1 .TH fio 1 "September 2007" "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 \-\-output \fR=\fPfilename
16 Write output to \fIfilename\fR.
18 .BI \-\-timeout \fR=\fPtimeout
19 Limit run time to \fItimeout\fR seconds.
22 Generate per-job latency logs.
25 Generate per-job bandwidth logs.
28 Print statistics in a terse, semicolon-delimited format.
30 .BI \-\-showcmd \fR=\fPjobfile
31 Convert \fIjobfile\fR to a set of command-line options.
34 Enable read-only safety checks.
36 .BI \-\-eta \fR=\fPwhen
37 Specifies when real-time ETA estimate should be printed. \fIwhen\fR may
38 be one of `always', `never' or `auto'.
40 .BI \-\-section \fR=\fPsec
41 Only run section \fIsec\fR from job file.
43 .BI \-\-cmdhelp \fR=\fPcommand
44 Print help information for \fIcommand\fR. May be `all' for all commands.
46 .BI \-\-debug \fR=\fPtype
47 Enable verbose tracing of various fio actions. May be `all' for all types
48 or individual types seperated by a comma (eg \-\-debug=io,file). `help' will
49 list all available tracing options.
52 Display usage information and exit.
55 Display version information and exit.
57 Job files are in `ini' format. They consist of one or more
58 job definitions, which begin with a job name in square brackets and
59 extend to the next job name. The job name can be any ASCII string
60 except `global', which has a special meaning. Following the job name is
61 a sequence of zero or more parameters, one per line, that define the
62 behavior of the job. Any line starting with a `;' or `#' character is
63 considered a comment and ignored.
65 If \fIjobfile\fR is specified as `-', the job file will be read from
68 The global section contains default parameters for jobs specified in the
69 job file. A job is only affected by global sections residing above it,
70 and there may be any number of global sections. Specific job definitions
71 may override any parameter set in global sections.
74 Some parameters may take arguments of a specific type. The types used are:
77 String: a sequence of alphanumeric characters.
80 SI integer: a whole number, possibly containing a suffix denoting the base unit
81 of the value. Accepted suffixes are `k', 'M', 'G', 'T', and 'P', denoting
82 kilo (1024), mega (1024^2), giga (1024^3), tera (1024^4), and peta (1024^5)
83 respectively. The suffix is not case sensitive. If prefixed with '0x', the
84 value is assumed to be base 16 (hexadecimal). A suffix may include a trailing
85 'b', for instance 'kb' is identical to 'k'. You can specify a base 10 value
86 by using 'KiB', 'MiB', 'GiB', etc. This is useful for disk drives where
87 values are often given in base 10 values. Specifying '30GiB' will get you
91 Boolean: a true or false value. `0' denotes false, `1' denotes true.
94 Integer range: a range of integers specified in the format
95 \fIlower\fR:\fIupper\fR or \fIlower\fR\-\fIupper\fR. \fIlower\fR and
96 \fIupper\fR may contain a suffix as described above. If an option allows two
97 sets of ranges, they are separated with a `,' or `/' character. For example:
102 May be used to override the job name. On the command line, this parameter
103 has the special purpose of signalling the start of a new job.
105 .BI description \fR=\fPstr
106 Human-readable description of the job. It is printed when the job is run, but
107 otherwise has no special purpose.
109 .BI directory \fR=\fPstr
110 Prefix filenames with this directory. Used to place files in a location other
113 .BI filename \fR=\fPstr
115 normally makes up a file name based on the job name, thread number, and file
116 number. If you want to share files between threads in a job or several jobs,
117 specify a \fIfilename\fR for each of them to override the default. If the I/O
118 engine used is `net', \fIfilename\fR is the host and port to connect to in the
119 format \fIhost\fR/\fIport\fR. If the I/O engine is file-based, you can specify
120 a number of files by separating the names with a `:' character. `\-' is a
121 reserved name, meaning stdin or stdout, depending on the read/write direction
124 .BI lockfile \fR=\fPstr
125 Fio defaults to not locking any files before it does IO to them. If a file or
126 file descriptor is shared, fio can serialize IO to that file to make the end
127 result consistent. This is usual for emulating real workloads that share files.
133 No locking. This is the default.
136 Only one thread or process may do IO at the time, excluding all others.
139 Read-write locking on the file. Many readers may access the file at the same
140 time, but writes get exclusive access.
143 The option may be post-fixed with a lock batch number. If set, then each
144 thread/process may do that amount of IOs to the file before giving up the lock.
145 Since lock acquisition is expensive, batching the lock/unlocks will speed up IO.
148 .BI opendir \fR=\fPstr
149 Recursively open any files below directory \fIstr\fR.
151 .BI readwrite \fR=\fPstr "\fR,\fP rw" \fR=\fPstr
152 Type of I/O pattern. Accepted values are:
169 Mixed sequential reads and writes.
172 Mixed random reads and writes.
175 For mixed I/O, the default split is 50/50. For random I/O, the number of I/Os
176 to perform before getting a new offset can be specified by appending
177 `:\fIint\fR' to the pattern type. The default is 1.
180 .BI kb_base \fR=\fPint
181 The base unit for a kilobyte. The defacto base is 2^10, 1024. Storage
182 manufacturers like to use 10^3 or 1000 as a base ten unit instead, for obvious
183 reasons. Allow values are 1024 or 1000, with 1024 being the default.
185 .BI randrepeat \fR=\fPbool
186 Seed the random number generator in a predictable way so results are repeatable
187 across runs. Default: true.
189 .BI fallocate \fR=\fPbool
190 By default, fio will use fallocate() to advise the system of the size of the
191 file we are going to write. This can be turned off with fallocate=0. May not
192 be available on all supported platforms.
194 .BI fadvise_hint \fR=\fPbool
195 Disable use of \fIposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
196 are likely to be issued. Default: true.
199 Total size of I/O for this job. \fBfio\fR will run until this many bytes have
200 been transfered, unless limited by other options (\fBruntime\fR, for instance).
201 Unless \fBnr_files\fR and \fBfilesize\fR options are given, this amount will be
202 divided between the available files for the job.
204 .BI fill_device \fR=\fPbool
205 Sets size to something really large and waits for ENOSPC (no space left on
206 device) as the terminating condition. Only makes sense with sequential write.
207 For a read workload, the mount point will be filled first then IO started on
210 .BI filesize \fR=\fPirange
211 Individual file sizes. May be a range, in which case \fBfio\fR will select sizes
212 for files at random within the given range, limited to \fBsize\fR in total (if
213 that is given). If \fBfilesize\fR is not specified, each created file is the
216 .BI blocksize \fR=\fPint[,int] "\fR,\fB bs" \fR=\fPint[,int]
217 Block size for I/O units. Default: 4k. Values for reads and writes can be
218 specified separately in the format \fIread\fR,\fIwrite\fR, either of
219 which may be empty to leave that value at its default.
221 .BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange]
222 Specify a range of I/O block sizes. The issued I/O unit will always be a
223 multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies
224 to both reads and writes if only one range is given, but can be specified
225 separately with a comma seperating the values. Example: bsrange=1k-4k,2k-8k.
226 Also (see \fBblocksize\fR).
228 .BI bssplit \fR=\fPstr
229 This option allows even finer grained control of the block sizes issued,
230 not just even splits between them. With this option, you can weight various
231 block sizes for exact control of the issued IO for a job that has mixed
232 block sizes. The format of the option is bssplit=blocksize/percentage,
233 optionally adding as many definitions as needed seperated by a colon.
234 Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k
235 blocks and 40% 32k blocks. \fBbssplit\fR also supports giving separate
236 splits to reads and writes. The format is identical to what the
237 \fBbs\fR option accepts, the read and write parts are separated with a
240 .B blocksize_unaligned\fR,\fP bs_unaligned
241 If set, any size in \fBblocksize_range\fR may be used. This typically won't
242 work with direct I/O, as that normally requires sector alignment.
244 .BI blockalign \fR=\fPint[,int] "\fR,\fB ba" \fR=\fPint[,int]
245 At what boundary to align random IO offsets. Defaults to the same as 'blocksize'
246 the minimum blocksize given. Minimum alignment is typically 512b
247 for using direct IO, though it usually depends on the hardware block size.
248 This option is mutually exclusive with using a random map for files, so it
249 will turn off that option.
252 Initialise buffers with all zeros. Default: fill buffers with random data.
255 If this option is given, fio will refill the IO buffers on every submit. The
256 default is to only fill it at init time and reuse that data. Only makes sense
257 if zero_buffers isn't specified, naturally. If data verification is enabled,
258 refill_buffers is also automatically enabled.
260 .BI nrfiles \fR=\fPint
261 Number of files to use for this job. Default: 1.
263 .BI openfiles \fR=\fPint
264 Number of files to keep open at the same time. Default: \fBnrfiles\fR.
266 .BI file_service_type \fR=\fPstr
267 Defines how files to service are selected. The following types are defined:
272 Choose a file at random
275 Round robin over open files (default).
277 Do each file in the set sequentially.
280 The number of I/Os to issue before switching a new file can be specified by
281 appending `:\fIint\fR' to the service type.
284 .BI ioengine \fR=\fPstr
285 Defines how the job issues I/O. The following types are defined:
290 Basic \fIread\fR\|(2) or \fIwrite\fR\|(2) I/O. \fIfseek\fR\|(2) is used to
291 position the I/O location.
294 Basic \fIpread\fR\|(2) or \fIpwrite\fR\|(2) I/O.
297 Basic \fIreadv\fR\|(2) or \fIwritev\fR\|(2) I/O. Will emulate queuing by
298 coalescing adjacents IOs into a single submission.
301 Linux native asynchronous I/O.
304 glibc POSIX asynchronous I/O using \fIaio_read\fR\|(3) and \fIaio_write\fR\|(3).
307 File is memory mapped with \fImmap\fR\|(2) and data copied using
311 \fIsplice\fR\|(2) is used to transfer the data and \fIvmsplice\fR\|(2) to
312 transfer data from user-space to the kernel.
315 Use the syslet system calls to make regular read/write asynchronous.
318 SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if
319 the target is an sg character device, we use \fIread\fR\|(2) and
320 \fIwrite\fR\|(2) for asynchronous I/O.
323 Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR
324 itself and for debugging and testing purposes.
327 Transfer over the network. \fBfilename\fR must be set appropriately to
328 `\fIhost\fR/\fIport\fR' regardless of data direction. If receiving, only the
329 \fIport\fR argument is used.
332 Like \fBnet\fR, but uses \fIsplice\fR\|(2) and \fIvmsplice\fR\|(2) to map data
336 Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and
337 \fBcpucycles\fR parameters.
340 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface
341 approach to asycnronous I/O.
343 See <http://www.xmailserver.org/guasi\-lib.html>.
346 Loads an external I/O engine object file. Append the engine filename as
351 .BI iodepth \fR=\fPint
352 Number of I/O units to keep in flight against the file. Default: 1.
354 .BI iodepth_batch \fR=\fPint
355 Number of I/Os to submit at once. Default: \fBiodepth\fR.
357 .BI iodepth_batch_complete \fR=\fPint
358 This defines how many pieces of IO to retrieve at once. It defaults to 1 which
359 means that we'll ask for a minimum of 1 IO in the retrieval process from the
360 kernel. The IO retrieval will go on until we hit the limit set by
361 \fBiodepth_low\fR. If this variable is set to 0, then fio will always check for
362 completed events before queuing more IO. This helps reduce IO latency, at the
363 cost of more retrieval system calls.
365 .BI iodepth_low \fR=\fPint
366 Low watermark indicating when to start filling the queue again. Default:
369 .BI direct \fR=\fPbool
370 If true, use non-buffered I/O (usually O_DIRECT). Default: false.
372 .BI buffered \fR=\fPbool
373 If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter.
376 .BI offset \fR=\fPint
377 Offset in the file to start I/O. Data before the offset will not be touched.
380 How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If
381 0, don't sync. Default: 0.
383 .BI fdatasync \fR=\fPint
384 Like \fBfsync\fR, but uses \fBfdatasync\fR\|(2) instead to only sync the
385 data parts of the file. Default: 0.
387 .BI sync_file_range \fR=\fPstr:int
388 Use sync_file_range() for every \fRval\fP number of write operations. Fio will
389 track range of writes that have happened since the last sync_file_range() call.
390 \fRstr\fP can currently be one or more of:
394 SYNC_FILE_RANGE_WAIT_BEFORE
397 SYNC_FILE_RANGE_WRITE
400 SYNC_FILE_RANGE_WRITE
404 So if you do sync_file_range=wait_before,write:8, fio would use
405 \fBSYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE\fP for every 8 writes.
406 Also see the sync_file_range(2) man page. This option is Linux specific.
408 .BI overwrite \fR=\fPbool
409 If writing, setup the file first and do overwrites. Default: false.
411 .BI end_fsync \fR=\fPbool
412 Sync file contents when job exits. Default: false.
414 .BI fsync_on_close \fR=\fPbool
415 If true, sync file contents on close. This differs from \fBend_fsync\fR in that
416 it will happen on every close, not just at the end of the job. Default: false.
418 .BI rwmixcycle \fR=\fPint
419 How many milliseconds before switching between reads and writes for a mixed
420 workload. Default: 500ms.
422 .BI rwmixread \fR=\fPint
423 Percentage of a mixed workload that should be reads. Default: 50.
425 .BI rwmixwrite \fR=\fPint
426 Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and
427 \fBrwmixwrite\fR are given and do not sum to 100%, the latter of the two
428 overrides the first. This may interfere with a given rate setting, if fio is
429 asked to limit reads or writes to a certain rate. If that is the case, then
430 the distribution may be skewed. Default: 50.
433 Normally \fBfio\fR will cover every block of the file when doing random I/O. If
434 this parameter is given, a new offset will be chosen without looking at past
435 I/O history. This parameter is mutually exclusive with \fBverify\fR.
438 See \fBnorandommap\fR. If fio runs with the random block map enabled and it
439 fails to allocate the map, if this option is set it will continue without a
440 random block map. As coverage will not be as complete as with random maps, this
441 option is disabled by default.
444 Run job with given nice value. See \fInice\fR\|(2).
447 Set I/O priority value of this job between 0 (highest) and 7 (lowest). See
450 .BI prioclass \fR=\fPint
451 Set I/O priority class. See \fIionice\fR\|(1).
453 .BI thinktime \fR=\fPint
454 Stall job for given number of microseconds between issuing I/Os.
456 .BI thinktime_spin \fR=\fPint
457 Pretend to spend CPU time for given number of microseconds, sleeping the rest
458 of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set.
460 .BI thinktime_blocks \fR=\fPint
461 Number of blocks to issue before waiting \fBthinktime\fR microseconds.
465 Cap bandwidth used by this job. The number is in bytes/sec, the normal postfix
466 rules apply. You can use \fBrate\fR=500k to limit reads and writes to 500k each,
467 or you can specify read and writes separately. Using \fBrate\fR=1m,500k would
468 limit reads to 1MB/sec and writes to 500KB/sec. Capping only reads or writes
469 can be done with \fBrate\fR=,500k or \fBrate\fR=500k,. The former will only
470 limit writes (to 500KB/sec), the latter will only limit reads.
472 .BI ratemin \fR=\fPint
473 Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth.
474 Failing to meet this requirement will cause the job to exit. The same format
475 as \fBrate\fR is used for read vs write separation.
477 .BI rate_iops \fR=\fPint
478 Cap the bandwidth to this number of IOPS. Basically the same as rate, just
479 specified independently of bandwidth. The same format as \fBrate\fR is used for
480 read vs write seperation. If \fBblocksize\fR is a range, the smallest block
481 size is used as the metric.
483 .BI rate_iops_min \fR=\fPint
484 If this rate of I/O is not met, the job will exit. The same format as \fBrate\fR
485 is used for read vs write seperation.
487 .BI ratecycle \fR=\fPint
488 Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of
489 milliseconds. Default: 1000ms.
491 .BI cpumask \fR=\fPint
492 Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job
493 may run on. See \fBsched_setaffinity\fR\|(2).
495 .BI cpus_allowed \fR=\fPstr
496 Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers.
498 .BI startdelay \fR=\fPint
499 Delay start of job for the specified number of seconds.
501 .BI runtime \fR=\fPint
502 Terminate processing after the specified number of seconds.
505 If given, run for the specified \fBruntime\fR duration even if the files are
506 completely read or written. The same workload will be repeated as many times
507 as \fBruntime\fR allows.
509 .BI ramp_time \fR=\fPint
510 If set, fio will run the specified workload for this amount of time before
511 logging any performance numbers. Useful for letting performance settle before
512 logging results, thus minimizing the runtime required for stable results. Note
513 that the \fBramp_time\fR is considered lead in time for a job, thus it will
514 increase the total runtime if a special timeout or runtime is specified.
516 .BI invalidate \fR=\fPbool
517 Invalidate buffer-cache for the file prior to starting I/O. Default: true.
520 Use synchronous I/O for buffered writes. For the majority of I/O engines,
521 this means using O_SYNC. Default: false.
523 .BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr
524 Allocation method for I/O unit buffer. Allowed values are:
529 Allocate memory with \fImalloc\fR\|(3).
532 Use shared memory buffers allocated through \fIshmget\fR\|(2).
535 Same as \fBshm\fR, but use huge pages as backing.
538 Use \fImmap\fR\|(2) for allocation. Uses anonymous memory unless a filename
539 is given after the option in the format `:\fIfile\fR'.
542 Same as \fBmmap\fR, but use huge files as backing.
545 The amount of memory allocated is the maximum allowed \fBblocksize\fR for the
546 job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work,
547 the system must have free huge pages allocated. \fBmmaphuge\fR also needs to
548 have hugetlbfs mounted, and \fIfile\fR must point there. At least on Linux,
549 huge pages must be manually allocated. See \fB/proc/sys/vm/nr_hugehages\fR
550 and the documentation for that. Normally you just need to echo an appropriate
551 number, eg echoing 8 will ensure that the OS has 8 huge pages ready for
555 .BI iomem_align \fR=\fPint
556 This indiciates the memory alignment of the IO memory buffers. Note that the
557 given alignment is applied to the first IO unit buffer, if using \fBiodepth\fR
558 the alignment of the following buffers are given by the \fBbs\fR used. In
559 other words, if using a \fBbs\fR that is a multiple of the page sized in the
560 system, all buffers will be aligned to this value. If using a \fBbs\fR that
561 is not page aligned, the alignment of subsequent IO memory buffers is the
562 sum of the \fBiomem_align\fR and \fBbs\fR used.
564 .BI hugepage\-size \fR=\fPint
565 Defines the size of a huge page. Must be at least equal to the system setting.
566 Should be a multiple of 1MB. Default: 4MB.
569 Terminate all jobs when one finishes. Default: wait for each job to finish.
571 .BI bwavgtime \fR=\fPint
572 Average bandwidth calculations over the given time in milliseconds. Default:
575 .BI create_serialize \fR=\fPbool
576 If true, serialize file creation for the jobs. Default: true.
578 .BI create_fsync \fR=\fPbool
579 \fIfsync\fR\|(2) data file after creation. Default: true.
581 .BI create_on_open \fR=\fPbool
582 If true, the files are not created until they are opened for IO by the job.
584 .BI pre_read \fR=\fPbool
585 If this is given, files will be pre-read into memory before starting the given
586 IO operation. This will also clear the \fR \fBinvalidate\fR flag, since it is
587 pointless to pre-read and then drop the cache. This will only work for IO
588 engines that are seekable, since they allow you to read the same data
589 multiple times. Thus it will not work on eg network or splice IO.
591 .BI unlink \fR=\fPbool
592 Unlink job files when done. Default: false.
595 Specifies the number of iterations (runs of the same workload) of this job.
598 .BI do_verify \fR=\fPbool
599 Run the verify phase after a write phase. Only valid if \fBverify\fR is set.
602 .BI verify \fR=\fPstr
603 Method of verifying file contents after each iteration of the job. Allowed
608 .B md5 crc16 crc32 crc32c crc32c-intel crc64 crc7 sha256 sha512 sha1
609 Store appropriate checksum in the header of each block. crc32c-intel is
610 hardware accelerated SSE4.2 driven, falls back to regular crc32c if
611 not supported by the system.
614 Write extra information about each I/O (timestamp, block number, etc.). The
615 block number is verified. See \fBverify_pattern\fR as well.
618 Pretend to verify. Used for testing internals.
621 This option can be used for repeated burn-in tests of a system to make sure
622 that the written data is also correctly read back. If the data direction given
623 is a read or random read, fio will assume that it should verify a previously
624 written file. If the data direction includes any form of write, the verify will
625 be of the newly written data.
628 .BI verify_sort \fR=\fPbool
629 If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to
630 read them back in a sorted manner. Default: true.
632 .BI verify_offset \fR=\fPint
633 Swap the verification header with data somewhere else in the block before
634 writing. It is swapped back before verifying.
636 .BI verify_interval \fR=\fPint
637 Write the verification header for this number of bytes, which should divide
638 \fBblocksize\fR. Default: \fBblocksize\fR.
640 .BI verify_pattern \fR=\fPstr
641 If set, fio will fill the io buffers with this pattern. Fio defaults to filling
642 with totally random bytes, but sometimes it's interesting to fill with a known
643 pattern for io verification purposes. Depending on the width of the pattern,
644 fio will fill 1/2/3/4 bytes of the buffer at the time(it can be either a
645 decimal or a hex number). The verify_pattern if larger than a 32-bit quantity
646 has to be a hex number that starts with either "0x" or "0X". Use with
649 .BI verify_fatal \fR=\fPbool
650 If true, exit the job on the first observed verification failure. Default:
653 .BI verify_async \fR=\fPint
654 Fio will normally verify IO inline from the submitting thread. This option
655 takes an integer describing how many async offload threads to create for IO
656 verification instead, causing fio to offload the duty of verifying IO contents
657 to one or more separate threads. If using this offload option, even sync IO
658 engines can benefit from using an \fBiodepth\fR setting higher than 1, as it
659 allows them to have IO in flight while verifies are running.
661 .BI verify_async_cpus \fR=\fPstr
662 Tell fio to set the given CPU affinity on the async IO verification threads.
663 See \fBcpus_allowed\fP for the format used.
665 .BI verify_backlog \fR=\fPint
666 Fio will normally verify the written contents of a job that utilizes verify
667 once that job has completed. In other words, everything is written then
668 everything is read back and verified. You may want to verify continually
669 instead for a variety of reasons. Fio stores the meta data associated with an
670 IO block in memory, so for large verify workloads, quite a bit of memory would
671 be used up holding this meta data. If this option is enabled, fio will verify
672 the previously written blocks before continuing to write new ones.
674 .BI verify_backlog_batch \fR=\fPint
675 Control how many blocks fio will verify if verify_backlog is set. If not set,
676 will default to the value of \fBverify_backlog\fR (meaning the entire queue is
677 read back and verified).
680 Wait for preceeding jobs in the job file to exit before starting this one.
681 \fBstonewall\fR implies \fBnew_group\fR.
684 Start a new reporting group. If not given, all jobs in a file will be part
685 of the same reporting group, unless separated by a stonewall.
687 .BI numjobs \fR=\fPint
688 Number of clones (processes/threads performing the same workload) of this job.
692 If set, display per-group reports instead of per-job when \fBnumjobs\fR is
696 Use threads created with \fBpthread_create\fR\|(3) instead of processes created
697 with \fBfork\fR\|(2).
699 .BI zonesize \fR=\fPint
700 Divide file into zones of the specified size in bytes. See \fBzoneskip\fR.
702 .BI zoneskip \fR=\fPint
703 Skip the specified number of bytes when \fBzonesize\fR bytes of data have been
706 .BI write_iolog \fR=\fPstr
707 Write the issued I/O patterns to the specified file.
709 .BI read_iolog \fR=\fPstr
710 Replay the I/O patterns contained in the specified file generated by
711 \fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file.
713 .B write_bw_log \fR=\fPstr
714 If given, write a bandwidth log of the jobs in this job file. Can be used to
715 store data of the bandwidth of the jobs in their lifetime. The included
716 fio_generate_plots script uses gnuplot to turn these text files into nice
717 graphs. See \fBwrite_log_log\fR for behaviour of given filename. For this
718 option, the postfix is _bw.log.
721 Same as \fBwrite_bw_log\fR, but writes I/O completion latencies. If no
722 filename is given with this option, the default filename of "jobname_type.log"
723 is used. Even if the filename is given, fio will still append the type of log.
725 .B disable_clat \fR=\fPbool
726 Disable measurements of completion latency numbers. Useful only for cutting
727 back the number of calls to gettimeofday, as that does impact performance at
728 really high IOPS rates. Note that to really get rid of a large amount of these
729 calls, this option must be used with disable_slat and disable_bw as well.
731 .B disable_slat \fR=\fPbool
732 Disable measurements of submission latency numbers. See \fBdisable_clat\fR.
734 .B disable_bw_measurement \fR=\fPbool
735 Disable measurements of throughput/bandwidth numbers. See \fBdisable_clat\fR.
737 .BI lockmem \fR=\fPint
738 Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to
739 simulate a smaller amount of memory.
741 .BI exec_prerun \fR=\fPstr
742 Before running the job, execute the specified command with \fBsystem\fR\|(3).
744 .BI exec_postrun \fR=\fPstr
745 Same as \fBexec_prerun\fR, but the command is executed after the job completes.
747 .BI ioscheduler \fR=\fPstr
748 Attempt to switch the device hosting the file to the specified I/O scheduler.
750 .BI cpuload \fR=\fPint
751 If the job is a CPU cycle-eater, attempt to use the specified percentage of
754 .BI cpuchunks \fR=\fPint
755 If the job is a CPU cycle-eater, split the load into cycles of the
756 given time in milliseconds.
758 .BI disk_util \fR=\fPbool
759 Generate disk utilization statistics if the platform supports it. Default: true.
761 .BI gtod_reduce \fR=\fPbool
762 Enable all of the gettimeofday() reducing options (disable_clat, disable_slat,
763 disable_bw) plus reduce precision of the timeout somewhat to really shrink the
764 gettimeofday() call count. With this option enabled, we only do about 0.4% of
765 the gtod() calls we would have done if all time keeping was enabled.
767 .BI gtod_cpu \fR=\fPint
768 Sometimes it's cheaper to dedicate a single thread of execution to just getting
769 the current time. Fio (and databases, for instance) are very intensive on
770 gettimeofday() calls. With this option, you can set one CPU aside for doing
771 nothing but logging current time to a shared memory location. Then the other
772 threads/processes that run IO workloads need only copy that segment, instead of
773 entering the kernel with a gettimeofday() call. The CPU set aside for doing
774 these time calls will be excluded from other uses. Fio will manually clear it
775 from the CPU mask of other jobs.
777 .BI cgroup \fR=\fPstr
778 Add job to this control group. If it doesn't exist, it will be created.
779 The system must have a mounted cgroup blkio mount point for this to work. If
780 your system doesn't have it mounted, you can do so with:
782 # mount -t cgroup -o blkio none /cgroup
784 .BI cgroup_weight \fR=\fPint
785 Set the weight of the cgroup to this value. See the documentation that comes
786 with the kernel, allowed values are in the range of 100..1000.
788 .BI cgroup_nodelete \fR=\fPbool
789 Normally fio will delete the cgroups it has created after the job completion.
790 To override this behavior and to leave cgroups around after the job completion,
791 set cgroup_nodelete=1. This can be useful if one wants to inspect various
792 cgroup files after job completion. Default: false
795 Instead of running as the invoking user, set the user ID to this value before
796 the thread/process does any work.
799 Set group ID, see \fBuid\fR.
801 While running, \fBfio\fR will display the status of the created jobs. For
805 Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
808 The characters in the first set of brackets denote the current status of each
809 threads. The possible values are:
815 Setup but not started.
821 Initialized, waiting.
824 Running, doing sequential reads.
827 Running, doing random reads.
830 Running, doing sequential writes.
833 Running, doing random writes.
836 Running, doing mixed sequential reads/writes.
839 Running, doing mixed random reads/writes.
842 Running, currently waiting for \fBfsync\fR\|(2).
845 Running, verifying written data.
848 Exited, not reaped by main thread.
851 Exited, thread reaped.
855 The second set of brackets shows the estimated completion percentage of
856 the current group. The third set shows the read and write I/O rate,
857 respectively. Finally, the estimated run time of the job is displayed.
859 When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data
860 for each thread, each group of threads, and each disk, in that order.
862 Per-thread statistics first show the threads client number, group-id, and
863 error code. The remaining figures are as follows:
867 Number of megabytes of I/O performed.
870 Average data rate (bandwidth).
876 Submission latency minimum, maximum, average and standard deviation. This is
877 the time it took to submit the I/O.
880 Completion latency minimum, maximum, average and standard deviation. This
881 is the time between submission and completion.
884 Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average
885 and standard deviation.
888 CPU usage statistics. Includes user and system time, number of context switches
889 this thread went through and number of major and minor page faults.
892 Distribution of I/O depths. Each depth includes everything less than (or equal)
893 to it, but greater than the previous depth.
896 Number of read/write requests issued, and number of short read/write requests.
899 Distribution of I/O completion latencies. The numbers follow the same pattern
903 The group statistics show:
908 Number of megabytes I/O performed.
911 Aggregate bandwidth of threads in the group.
914 Minimum average bandwidth a thread saw.
917 Maximum average bandwidth a thread saw.
920 Shortest runtime of threads in the group.
923 Longest runtime of threads in the group.
927 Finally, disk statistics are printed with reads first:
932 Number of I/Os performed by all groups.
935 Number of merges in the I/O scheduler.
938 Number of ticks we kept the disk busy.
941 Total time spent in the disk queue.
948 If the \fB\-\-minimal\fR option is given, the results will be printed in a
949 semicolon-delimited format suitable for scripted use. The fields are:
952 .B jobname, groupid, error
956 .B KB I/O, bandwidth \fR(KB/s)\fP, runtime \fR(ms)\fP
960 .B min, max, mean, standard deviation
964 .B min, max, mean, standard deviation
968 .B min, max, aggregate percentage of total, mean, standard deviation
974 .B KB I/O, bandwidth \fR(KB/s)\fP, runtime \fR(ms)\fP
978 .B min, max, mean, standard deviation
982 .B min, max, mean, standard deviation
986 .B min, max, aggregate percentage of total, mean, standard deviation
992 .B user, system, context switches, major page faults, minor page faults
995 IO depth distribution:
997 .B <=1, 2, 4, 8, 16, 32, >=64
1000 IO latency distribution (ms):
1002 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, >=2000
1009 was written by Jens Axboe <jens.axboe@oracle.com>,
1010 now Jens Axboe <jaxboe@fusionio.com>.
1012 This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based
1013 on documentation by Jens Axboe.
1014 .SH "REPORTING BUGS"
1015 Report bugs to the \fBfio\fR mailing list <fio@vger.kernel.org>.
1018 For further documentation see \fBHOWTO\fR and \fBREADME\fR.
1020 Sample jobfiles are available in the \fBexamples\fR directory.