The format is identical to what the
443 \fBbs\fR option accepts, the read and write parts are separated with a
446 .B blocksize_unaligned\fR,\fP bs_unaligned
447 If set, any size in \fBblocksize_range\fR may be used. This typically won't
448 work with direct I/O, as that normally requires sector alignment.
450 .BI blockalign \fR=\fPint[,int] "\fR,\fB ba" \fR=\fPint[,int]
451 At what boundary to align random IO offsets. Defaults to the same as 'blocksize'
452 the minimum blocksize given. Minimum alignment is typically 512b
453 for using direct IO, though it usually depends on the hardware block size.
454 This option is mutually exclusive with using a random map for files, so it
455 will turn off that option.
457 .BI bs_is_seq_rand \fR=\fPbool
458 If this option is set, fio will use the normal read,write blocksize settings as
459 sequential,random instead. Any random read or write will use the WRITE
460 blocksize settings, and any sequential read or write will use the READ
464 Initialize buffers with all zeros. Default: fill buffers with random data.
465 The resulting IO buffers will not be completely zeroed, unless
466 \fPscramble_buffers\fR is also turned off.
469 If this option is given, fio will refill the IO buffers on every submit. The
470 default is to only fill it at init time and reuse that data. Only makes sense
471 if zero_buffers isn't specified, naturally. If data verification is enabled,
472 refill_buffers is also automatically enabled.
474 .BI scramble_buffers \fR=\fPbool
475 If \fBrefill_buffers\fR is too costly and the target is using data
476 deduplication, then setting this option will slightly modify the IO buffer
477 contents to defeat normal de-dupe attempts. This is not enough to defeat
478 more clever block compression attempts, but it will stop naive dedupe
479 of blocks. Default: true.
481 .BI buffer_compress_percentage \fR=\fPint
482 If this is set, then fio will attempt to provide IO buffer content (on WRITEs)
483 that compress to the specified level. Fio does this by providing a mix of
484 random data and a fixed pattern. The fixed pattern is either zeroes, or the
485 pattern specified by \fBbuffer_pattern\fR. If the pattern option is used, it
486 might skew the compression ratio slightly. Note that this is per block size
487 unit, for file/disk wide compression level that matches this setting. Note
488 that this is per block size unit, for file/disk wide compression level that
489 matches this setting, you'll also want to set refill_buffers.
491 .BI buffer_compress_chunk \fR=\fPint
492 See \fBbuffer_compress_percentage\fR. This setting allows fio to manage how
493 big the ranges of random data and zeroed data is. Without this set, fio will
494 provide \fBbuffer_compress_percentage\fR of blocksize random data, followed by
495 the remaining zeroed. With this set to some chunk size smaller than the block
496 size, fio can alternate random and zeroed data throughout the IO buffer.
498 .BI buffer_pattern \fR=\fPstr
499 If set, fio will fill the IO buffers with this pattern. If not set, the contents
500 of IO buffers is defined by the other options related to buffer contents. The
501 setting can be any pattern of bytes, and can be prefixed with 0x for hex
502 values. It may also be a string, where the string must then be wrapped with
505 .BI dedupe_percentage \fR=\fPint
506 If set, fio will generate this percentage of identical buffers when writing.
507 These buffers will be naturally dedupable. The contents of the buffers depend
508 on what other buffer compression settings have been set. It's possible to have
509 the individual buffers either fully compressible, or not at all. This option
510 only controls the distribution of unique buffers.
512 .BI nrfiles \fR=\fPint
513 Number of files to use for this job. Default: 1.
515 .BI openfiles \fR=\fPint
516 Number of files to keep open at the same time. Default: \fBnrfiles\fR.
518 .BI file_service_type \fR=\fPstr
519 Defines how files to service are selected. The following types are defined:
524 Choose a file at random.
527 Round robin over opened files (default).
530 Do each file in the set sequentially.
533 The number of I/Os to issue before switching to a new file can be specified by
534 appending `:\fIint\fR' to the service type.
537 .BI ioengine \fR=\fPstr
538 Defines how the job issues I/O. The following types are defined:
543 Basic \fBread\fR\|(2) or \fBwrite\fR\|(2) I/O. \fBfseek\fR\|(2) is used to
544 position the I/O location.
547 Basic \fBpread\fR\|(2) or \fBpwrite\fR\|(2) I/O.
550 Basic \fBreadv\fR\|(2) or \fBwritev\fR\|(2) I/O. Will emulate queuing by
551 coalescing adjacent IOs into a single submission.
554 Basic \fBpreadv\fR\|(2) or \fBpwritev\fR\|(2) I/O.
557 Linux native asynchronous I/O. This ioengine defines engine specific options.
560 POSIX asynchronous I/O using \fBaio_read\fR\|(3) and \fBaio_write\fR\|(3).
563 Solaris native asynchronous I/O.
566 Windows native asynchronous I/O.
569 File is memory mapped with \fBmmap\fR\|(2) and data copied using
573 \fBsplice\fR\|(2) is used to transfer the data and \fBvmsplice\fR\|(2) to
574 transfer data from user-space to the kernel.
577 Use the syslet system calls to make regular read/write asynchronous.
580 SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if
581 the target is an sg character device, we use \fBread\fR\|(2) and
582 \fBwrite\fR\|(2) for asynchronous I/O.
585 Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR
586 itself and for debugging and testing purposes.
589 Transfer over the network. The protocol to be used can be defined with the
590 \fBprotocol\fR parameter. Depending on the protocol, \fBfilename\fR,
591 \fBhostname\fR, \fBport\fR, or \fBlisten\fR must be specified.
592 This ioengine defines engine specific options.
595 Like \fBnet\fR, but uses \fBsplice\fR\|(2) and \fBvmsplice\fR\|(2) to map data
596 and send/receive. This ioengine defines engine specific options.
599 Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and
600 \fBcpucycles\fR parameters.
603 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface
604 approach to asynchronous I/O.
606 See <http://www.xmailserver.org/guasi\-lib.html>.
609 The RDMA I/O engine supports both RDMA memory semantics (RDMA_WRITE/RDMA_READ)
610 and channel semantics (Send/Recv) for the InfiniBand, RoCE and iWARP protocols.
613 Loads an external I/O engine object file. Append the engine filename as
617 IO engine that does regular linux native fallocate call to simulate data
618 transfer as fio ioengine
620 DDIR_READ does fallocate(,mode = FALLOC_FL_KEEP_SIZE,)
622 DIR_WRITE does fallocate(,mode = 0)
624 DDIR_TRIM does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE)
627 IO engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate defragment activity
628 request to DDIR_WRITE event
631 IO engine supporting direct access to Ceph Rados Block Devices (RBD) via librbd
632 without the need to use the kernel rbd driver. This ioengine defines engine specific
636 Using Glusterfs libgfapi sync interface to direct access to Glusterfs volumes without
637 having to go through FUSE. This ioengine defines engine specific
641 Using Glusterfs libgfapi async interface to direct access to Glusterfs volumes without
642 having to go through FUSE. This ioengine defines engine specific
646 Read and write through Hadoop (HDFS). The \fBfilename\fR option is used to
647 specify host,port of the hdfs name-node to connect. This engine interprets
648 offsets a little differently. In HDFS, files once created cannot be modified.
649 So random writes are not possible. To imitate this, libhdfs engine expects
650 bunch of small files to be created over HDFS, and engine will randomly pick a
651 file out of those files based on the offset generated by fio backend. (see the
652 example job file to create such files, use rw=write option). Please note, you
653 might want to set necessary environment variables to work with hdfs/libhdfs
659 .BI iodepth \fR=\fPint
660 Number of I/O units to keep in flight against the file. Note that increasing
661 iodepth beyond 1 will not affect synchronous ioengines (except for small
662 degress when verify_async is in use). Even async engines may impose OS
663 restrictions causing the desired depth not to be achieved. This may happen on
664 Linux when using libaio and not setting \fBdirect\fR=1, since buffered IO is
665 not async on that OS. Keep an eye on the IO depth distribution in the
666 fio output to verify that the achieved depth is as expected. Default: 1.
668 .BI iodepth_batch \fR=\fPint
669 Number of I/Os to submit at once. Default: \fBiodepth\fR.
671 .BI iodepth_batch_complete \fR=\fPint
672 This defines how many pieces of IO to retrieve at once. It defaults to 1 which
673 means that we'll ask for a minimum of 1 IO in the retrieval process from the
674 kernel. The IO retrieval will go on until we hit the limit set by
675 \fBiodepth_low\fR. If this variable is set to 0, then fio will always check for
676 completed events before queuing more IO. This helps reduce IO latency, at the
677 cost of more retrieval system calls.
679 .BI iodepth_low \fR=\fPint
680 Low watermark indicating when to start filling the queue again. Default:
683 .BI direct \fR=\fPbool
684 If true, use non-buffered I/O (usually O_DIRECT). Default: false.
686 .BI atomic \fR=\fPbool
687 If value is true, attempt to use atomic direct IO. Atomic writes are guaranteed
688 to be stable once acknowledged by the operating system. Only Linux supports
691 .BI buffered \fR=\fPbool
692 If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter.
695 .BI offset \fR=\fPint
696 Offset in the file to start I/O. Data before the offset will not be touched.
698 .BI offset_increment \fR=\fPint
699 If this is provided, then the real offset becomes the
700 offset + offset_increment * thread_number, where the thread number is a
701 counter that starts at 0 and is incremented for each sub-job (i.e. when
702 numjobs option is specified). This option is useful if there are several jobs
703 which are intended to operate on a file in parallel disjoint segments, with
704 even spacing between the starting points.
706 .BI number_ios \fR=\fPint
707 Fio will normally perform IOs until it has exhausted the size of the region
708 set by \fBsize\fR, or if it exhaust the allocated time (or hits an error
709 condition). With this setting, the range/size can be set independently of
710 the number of IOs to perform. When fio reaches this number, it will exit
711 normally and report status. Note that this does not extend the amount
712 of IO that will be done, it will only stop fio if this condition is met
713 before other end-of-job criteria.
716 How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If
717 0, don't sync. Default: 0.
719 .BI fdatasync \fR=\fPint
720 Like \fBfsync\fR, but uses \fBfdatasync\fR\|(2) instead to only sync the
721 data parts of the file. Default: 0.
723 .BI write_barrier \fR=\fPint
724 Make every Nth write a barrier write.
726 .BI sync_file_range \fR=\fPstr:int
727 Use \fBsync_file_range\fR\|(2) for every \fRval\fP number of write operations. Fio will
728 track range of writes that have happened since the last \fBsync_file_range\fR\|(2) call.
729 \fRstr\fP can currently be one or more of:
733 SYNC_FILE_RANGE_WAIT_BEFORE
736 SYNC_FILE_RANGE_WRITE
739 SYNC_FILE_RANGE_WRITE
743 So if you do sync_file_range=wait_before,write:8, fio would use
744 \fBSYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE\fP for every 8 writes.
745 Also see the \fBsync_file_range\fR\|(2) man page. This option is Linux specific.
747 .BI overwrite \fR=\fPbool
748 If writing, setup the file first and do overwrites. Default: false.
750 .BI end_fsync \fR=\fPbool
751 Sync file contents when a write stage has completed. Default: false.
753 .BI fsync_on_close \fR=\fPbool
754 If true, sync file contents on close. This differs from \fBend_fsync\fR in that
755 it will happen on every close, not just at the end of the job. Default: false.
757 .BI rwmixread \fR=\fPint
758 Percentage of a mixed workload that should be reads. Default: 50.
760 .BI rwmixwrite \fR=\fPint
761 Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and
762 \fBrwmixwrite\fR are given and do not sum to 100%, the latter of the two
763 overrides the first. This may interfere with a given rate setting, if fio is
764 asked to limit reads or writes to a certain rate. If that is the case, then
765 the distribution may be skewed. Default: 50.
767 .BI random_distribution \fR=\fPstr:float
768 By default, fio will use a completely uniform random distribution when asked
769 to perform random IO. Sometimes it is useful to skew the distribution in
770 specific ways, ensuring that some parts of the data is more hot than others.
771 Fio includes the following distribution models:
775 Uniform random distribution
785 When using a zipf or pareto distribution, an input value is also needed to
786 define the access pattern. For zipf, this is the zipf theta. For pareto,
787 it's the pareto power. Fio includes a test program, genzipf, that can be
788 used visualize what the given input values will yield in terms of hit rates.
789 If you wanted to use zipf with a theta of 1.2, you would use
790 random_distribution=zipf:1.2 as the option. If a non-uniform model is used,
791 fio will disable use of the random map.
793 .BI percentage_random \fR=\fPint
794 For a random workload, set how big a percentage should be random. This defaults
795 to 100%, in which case the workload is fully random. It can be set from
796 anywhere from 0 to 100. Setting it to 0 would make the workload fully
797 sequential. It is possible to set different values for reads, writes, and
798 trim. To do so, simply use a comma separated list. See \fBblocksize\fR.
801 Normally \fBfio\fR will cover every block of the file when doing random I/O. If
802 this parameter is given, a new offset will be chosen without looking at past
803 I/O history. This parameter is mutually exclusive with \fBverify\fR.
805 .BI softrandommap \fR=\fPbool
806 See \fBnorandommap\fR. If fio runs with the random block map enabled and it
807 fails to allocate the map, if this option is set it will continue without a
808 random block map. As coverage will not be as complete as with random maps, this
809 option is disabled by default.
811 .BI random_generator \fR=\fPstr
812 Fio supports the following engines for generating IO offsets for random IO:
816 Strong 2^88 cycle random number generator
819 Linear feedback shift register generator
823 Tausworthe is a strong random number generator, but it requires tracking on the
824 side if we want to ensure that blocks are only read or written once. LFSR
825 guarantees that we never generate the same offset twice, and it's also less
826 computationally expensive. It's not a true random generator, however, though
827 for IO purposes it's typically good enough. LFSR only works with single block
828 sizes, not with workloads that use multiple block sizes. If used with such a
829 workload, fio may read or write some blocks multiple times.
832 Run job with given nice value. See \fBnice\fR\|(2).
835 Set I/O priority value of this job between 0 (highest) and 7 (lowest). See
838 .BI prioclass \fR=\fPint
839 Set I/O priority class. See \fBionice\fR\|(1).
841 .BI thinktime \fR=\fPint
842 Stall job for given number of microseconds between issuing I/Os.
844 .BI thinktime_spin \fR=\fPint
845 Pretend to spend CPU time for given number of microseconds, sleeping the rest
846 of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set.
848 .BI thinktime_blocks \fR=\fPint
849 Only valid if thinktime is set - control how many blocks to issue, before
850 waiting \fBthinktime\fR microseconds. If not set, defaults to 1 which will
851 make fio wait \fBthinktime\fR microseconds after every block. This
852 effectively makes any queue depth setting redundant, since no more than 1 IO
853 will be queued before we have to complete it and do our thinktime. In other
854 words, this setting effectively caps the queue depth if the latter is larger.
858 Cap bandwidth used by this job. The number is in bytes/sec, the normal postfix
859 rules apply. You can use \fBrate\fR=500k to limit reads and writes to 500k each,
860 or you can specify read and writes separately. Using \fBrate\fR=1m,500k would
861 limit reads to 1MB/sec and writes to 500KB/sec. Capping only reads or writes
862 can be done with \fBrate\fR=,500k or \fBrate\fR=500k,. The former will only
863 limit writes (to 500KB/sec), the latter will only limit reads.
865 .BI ratemin \fR=\fPint
866 Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth.
867 Failing to meet this requirement will cause the job to exit. The same format
868 as \fBrate\fR is used for read vs write separation.
870 .BI rate_iops \fR=\fPint
871 Cap the bandwidth to this number of IOPS. Basically the same as rate, just
872 specified independently of bandwidth. The same format as \fBrate\fR is used for
873 read vs write separation. If \fBblocksize\fR is a range, the smallest block
874 size is used as the metric.
876 .BI rate_iops_min \fR=\fPint
877 If this rate of I/O is not met, the job will exit. The same format as \fBrate\fR
878 is used for read vs write separation.
880 .BI ratecycle \fR=\fPint
881 Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of
882 milliseconds. Default: 1000ms.
884 .BI latency_target \fR=\fPint
885 If set, fio will attempt to find the max performance point that the given
886 workload will run at while maintaining a latency below this target. The
887 values is given in microseconds. See \fBlatency_window\fR and
888 \fBlatency_percentile\fR.
890 .BI latency_window \fR=\fPint
891 Used with \fBlatency_target\fR to specify the sample window that the job
892 is run at varying queue depths to test the performance. The value is given
895 .BI latency_percentile \fR=\fPfloat
896 The percentage of IOs that must fall within the criteria specified by
897 \fBlatency_target\fR and \fBlatency_window\fR. If not set, this defaults
898 to 100.0, meaning that all IOs must be equal or below to the value set
899 by \fBlatency_target\fR.
901 .BI max_latency \fR=\fPint
902 If set, fio will exit the job if it exceeds this maximum latency. It will exit
905 .BI cpumask \fR=\fPint
906 Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job
907 may run on. See \fBsched_setaffinity\fR\|(2).
909 .BI cpus_allowed \fR=\fPstr
910 Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers.
912 .BI cpus_allowed_policy \fR=\fPstr
913 Set the policy of how fio distributes the CPUs specified by \fBcpus_allowed\fR
914 or \fBcpumask\fR. Two policies are supported:
919 All jobs will share the CPU set specified.
922 Each job will get a unique CPU from the CPU set.
925 \fBshared\fR is the default behaviour, if the option isn't specified. If
926 \fBsplit\fR is specified, then fio will assign one cpu per job. If not enough
927 CPUs are given for the jobs listed, then fio will roundrobin the CPUs in
932 .BI numa_cpu_nodes \fR=\fPstr
933 Set this job running on specified NUMA nodes' CPUs. The arguments allow
934 comma delimited list of cpu numbers, A-B ranges, or 'all'.
936 .BI numa_mem_policy \fR=\fPstr
937 Set this job's memory policy and corresponding NUMA nodes. Format of
941 .B <mode>[:<nodelist>]
944 is one of the following memory policy:
946 .B default, prefer, bind, interleave, local
949 For \fBdefault\fR and \fBlocal\fR memory policy, no \fBnodelist\fR is
950 needed to be specified. For \fBprefer\fR, only one node is
951 allowed. For \fBbind\fR and \fBinterleave\fR, \fBnodelist\fR allows
952 comma delimited list of numbers, A-B ranges, or 'all'.
954 .BI startdelay \fR=\fPirange
955 Delay start of job for the specified number of seconds. Supports all time
956 suffixes to allow specification of hours, minutes, seconds and
957 milliseconds - seconds are the default if a unit is omitted.
958 Can be given as a range which causes each thread to choose randomly out of the
961 .BI runtime \fR=\fPint
962 Terminate processing after the specified number of seconds.
965 If given, run for the specified \fBruntime\fR duration even if the files are
966 completely read or written. The same workload will be repeated as many times
967 as \fBruntime\fR allows.
969 .BI ramp_time \fR=\fPint
970 If set, fio will run the specified workload for this amount of time before
971 logging any performance numbers. Useful for letting performance settle before
972 logging results, thus minimizing the runtime required for stable results. Note
973 that the \fBramp_time\fR is considered lead in time for a job, thus it will
974 increase the total runtime if a special timeout or runtime is specified.
976 .BI invalidate \fR=\fPbool
977 Invalidate buffer-cache for the file prior to starting I/O. Default: true.
980 Use synchronous I/O for buffered writes. For the majority of I/O engines,
981 this means using O_SYNC. Default: false.
983 .BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr
984 Allocation method for I/O unit buffer. Allowed values are:
989 Allocate memory with \fBmalloc\fR\|(3).
992 Use shared memory buffers allocated through \fBshmget\fR\|(2).
995 Same as \fBshm\fR, but use huge pages as backing.
998 Use \fBmmap\fR\|(2) for allocation. Uses anonymous memory unless a filename
999 is given after the option in the format `:\fIfile\fR'.
1002 Same as \fBmmap\fR, but use huge files as backing.
1005 The amount of memory allocated is the maximum allowed \fBblocksize\fR for the
1006 job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work,
1007 the system must have free huge pages allocated. \fBmmaphuge\fR also needs to
1008 have hugetlbfs mounted, and \fIfile\fR must point there. At least on Linux,
1009 huge pages must be manually allocated. See \fB/proc/sys/vm/nr_hugehages\fR
1010 and the documentation for that. Normally you just need to echo an appropriate
1011 number, eg echoing 8 will ensure that the OS has 8 huge pages ready for
1015 .BI iomem_align \fR=\fPint "\fR,\fP mem_align" \fR=\fPint
1016 This indicates the memory alignment of the IO memory buffers. Note that the
1017 given alignment is applied to the first IO unit buffer, if using \fBiodepth\fR
1018 the alignment of the following buffers are given by the \fBbs\fR used. In
1019 other words, if using a \fBbs\fR that is a multiple of the page sized in the
1020 system, all buffers will be aligned to this value. If using a \fBbs\fR that
1021 is not page aligned, the alignment of subsequent IO memory buffers is the
1022 sum of the \fBiomem_align\fR and \fBbs\fR used.
1024 .BI hugepage\-size \fR=\fPint
1025 Defines the size of a huge page. Must be at least equal to the system setting.
1026 Should be a multiple of 1MB. Default: 4MB.
1029 Terminate all jobs when one finishes. Default: wait for each job to finish.
1031 .BI bwavgtime \fR=\fPint
1032 Average bandwidth calculations over the given time in milliseconds. Default:
1035 .BI iopsavgtime \fR=\fPint
1036 Average IOPS calculations over the given time in milliseconds. Default:
1039 .BI create_serialize \fR=\fPbool
1040 If true, serialize file creation for the jobs. Default: true.
1042 .BI create_fsync \fR=\fPbool
1043 \fBfsync\fR\|(2) data file after creation. Default: true.
1045 .BI create_on_open \fR=\fPbool
1046 If true, the files are not created until they are opened for IO by the job.
1048 .BI create_only \fR=\fPbool
1049 If true, fio will only run the setup phase of the job. If files need to be
1050 laid out or updated on disk, only that will be done. The actual job contents
1053 .BI pre_read \fR=\fPbool
1054 If this is given, files will be pre-read into memory before starting the given
1055 IO operation. This will also clear the \fR \fBinvalidate\fR flag, since it is
1056 pointless to pre-read and then drop the cache. This will only work for IO
1057 engines that are seekable, since they allow you to read the same data
1058 multiple times. Thus it will not work on eg network or splice IO.
1060 .BI unlink \fR=\fPbool
1061 Unlink job files when done. Default: false.
1063 .BI loops \fR=\fPint
1064 Specifies the number of iterations (runs of the same workload) of this job.
1067 .BI verify_only \fR=\fPbool
1068 Do not perform the specified workload, only verify data still matches previous
1069 invocation of this workload. This option allows one to check data multiple
1070 times at a later date without overwriting it. This option makes sense only for
1071 workloads that write data, and does not support workloads with the
1072 \fBtime_based\fR option set.
1074 .BI do_verify \fR=\fPbool
1075 Run the verify phase after a write phase. Only valid if \fBverify\fR is set.
1078 .BI verify \fR=\fPstr
1079 Method of verifying file contents after each iteration of the job. Allowed
1084 .B md5 crc16 crc32 crc32c crc32c-intel crc64 crc7 sha256 sha512 sha1 xxhash
1085 Store appropriate checksum in the header of each block. crc32c-intel is
1086 hardware accelerated SSE4.2 driven, falls back to regular crc32c if
1087 not supported by the system.
1090 Write extra information about each I/O (timestamp, block number, etc.). The
1091 block number is verified. See \fBverify_pattern\fR as well.
1094 Pretend to verify. Used for testing internals.
1097 This option can be used for repeated burn-in tests of a system to make sure
1098 that the written data is also correctly read back. If the data direction given
1099 is a read or random read, fio will assume that it should verify a previously
1100 written file. If the data direction includes any form of write, the verify will
1101 be of the newly written data.
1104 .BI verifysort \fR=\fPbool
1105 If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to
1106 read them back in a sorted manner. Default: true.
1108 .BI verifysort_nr \fR=\fPint
1109 Pre-load and sort verify blocks for a read workload.
1111 .BI verify_offset \fR=\fPint
1112 Swap the verification header with data somewhere else in the block before
1113 writing. It is swapped back before verifying.
1115 .BI verify_interval \fR=\fPint
1116 Write the verification header for this number of bytes, which should divide
1117 \fBblocksize\fR. Default: \fBblocksize\fR.
1119 .BI verify_pattern \fR=\fPstr
1120 If set, fio will fill the io buffers with this pattern. Fio defaults to filling
1121 with totally random bytes, but sometimes it's interesting to fill with a known
1122 pattern for io verification purposes. Depending on the width of the pattern,
1123 fio will fill 1/2/3/4 bytes of the buffer at the time(it can be either a
1124 decimal or a hex number). The verify_pattern if larger than a 32-bit quantity
1125 has to be a hex number that starts with either "0x" or "0X". Use with
1128 .BI verify_fatal \fR=\fPbool
1129 If true, exit the job on the first observed verification failure. Default:
1132 .BI verify_dump \fR=\fPbool
1133 If set, dump the contents of both the original data block and the data block we
1134 read off disk to files. This allows later analysis to inspect just what kind of
1135 data corruption occurred. Off by default.
1137 .BI verify_async \fR=\fPint
1138 Fio will normally verify IO inline from the submitting thread. This option
1139 takes an integer describing how many async offload threads to create for IO
1140 verification instead, causing fio to offload the duty of verifying IO contents
1141 to one or more separate threads. If using this offload option, even sync IO
1142 engines can benefit from using an \fBiodepth\fR setting higher than 1, as it
1143 allows them to have IO in flight while verifies are running.
1145 .BI verify_async_cpus \fR=\fPstr
1146 Tell fio to set the given CPU affinity on the async IO verification threads.
1147 See \fBcpus_allowed\fP for the format used.
1149 .BI verify_backlog \fR=\fPint
1150 Fio will normally verify the written contents of a job that utilizes verify
1151 once that job has completed. In other words, everything is written then
1152 everything is read back and verified. You may want to verify continually
1153 instead for a variety of reasons. Fio stores the meta data associated with an
1154 IO block in memory, so for large verify workloads, quite a bit of memory would
1155 be used up holding this meta data. If this option is enabled, fio will write
1156 only N blocks before verifying these blocks.
1158 .BI verify_backlog_batch \fR=\fPint
1159 Control how many blocks fio will verify if verify_backlog is set. If not set,
1160 will default to the value of \fBverify_backlog\fR (meaning the entire queue is
1161 read back and verified). If \fBverify_backlog_batch\fR is less than
1162 \fBverify_backlog\fR then not all blocks will be verified, if
1163 \fBverify_backlog_batch\fR is larger than \fBverify_backlog\fR, some blocks
1164 will be verified more than once.
1166 .BI trim_percentage \fR=\fPint
1167 Number of verify blocks to discard/trim.
1169 .BI trim_verify_zero \fR=\fPbool
1170 Verify that trim/discarded blocks are returned as zeroes.
1172 .BI trim_backlog \fR=\fPint
1173 Trim after this number of blocks are written.
1175 .BI trim_backlog_batch \fR=\fPint
1176 Trim this number of IO blocks.
1178 .BI experimental_verify \fR=\fPbool
1179 Enable experimental verification.
1181 .BI verify_state_save \fR=\fPbool
1182 When a job exits during the write phase of a verify workload, save its
1183 current state. This allows fio to replay up until that point, if the
1184 verify state is loaded for the verify read phase.
1186 .BI verify_state_load \fR=\fPbool
1187 If a verify termination trigger was used, fio stores the current write
1188 state of each thread. This can be used at verification time so that fio
1189 knows how far it should verify. Without this information, fio will run
1190 a full verification pass, according to the settings in the job file used.
1192 .B stonewall "\fR,\fP wait_for_previous"
1193 Wait for preceding jobs in the job file to exit before starting this one.
1194 \fBstonewall\fR implies \fBnew_group\fR.
1197 Start a new reporting group. If not given, all jobs in a file will be part
1198 of the same reporting group, unless separated by a stonewall.
1200 .BI numjobs \fR=\fPint
1201 Number of clones (processes/threads performing the same workload) of this job.
1205 If set, display per-group reports instead of per-job when \fBnumjobs\fR is
1209 Use threads created with \fBpthread_create\fR\|(3) instead of processes created
1210 with \fBfork\fR\|(2).
1212 .BI zonesize \fR=\fPint
1213 Divide file into zones of the specified size in bytes. See \fBzoneskip\fR.
1215 .BI zonerange \fR=\fPint
1216 Give size of an IO zone. See \fBzoneskip\fR.
1218 .BI zoneskip \fR=\fPint
1219 Skip the specified number of bytes when \fBzonesize\fR bytes of data have been
1222 .BI write_iolog \fR=\fPstr
1223 Write the issued I/O patterns to the specified file. Specify a separate file
1224 for each job, otherwise the iologs will be interspersed and the file may be
1227 .BI read_iolog \fR=\fPstr
1228 Replay the I/O patterns contained in the specified file generated by
1229 \fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file.
1231 .BI replay_no_stall \fR=\fPint
1232 While replaying I/O patterns using \fBread_iolog\fR the default behavior
1233 attempts to respect timing information between I/Os. Enabling
1234 \fBreplay_no_stall\fR causes I/Os to be replayed as fast as possible while
1235 still respecting ordering.
1237 .BI replay_redirect \fR=\fPstr
1238 While replaying I/O patterns using \fBread_iolog\fR the default behavior
1239 is to replay the IOPS onto the major/minor device that each IOP was recorded
1240 from. Setting \fBreplay_redirect\fR causes all IOPS to be replayed onto the
1241 single specified device regardless of the device it was recorded from.
1243 .BI write_bw_log \fR=\fPstr
1244 If given, write a bandwidth log of the jobs in this job file. Can be used to
1245 store data of the bandwidth of the jobs in their lifetime. The included
1246 fio_generate_plots script uses gnuplot to turn these text files into nice
1247 graphs. See \fBwrite_lat_log\fR for behaviour of given filename. For this
1248 option, the postfix is _bw.x.log, where x is the index of the job (1..N,
1249 where N is the number of jobs)
1251 .BI write_lat_log \fR=\fPstr
1252 Same as \fBwrite_bw_log\fR, but writes I/O completion latencies. If no
1253 filename is given with this option, the default filename of
1254 "jobname_type.x.log" is used, where x is the index of the job (1..N, where
1255 N is the number of jobs). Even if the filename is given, fio will still
1256 append the type of log.
1258 .BI write_iops_log \fR=\fPstr
1259 Same as \fBwrite_bw_log\fR, but writes IOPS. If no filename is given with this
1260 option, the default filename of "jobname_type.x.log" is used, where x is the
1261 index of the job (1..N, where N is the number of jobs). Even if the filename
1262 is given, fio will still append the type of log.
1264 .BI log_avg_msec \fR=\fPint
1265 By default, fio will log an entry in the iops, latency, or bw log for every
1266 IO that completes. When writing to the disk log, that can quickly grow to a
1267 very large size. Setting this option makes fio average the each log entry
1268 over the specified period of time, reducing the resolution of the log.
1271 .BI log_offset \fR=\fPbool
1272 If this is set, the iolog options will include the byte offset for the IO
1273 entry as well as the other data values.
1275 .BI log_compression \fR=\fPint
1276 If this is set, fio will compress the IO logs as it goes, to keep the memory
1277 footprint lower. When a log reaches the specified size, that chunk is removed
1278 and compressed in the background. Given that IO logs are fairly highly
1279 compressible, this yields a nice memory savings for longer runs. The downside
1280 is that the compression will consume some background CPU cycles, so it may
1281 impact the run. This, however, is also true if the logging ends up consuming
1282 most of the system memory. So pick your poison. The IO logs are saved
1283 normally at the end of a run, by decompressing the chunks and storing them
1284 in the specified log file. This feature depends on the availability of zlib.
1286 .BI log_store_compressed \fR=\fPbool
1287 If set, and \fBlog\fR_compression is also set, fio will store the log files in
1288 a compressed format. They can be decompressed with fio, using the
1289 \fB\-\-inflate-log\fR command line parameter. The files will be stored with a
1292 .BI disable_lat \fR=\fPbool
1293 Disable measurements of total latency numbers. Useful only for cutting
1294 back the number of calls to \fBgettimeofday\fR\|(2), as that does impact performance at
1295 really high IOPS rates. Note that to really get rid of a large amount of these
1296 calls, this option must be used with disable_slat and disable_bw as well.
1298 .BI disable_clat \fR=\fPbool
1299 Disable measurements of completion latency numbers. See \fBdisable_lat\fR.
1301 .BI disable_slat \fR=\fPbool
1302 Disable measurements of submission latency numbers. See \fBdisable_lat\fR.
1304 .BI disable_bw_measurement \fR=\fPbool
1305 Disable measurements of throughput/bandwidth numbers. See \fBdisable_lat\fR.
1307 .BI lockmem \fR=\fPint
1308 Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to
1309 simulate a smaller amount of memory. The amount specified is per worker.
1311 .BI exec_prerun \fR=\fPstr
1312 Before running the job, execute the specified command with \fBsystem\fR\|(3).
1314 Output is redirected in a file called \fBjobname.prerun.txt\fR
1317 .BI exec_postrun \fR=\fPstr
1318 Same as \fBexec_prerun\fR, but the command is executed after the job completes.
1320 Output is redirected in a file called \fBjobname.postrun.txt\fR
1323 .BI ioscheduler \fR=\fPstr
1324 Attempt to switch the device hosting the file to the specified I/O scheduler.
1326 .BI disk_util \fR=\fPbool
1327 Generate disk utilization statistics if the platform supports it. Default: true.
1329 .BI clocksource \fR=\fPstr
1330 Use the given clocksource as the base of timing. The supported options are:
1334 \fBgettimeofday\fR\|(2)
1337 \fBclock_gettime\fR\|(2)
1340 Internal CPU clock source
1344 \fBcpu\fR is the preferred clocksource if it is reliable, as it is very fast
1345 (and fio is heavy on time calls). Fio will automatically use this clocksource
1346 if it's supported and considered reliable on the system it is running on,
1347 unless another clocksource is specifically set. For x86/x86-64 CPUs, this
1348 means supporting TSC Invariant.
1350 .BI gtod_reduce \fR=\fPbool
1351 Enable all of the \fBgettimeofday\fR\|(2) reducing options (disable_clat, disable_slat,
1352 disable_bw) plus reduce precision of the timeout somewhat to really shrink the
1353 \fBgettimeofday\fR\|(2) call count. With this option enabled, we only do about 0.4% of
1354 the gtod() calls we would have done if all time keeping was enabled.
1356 .BI gtod_cpu \fR=\fPint
1357 Sometimes it's cheaper to dedicate a single thread of execution to just getting
1358 the current time. Fio (and databases, for instance) are very intensive on
1359 \fBgettimeofday\fR\|(2) calls. With this option, you can set one CPU aside for doing
1360 nothing but logging current time to a shared memory location. Then the other
1361 threads/processes that run IO workloads need only copy that segment, instead of
1362 entering the kernel with a \fBgettimeofday\fR\|(2) call. The CPU set aside for doing
1363 these time calls will be excluded from other uses. Fio will manually clear it
1364 from the CPU mask of other jobs.
1366 .BI ignore_error \fR=\fPstr
1367 Sometimes you want to ignore some errors during test in that case you can specify
1368 error list for each error type.
1370 ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST
1372 errors for given error type is separated with ':'.
1373 Error may be symbol ('ENOSPC', 'ENOMEM') or an integer.
1375 Example: ignore_error=EAGAIN,ENOSPC:122 .
1377 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from WRITE.
1379 .BI error_dump \fR=\fPbool
1380 If set dump every error even if it is non fatal, true by default. If disabled
1381 only fatal error will be dumped
1383 .BI profile \fR=\fPstr
1384 Select a specific builtin performance test.
1386 .BI cgroup \fR=\fPstr
1387 Add job to this control group. If it doesn't exist, it will be created.
1388 The system must have a mounted cgroup blkio mount point for this to work. If
1389 your system doesn't have it mounted, you can do so with:
1391 # mount \-t cgroup \-o blkio none /cgroup
1393 .BI cgroup_weight \fR=\fPint
1394 Set the weight of the cgroup to this value. See the documentation that comes
1395 with the kernel, allowed values are in the range of 100..1000.
1397 .BI cgroup_nodelete \fR=\fPbool
1398 Normally fio will delete the cgroups it has created after the job completion.
1399 To override this behavior and to leave cgroups around after the job completion,
1400 set cgroup_nodelete=1. This can be useful if one wants to inspect various
1401 cgroup files after job completion. Default: false
1404 Instead of running as the invoking user, set the user ID to this value before
1405 the thread/process does any work.
1408 Set group ID, see \fBuid\fR.
1410 .BI unit_base \fR=\fPint
1411 Base unit for reporting. Allowed values are:
1415 Use auto-detection (default).
1425 .BI flow_id \fR=\fPint
1426 The ID of the flow. If not specified, it defaults to being a global flow. See
1430 Weight in token-based flow control. If this value is used, then there is a
1431 \fBflow counter\fR which is used to regulate the proportion of activity between
1432 two or more jobs. fio attempts to keep this flow counter near zero. The
1433 \fBflow\fR parameter stands for how much should be added or subtracted to the
1434 flow counter on each iteration of the main I/O loop. That is, if one job has
1435 \fBflow=8\fR and another job has \fBflow=-1\fR, then there will be a roughly
1436 1:8 ratio in how much one runs vs the other.
1438 .BI flow_watermark \fR=\fPint
1439 The maximum value that the absolute value of the flow counter is allowed to
1440 reach before the job must wait for a lower value of the counter.
1442 .BI flow_sleep \fR=\fPint
1443 The period of time, in microseconds, to wait after the flow watermark has been
1444 exceeded before retrying operations
1446 .BI clat_percentiles \fR=\fPbool
1447 Enable the reporting of percentiles of completion latencies.
1449 .BI percentile_list \fR=\fPfloat_list
1450 Overwrite the default list of percentiles for completion
1451 latencies. Each number is a floating number in the range (0,100], and
1452 the maximum length of the list is 20. Use ':' to separate the
1453 numbers. For example, \-\-percentile_list=99.5:99.9 will cause fio to
1454 report the values of completion latency below which 99.5% and 99.9% of
1455 the observed latencies fell, respectively.
1456 .SS "Ioengine Parameters List"
1457 Some parameters are only valid when a specific ioengine is in use. These are
1458 used identically to normal parameters, with the caveat that when used on the
1459 command line, they must come after the ioengine.
1461 .BI (cpu)cpuload \fR=\fPint
1462 Attempt to use the specified percentage of CPU cycles.
1464 .BI (cpu)cpuchunks \fR=\fPint
1465 Split the load into cycles of the given time. In microseconds.
1467 .BI (cpu)exit_on_io_done \fR=\fPbool
1468 Detect when IO threads are done, then exit.
1470 .BI (libaio)userspace_reap
1471 Normally, with the libaio engine in use, fio will use
1472 the io_getevents system call to reap newly returned events.
1473 With this flag turned on, the AIO ring will be read directly
1474 from user-space to reap events. The reaping mode is only
1475 enabled when polling for a minimum of 0 events (eg when
1476 iodepth_batch_complete=0).
1478 .BI (net,netsplice)hostname \fR=\fPstr
1479 The host name or IP address to use for TCP or UDP based IO.
1480 If the job is a TCP listener or UDP reader, the hostname is not
1481 used and must be omitted unless it is a valid UDP multicast address.
1483 .BI (net,netsplice)port \fR=\fPint
1484 The TCP or UDP port to bind to or connect to. If this is used with
1485 \fBnumjobs\fR to spawn multiple instances of the same job type, then
1486 this will be the starting port number since fio will use a range of ports.
1488 .BI (net,netsplice)interface \fR=\fPstr
1489 The IP address of the network interface used to send or receive UDP multicast
1492 .BI (net,netsplice)ttl \fR=\fPint
1493 Time-to-live value for outgoing UDP multicast packets. Default: 1
1495 .BI (net,netsplice)nodelay \fR=\fPbool
1496 Set TCP_NODELAY on TCP connections.
1498 .BI (net,netsplice)protocol \fR=\fPstr "\fR,\fP proto" \fR=\fPstr
1499 The network protocol to use. Accepted values are:
1504 Transmission control protocol
1507 Transmission control protocol V6
1510 User datagram protocol
1513 User datagram protocol V6
1519 When the protocol is TCP or UDP, the port must also be given,
1520 as well as the hostname if the job is a TCP listener or UDP
1521 reader. For unix sockets, the normal filename option should be
1522 used and the port is invalid.
1525 .BI (net,netsplice)listen
1526 For TCP network connections, tell fio to listen for incoming
1527 connections rather than initiating an outgoing connection. The
1528 hostname must be omitted if this option is used.
1530 .BI (net, pingpong) \fR=\fPbool
1531 Normally a network writer will just continue writing data, and a network reader
1532 will just consume packets. If pingpong=1 is set, a writer will send its normal
1533 payload to the reader, then wait for the reader to send the same payload back.
1534 This allows fio to measure network latencies. The submission and completion
1535 latencies then measure local time spent sending or receiving, and the
1536 completion latency measures how long it took for the other end to receive and
1537 send back. For UDP multicast traffic pingpong=1 should only be set for a single
1538 reader when multiple readers are listening to the same address.
1540 .BI (net, window_size) \fR=\fPint
1541 Set the desired socket buffer size for the connection.
1543 .BI (net, mss) \fR=\fPint
1544 Set the TCP maximum segment size (TCP_MAXSEG).
1546 .BI (e4defrag,donorname) \fR=\fPstr
1547 File will be used as a block donor (swap extents between files)
1549 .BI (e4defrag,inplace) \fR=\fPint
1550 Configure donor file block allocation strategy
1553 Preallocate donor's file on init
1556 allocate space immediately inside defragment event, and free right after event
1559 .BI (rbd)rbdname \fR=\fPstr
1560 Specifies the name of the RBD.
1562 .BI (rbd)pool \fR=\fPstr
1563 Specifies the name of the Ceph pool containing the RBD.
1565 .BI (rbd)clientname \fR=\fPstr
1566 Specifies the username (without the 'client.' prefix) used to access the Ceph cluster.
1568 While running, \fBfio\fR will display the status of the created jobs. For
1572 Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
1575 The characters in the first set of brackets denote the current status of each
1576 threads. The possible values are:
1582 Setup but not started.
1588 Initialized, waiting.
1591 Running, doing sequential reads.
1594 Running, doing random reads.
1597 Running, doing sequential writes.
1600 Running, doing random writes.
1603 Running, doing mixed sequential reads/writes.
1606 Running, doing mixed random reads/writes.
1609 Running, currently waiting for \fBfsync\fR\|(2).
1612 Running, verifying written data.
1615 Exited, not reaped by main thread.
1618 Exited, thread reaped.
1622 The second set of brackets shows the estimated completion percentage of
1623 the current group. The third set shows the read and write I/O rate,
1624 respectively. Finally, the estimated run time of the job is displayed.
1626 When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data
1627 for each thread, each group of threads, and each disk, in that order.
1629 Per-thread statistics first show the threads client number, group-id, and
1630 error code. The remaining figures are as follows:
1634 Number of megabytes of I/O performed.
1637 Average data rate (bandwidth).
1643 Submission latency minimum, maximum, average and standard deviation. This is
1644 the time it took to submit the I/O.
1647 Completion latency minimum, maximum, average and standard deviation. This
1648 is the time between submission and completion.
1651 Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average
1652 and standard deviation.
1655 CPU usage statistics. Includes user and system time, number of context switches
1656 this thread went through and number of major and minor page faults.
1659 Distribution of I/O depths. Each depth includes everything less than (or equal)
1660 to it, but greater than the previous depth.
1663 Number of read/write requests issued, and number of short read/write requests.
1666 Distribution of I/O completion latencies. The numbers follow the same pattern
1670 The group statistics show:
1675 Number of megabytes I/O performed.
1678 Aggregate bandwidth of threads in the group.
1681 Minimum average bandwidth a thread saw.
1684 Maximum average bandwidth a thread saw.
1687 Shortest runtime of threads in the group.
1690 Longest runtime of threads in the group.
1694 Finally, disk statistics are printed with reads first:
1699 Number of I/Os performed by all groups.
1702 Number of merges in the I/O scheduler.
1705 Number of ticks we kept the disk busy.
1708 Total time spent in the disk queue.
1715 It is also possible to get fio to dump the current output while it is
1716 running, without terminating the job. To do that, send fio the \fBUSR1\fR
1719 If the \fB\-\-minimal\fR / \fB\-\-append-terse\fR options are given, the
1720 results will be printed/appended in a semicolon-delimited format suitable for
1722 A job description (if provided) follows on a new line. Note that the first
1723 number in the line is the version number. If the output has to be changed
1724 for some reason, this number will be incremented by 1 to signify that
1725 change. The fields are:
1728 .B terse version, fio version, jobname, groupid, error
1732 .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP
1736 .B min, max, mean, standard deviation
1740 .B min, max, mean, standard deviation
1742 Completion latency percentiles (20 fields):
1744 .B Xth percentile=usec
1748 .B min, max, mean, standard deviation
1752 .B min, max, aggregate percentage of total, mean, standard deviation
1758 .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP
1762 .B min, max, mean, standard deviation
1766 .B min, max, mean, standard deviation
1768 Completion latency percentiles (20 fields):
1770 .B Xth percentile=usec
1774 .B min, max, mean, standard deviation
1778 .B min, max, aggregate percentage of total, mean, standard deviation
1784 .B user, system, context switches, major page faults, minor page faults
1787 IO depth distribution:
1789 .B <=1, 2, 4, 8, 16, 32, >=64
1792 IO latency distribution:
1796 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
1800 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
1804 Disk utilization (1 for each disk used):
1806 .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
1809 Error Info (dependent on continue_on_error, default off):
1811 .B total # errors, first error code
1814 .B text description (if provided in config - appears on newline)
1817 Normally you would run fio as a stand-alone application on the machine
1818 where the IO workload should be generated. However, it is also possible to
1819 run the frontend and backend of fio separately. This makes it possible to
1820 have a fio server running on the machine(s) where the IO workload should
1821 be running, while controlling it from another machine.
1823 To start the server, you would do:
1825 \fBfio \-\-server=args\fR
1827 on that machine, where args defines what fio listens to. The arguments
1828 are of the form 'type:hostname or IP:port'. 'type' is either 'ip' (or ip4)
1829 for TCP/IP v4, 'ip6' for TCP/IP v6, or 'sock' for a local unix domain
1830 socket. 'hostname' is either a hostname or IP address, and 'port' is the port to
1831 listen to (only valid for TCP/IP, not a local socket). Some examples:
1835 Start a fio server, listening on all interfaces on the default port (8765).
1837 2) fio \-\-server=ip:hostname,4444
1839 Start a fio server, listening on IP belonging to hostname and on port 4444.
1841 3) fio \-\-server=ip6:::1,4444
1843 Start a fio server, listening on IPv6 localhost ::1 and on port 4444.
1845 4) fio \-\-server=,4444
1847 Start a fio server, listening on all interfaces on port 4444.
1849 5) fio \-\-server=1.2.3.4
1851 Start a fio server, listening on IP 1.2.3.4 on the default port.
1853 6) fio \-\-server=sock:/tmp/fio.sock
1855 Start a fio server, listening on the local socket /tmp/fio.sock.
1857 When a server is running, you can connect to it from a client. The client
1860 fio \-\-local-args \-\-client=server \-\-remote-args <job file(s)>
1862 where \-\-local-args are arguments that are local to the client where it is
1863 running, 'server' is the connect string, and \-\-remote-args and <job file(s)>
1864 are sent to the server. The 'server' string follows the same format as it
1865 does on the server side, to allow IP/hostname/socket and port strings.
1866 You can connect to multiple clients as well, to do that you could run:
1868 fio \-\-client=server2 \-\-client=server2 <job file(s)>
1870 If the job file is located on the fio server, then you can tell the server
1871 to load a local file as well. This is done by using \-\-remote-config:
1873 fio \-\-client=server \-\-remote-config /path/to/file.fio
1875 Then the fio serer will open this local (to the server) job file instead
1876 of being passed one from the client.
1880 was written by Jens Axboe <jens.axboe@oracle.com>,
1881 now Jens Axboe <axboe@fb.com>.
1883 This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based
1884 on documentation by Jens Axboe.
1885 .SH "REPORTING BUGS"
1886 Report bugs to the \fBfio\fR mailing list <fio@vger.kernel.org>.
1889 For further documentation see \fBHOWTO\fR and \fBREADME\fR.
1891 Sample jobfiles are available in the \fBexamples\fR directory.
projects / fio.git / blob