Add write barriers
[fio.git] / fio.1
CommitLineData
d60e92d1
AC
1.TH fio 1 "September 2007" "User Manual"
2.SH NAME
3fio \- flexible I/O tester
4.SH SYNOPSIS
5.B fio
6[\fIoptions\fR] [\fIjobfile\fR]...
7.SH DESCRIPTION
8.B fio
9is a tool that will spawn a number of threads or processes doing a
10particular type of I/O action as specified by the user.
11The typical use of fio is to write a job file matching the I/O load
12one wants to simulate.
13.SH OPTIONS
14.TP
15.BI \-\-output \fR=\fPfilename
16Write output to \fIfilename\fR.
17.TP
18.BI \-\-timeout \fR=\fPtimeout
19Limit run time to \fItimeout\fR seconds.
20.TP
21.B \-\-latency\-log
22Generate per-job latency logs.
23.TP
24.B \-\-bandwidth\-log
25Generate per-job bandwidth logs.
26.TP
27.B \-\-minimal
d1429b5c 28Print statistics in a terse, semicolon-delimited format.
d60e92d1
AC
29.TP
30.BI \-\-showcmd \fR=\fPjobfile
31Convert \fIjobfile\fR to a set of command-line options.
32.TP
33.B \-\-readonly
34Enable read-only safety checks.
35.TP
36.BI \-\-eta \fR=\fPwhen
37Specifies when real-time ETA estimate should be printed. \fIwhen\fR may
38be one of `always', `never' or `auto'.
39.TP
c0a5d35e
AC
40.BI \-\-section \fR=\fPsec
41Only run section \fIsec\fR from job file.
42.TP
d60e92d1
AC
43.BI \-\-cmdhelp \fR=\fPcommand
44Print help information for \fIcommand\fR. May be `all' for all commands.
45.TP
9183788d
JA
46.BI \-\-debug \fR=\fPtype
47Enable verbose tracing of various fio actions. May be `all' for all types
48or individual types seperated by a comma (eg --debug=io,file). `help' will
49list all available tracing options.
50.TP
d60e92d1
AC
51.B \-\-help
52Display usage information and exit.
53.TP
54.B \-\-version
55Display version information and exit.
56.SH "JOB FILE FORMAT"
57Job files are in `ini' format. They consist of one or more
58job definitions, which begin with a job name in square brackets and
59extend to the next job name. The job name can be any ASCII string
60except `global', which has a special meaning. Following the job name is
61a sequence of zero or more parameters, one per line, that define the
62behavior of the job. Any line starting with a `;' or `#' character is
d1429b5c 63considered a comment and ignored.
d9956b64
AC
64.P
65If \fIjobfile\fR is specified as `-', the job file will be read from
66standard input.
d60e92d1
AC
67.SS "Global Section"
68The global section contains default parameters for jobs specified in the
69job file. A job is only affected by global sections residing above it,
70and there may be any number of global sections. Specific job definitions
71may override any parameter set in global sections.
72.SH "JOB PARAMETERS"
73.SS Types
74Some parameters may take arguments of a specific type. The types used are:
75.TP
76.I str
77String: a sequence of alphanumeric characters.
78.TP
79.I int
80Integer: a whole number, possibly negative. If prefixed with `0x', the value
81is assumed to be base 16 (hexadecimal).
82.TP
83.I siint
84SI integer: a whole number, possibly containing a suffix denoting the base unit
85of the value. Accepted suffixes are `k', 'M' and 'G', denoting kilo (1024),
86mega (1024*1024) and giga (1024*1024*1024) respectively.
87.TP
88.I bool
89Boolean: a true or false value. `0' denotes false, `1' denotes true.
90.TP
91.I irange
92Integer range: a range of integers specified in the format
d1429b5c
AC
93\fIlower\fR:\fIupper\fR or \fIlower\fR\-\fIupper\fR. \fIlower\fR and
94\fIupper\fR may contain a suffix as described above. If an option allows two
95sets of ranges, they are separated with a `,' or `/' character. For example:
96`8\-8k/8M\-4G'.
d60e92d1
AC
97.SS "Parameter List"
98.TP
99.BI name \fR=\fPstr
d9956b64 100May be used to override the job name. On the command line, this parameter
d60e92d1
AC
101has the special purpose of signalling the start of a new job.
102.TP
103.BI description \fR=\fPstr
104Human-readable description of the job. It is printed when the job is run, but
105otherwise has no special purpose.
106.TP
107.BI directory \fR=\fPstr
108Prefix filenames with this directory. Used to place files in a location other
109than `./'.
110.TP
111.BI filename \fR=\fPstr
112.B fio
113normally makes up a file name based on the job name, thread number, and file
d1429b5c
AC
114number. If you want to share files between threads in a job or several jobs,
115specify a \fIfilename\fR for each of them to override the default. If the I/O
116engine used is `net', \fIfilename\fR is the host and port to connect to in the
117format \fIhost\fR/\fIport\fR. If the I/O engine is file-based, you can specify
118a number of files by separating the names with a `:' character. `\-' is a
119reserved name, meaning stdin or stdout, depending on the read/write direction
120set.
d60e92d1
AC
121.TP
122.BI opendir \fR=\fPstr
123Recursively open any files below directory \fIstr\fR.
124.TP
125.BI readwrite \fR=\fPstr "\fR,\fP rw" \fR=\fPstr
126Type of I/O pattern. Accepted values are:
127.RS
128.RS
129.TP
130.B read
d1429b5c 131Sequential reads.
d60e92d1
AC
132.TP
133.B write
d1429b5c 134Sequential writes.
d60e92d1
AC
135.TP
136.B randread
d1429b5c 137Random reads.
d60e92d1
AC
138.TP
139.B randwrite
d1429b5c 140Random writes.
d60e92d1
AC
141.TP
142.B rw
d1429b5c 143Mixed sequential reads and writes.
d60e92d1
AC
144.TP
145.B randrw
d1429b5c 146Mixed random reads and writes.
d60e92d1
AC
147.RE
148.P
d1429b5c
AC
149For mixed I/O, the default split is 50/50. For random I/O, the number of I/Os
150to perform before getting a new offset can be specified by appending
151`:\fIint\fR' to the pattern type. The default is 1.
d60e92d1
AC
152.RE
153.TP
154.BI randrepeat \fR=\fPbool
155Seed the random number generator in a predictable way so results are repeatable
d1429b5c 156across runs. Default: true.
d60e92d1
AC
157.TP
158.BI fadvise_hint \fR=\fPbool
d1429b5c
AC
159Disable use of \fIposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
160are likely to be issued. Default: true.
d60e92d1
AC
161.TP
162.BI size \fR=\fPsiint
163Total size of I/O for this job. \fBfio\fR will run until this many bytes have
164been transfered, unless limited by other options (\fBruntime\fR, for instance).
165Unless \fBnr_files\fR and \fBfilesize\fR options are given, this amount will be
166divided between the available files for the job.
167.TP
168.BI filesize \fR=\fPirange
169Individual file sizes. May be a range, in which case \fBfio\fR will select sizes
d1429b5c
AC
170for files at random within the given range, limited to \fBsize\fR in total (if
171that is given). If \fBfilesize\fR is not specified, each created file is the
172same size.
d60e92d1 173.TP
9183788d 174.BI blocksize \fR=\fPsiint[,siint] "\fR,\fB bs" \fR=\fPsiint[,siint]
d60e92d1
AC
175Block size for I/O units. Default: 4k. Values for reads and writes can be
176specified seperately in the format \fIread\fR,\fIwrite\fR, either of
177which may be empty to leave that value at its default.
178.TP
9183788d 179.BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange]
d1429b5c
AC
180Specify a range of I/O block sizes. The issued I/O unit will always be a
181multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies
9183788d
JA
182to both reads and writes if only one range is given, but can be specified
183seperately with a comma seperating the values. Example: bsrange=1k-4k,2k-8k.
184Also (see \fBblocksize\fR).
185.TP
186.BI bssplit \fR=\fPstr
187This option allows even finer grained control of the block sizes issued,
188not just even splits between them. With this option, you can weight various
189block sizes for exact control of the issued IO for a job that has mixed
190block sizes. The format of the option is bssplit=blocksize/percentage,
191optionally adding as many definitions as needed seperated by a colon.
192Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k
193blocks and 40% 32k blocks.
d60e92d1
AC
194.TP
195.B blocksize_unaligned\fR,\fP bs_unaligned
d1429b5c
AC
196If set, any size in \fBblocksize_range\fR may be used. This typically won't
197work with direct I/O, as that normally requires sector alignment.
d60e92d1
AC
198.TP
199.B zero_buffers
200Initialise buffers with all zeros. Default: fill buffers with random data.
201.TP
202.BI nrfiles \fR=\fPint
203Number of files to use for this job. Default: 1.
204.TP
205.BI openfiles \fR=\fPint
206Number of files to keep open at the same time. Default: \fBnrfiles\fR.
207.TP
208.BI file_service_type \fR=\fPstr
209Defines how files to service are selected. The following types are defined:
210.RS
211.RS
212.TP
213.B random
214Choose a file at random
215.TP
216.B roundrobin
217Round robin over open files (default).
218.RE
219.P
220The number of I/Os to issue before switching a new file can be specified by
221appending `:\fIint\fR' to the service type.
222.RE
223.TP
224.BI ioengine \fR=\fPstr
225Defines how the job issues I/O. The following types are defined:
226.RS
227.RS
228.TP
229.B sync
230Basic \fIread\fR\|(2) or \fIwrite\fR\|(2) I/O. \fIfseek\fR\|(2) is used to
231position the I/O location.
232.TP
a31041ea 233.B psync
234Basic \fIpread\fR\|(2) or \fIpwrite\fR\|(2) I/O.
235.TP
9183788d
JA
236.B vsync
237Basic \fIreadv\fR\|(2) or \fIwritev\fR\|(2) I/O. Will emulate queuing by
238coalescing adjacents IOs into a single submission.
239.TP
d60e92d1
AC
240.B libaio
241Linux native asynchronous I/O.
242.TP
243.B posixaio
244glibc POSIX asynchronous I/O using \fIaio_read\fR\|(3) and \fIaio_write\fR\|(3).
245.TP
246.B mmap
d1429b5c
AC
247File is memory mapped with \fImmap\fR\|(2) and data copied using
248\fImemcpy\fR\|(3).
d60e92d1
AC
249.TP
250.B splice
d1429b5c
AC
251\fIsplice\fR\|(2) is used to transfer the data and \fIvmsplice\fR\|(2) to
252transfer data from user-space to the kernel.
d60e92d1
AC
253.TP
254.B syslet-rw
255Use the syslet system calls to make regular read/write asynchronous.
256.TP
257.B sg
258SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if
d1429b5c
AC
259the target is an sg character device, we use \fIread\fR\|(2) and
260\fIwrite\fR\|(2) for asynchronous I/O.
d60e92d1
AC
261.TP
262.B null
263Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR
264itself and for debugging and testing purposes.
265.TP
266.B net
267Transfer over the network. \fBfilename\fR must be set appropriately to
268`\fIhost\fR/\fIport\fR' regardless of data direction. If receiving, only the
269\fIport\fR argument is used.
270.TP
271.B netsplice
272Like \fBnet\fR, but uses \fIsplice\fR\|(2) and \fIvmsplice\fR\|(2) to map data
273and send/receive.
274.TP
53aec0a4 275.B cpuio
d60e92d1
AC
276Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and
277\fBcpucycles\fR parameters.
278.TP
279.B guasi
280The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface
281approach to asycnronous I/O.
d1429b5c
AC
282.br
283See <http://www.xmailserver.org/guasi\-lib.html>.
d60e92d1
AC
284.TP
285.B external
286Loads an external I/O engine object file. Append the engine filename as
287`:\fIenginepath\fR'.
288.RE
289.RE
290.TP
291.BI iodepth \fR=\fPint
292Number of I/O units to keep in flight against the file. Default: 1.
293.TP
294.BI iodepth_batch \fR=\fPint
295Number of I/Os to submit at once. Default: \fBiodepth\fR.
296.TP
297.BI iodepth_low \fR=\fPint
298Low watermark indicating when to start filling the queue again. Default:
299\fBiodepth\fR.
300.TP
301.BI direct \fR=\fPbool
302If true, use non-buffered I/O (usually O_DIRECT). Default: false.
303.TP
304.BI buffered \fR=\fPbool
305If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter.
306Default: true.
307.TP
308.BI offset \fR=\fPsiint
309Offset in the file to start I/O. Data before the offset will not be touched.
310.TP
311.BI fsync \fR=\fPint
d1429b5c
AC
312How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If
3130, don't sync. Default: 0.
d60e92d1
AC
314.TP
315.BI overwrite \fR=\fPbool
d1429b5c 316If writing, setup the file first and do overwrites. Default: false.
d60e92d1
AC
317.TP
318.BI end_fsync \fR=\fPbool
d1429b5c 319Sync file contents when job exits. Default: false.
d60e92d1
AC
320.TP
321.BI fsync_on_close \fR=\fPbool
322If true, sync file contents on close. This differs from \fBend_fsync\fR in that
d1429b5c 323it will happen on every close, not just at the end of the job. Default: false.
d60e92d1
AC
324.TP
325.BI rwmixcycle \fR=\fPint
326How many milliseconds before switching between reads and writes for a mixed
327workload. Default: 500ms.
328.TP
329.BI rwmixread \fR=\fPint
330Percentage of a mixed workload that should be reads. Default: 50.
331.TP
332.BI rwmixwrite \fR=\fPint
d1429b5c 333Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and
d60e92d1 334\fBwrmixwrite\fR are given and do not sum to 100%, the latter of the two
d1429b5c 335overrides the first. Default: 50.
d60e92d1
AC
336.TP
337.B norandommap
338Normally \fBfio\fR will cover every block of the file when doing random I/O. If
339this parameter is given, a new offset will be chosen without looking at past
340I/O history. This parameter is mutually exclusive with \fBverify\fR.
341.TP
342.BI nice \fR=\fPint
343Run job with given nice value. See \fInice\fR\|(2).
344.TP
345.BI prio \fR=\fPint
346Set I/O priority value of this job between 0 (highest) and 7 (lowest). See
347\fIionice\fR\|(1).
348.TP
349.BI prioclass \fR=\fPint
350Set I/O priority class. See \fIionice\fR\|(1).
351.TP
352.BI thinktime \fR=\fPint
353Stall job for given number of microseconds between issuing I/Os.
354.TP
355.BI thinktime_spin \fR=\fPint
356Pretend to spend CPU time for given number of microseconds, sleeping the rest
357of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set.
358.TP
359.BI thinktime_blocks \fR=\fPint
360Number of blocks to issue before waiting \fBthinktime\fR microseconds.
361Default: 1.
362.TP
363.BI rate \fR=\fPint
364Cap bandwidth used by this job to this number of KiB/s.
365.TP
366.BI ratemin \fR=\fPint
367Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth.
368Failing to meet this requirement will cause the job to exit.
369.TP
370.BI rate_iops \fR=\fPint
371Cap the bandwidth to this number of IOPS. If \fBblocksize\fR is a range, the
372smallest block size is used as the metric.
373.TP
374.BI rate_iops_min \fR=\fPint
375If this rate of I/O is not met, the job will exit.
376.TP
377.BI ratecycle \fR=\fPint
378Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of
379milliseconds. Default: 1000ms.
380.TP
381.BI cpumask \fR=\fPint
382Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job
383may run on. See \fBsched_setaffinity\fR\|(2).
384.TP
385.BI cpus_allowed \fR=\fPstr
386Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers.
387.TP
388.BI startdelay \fR=\fPint
389Delay start of job for the specified number of seconds.
390.TP
391.BI runtime \fR=\fPint
392Terminate processing after the specified number of seconds.
393.TP
394.B time_based
395If given, run for the specified \fBruntime\fR duration even if the files are
396completely read or written. The same workload will be repeated as many times
397as \fBruntime\fR allows.
398.TP
399.BI invalidate \fR=\fPbool
400Invalidate buffer-cache for the file prior to starting I/O. Default: true.
401.TP
402.BI sync \fR=\fPbool
403Use synchronous I/O for buffered writes. For the majority of I/O engines,
d1429b5c 404this means using O_SYNC. Default: false.
d60e92d1
AC
405.TP
406.BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr
407Allocation method for I/O unit buffer. Allowed values are:
408.RS
409.RS
410.TP
411.B malloc
412Allocate memory with \fImalloc\fR\|(3).
413.TP
414.B shm
415Use shared memory buffers allocated through \fIshmget\fR\|(2).
416.TP
417.B shmhuge
418Same as \fBshm\fR, but use huge pages as backing.
419.TP
420.B mmap
421Use \fImmap\fR\|(2) for allocation. Uses anonymous memory unless a filename
422is given after the option in the format `:\fIfile\fR'.
423.TP
424.B mmaphuge
425Same as \fBmmap\fR, but use huge files as backing.
426.RE
427.P
428The amount of memory allocated is the maximum allowed \fBblocksize\fR for the
429job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work,
430the system must have free huge pages allocated. \fBmmaphuge\fR also needs to
431have hugetlbfs mounted, and \fIfile\fR must point there.
432.RE
433.TP
d1429b5c 434.BI hugepage\-size \fR=\fPsiint
d60e92d1
AC
435Defines the size of a huge page. Must be at least equal to the system setting.
436Should be a multiple of 1MiB. Default: 4MiB.
437.TP
438.B exitall
439Terminate all jobs when one finishes. Default: wait for each job to finish.
440.TP
441.BI bwavgtime \fR=\fPint
442Average bandwidth calculations over the given time in milliseconds. Default:
443500ms.
444.TP
445.BI create_serialize \fR=\fPbool
d1429b5c 446If true, serialize file creation for the jobs. Default: true.
d60e92d1
AC
447.TP
448.BI create_fsync \fR=\fPbool
449\fIfsync\fR\|(2) data file after creation. Default: true.
450.TP
451.BI unlink \fR=\fPbool
452Unlink job files when done. Default: false.
453.TP
454.BI loops \fR=\fPint
455Specifies the number of iterations (runs of the same workload) of this job.
456Default: 1.
457.TP
458.BI do_verify \fR=\fPbool
459Run the verify phase after a write phase. Only valid if \fBverify\fR is set.
460Default: true.
461.TP
462.BI verify \fR=\fPstr
463Method of verifying file contents after each iteration of the job. Allowed
464values are:
465.RS
466.RS
467.TP
468.B md5 crc16 crc32 crc64 crc7 sha256 sha512
469Store appropriate checksum in the header of each block.
470.TP
471.B meta
472Write extra information about each I/O (timestamp, block number, etc.). The
473block number is verified.
474.TP
475.B pattern
476Fill I/O buffers with a specific pattern that is used to verify. The pattern is
477specified by appending `:\fIint\fR' to the parameter. \fIint\fR cannot be larger
478than 32-bits.
479.TP
480.B null
481Pretend to verify. Used for testing internals.
482.RE
483.RE
484.TP
485.BI verify_sort \fR=\fPbool
486If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to
487read them back in a sorted manner. Default: true.
488.TP
489.BI verify_offset \fR=\fPsiint
490Swap the verification header with data somewhere else in the block before
d1429b5c 491writing. It is swapped back before verifying.
d60e92d1
AC
492.TP
493.BI verify_interval \fR=\fPsiint
494Write the verification header for this number of bytes, which should divide
495\fBblocksize\fR. Default: \fBblocksize\fR.
496.TP
497.BI verify_fatal \fR=\fPbool
498If true, exit the job on the first observed verification failure. Default:
499false.
500.TP
501.B stonewall
d1429b5c 502Wait for preceeding jobs in the job file to exit before starting this one.
d60e92d1
AC
503\fBstonewall\fR implies \fBnew_group\fR.
504.TP
505.B new_group
506Start a new reporting group. If not given, all jobs in a file will be part
507of the same reporting group, unless separated by a stonewall.
508.TP
509.BI numjobs \fR=\fPint
510Number of clones (processes/threads performing the same workload) of this job.
511Default: 1.
512.TP
513.B group_reporting
514If set, display per-group reports instead of per-job when \fBnumjobs\fR is
515specified.
516.TP
517.B thread
518Use threads created with \fBpthread_create\fR\|(3) instead of processes created
519with \fBfork\fR\|(2).
520.TP
521.BI zonesize \fR=\fPsiint
522Divide file into zones of the specified size in bytes. See \fBzoneskip\fR.
523.TP
524.BI zoneskip \fR=\fPsiint
d1429b5c 525Skip the specified number of bytes when \fBzonesize\fR bytes of data have been
d60e92d1
AC
526read.
527.TP
528.BI write_iolog \fR=\fPstr
529Write the issued I/O patterns to the specified file.
530.TP
531.BI read_iolog \fR=\fPstr
532Replay the I/O patterns contained in the specified file generated by
533\fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file.
534.TP
535.B write_bw_log
536If given, write bandwidth logs of the jobs in this file.
537.TP
538.B write_lat_log
539Same as \fBwrite_bw_log\fR, but writes I/O completion latencies.
540.TP
541.BI lockmem \fR=\fPsiint
542Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to
543simulate a smaller amount of memory.
544.TP
545.BI exec_prerun \fR=\fPstr
546Before running the job, execute the specified command with \fBsystem\fR\|(3).
547.TP
548.BI exec_postrun \fR=\fPstr
549Same as \fBexec_prerun\fR, but the command is executed after the job completes.
550.TP
551.BI ioscheduler \fR=\fPstr
552Attempt to switch the device hosting the file to the specified I/O scheduler.
553.TP
554.BI cpuload \fR=\fPint
555If the job is a CPU cycle-eater, attempt to use the specified percentage of
556CPU cycles.
557.TP
558.BI cpuchunks \fR=\fPint
559If the job is a CPU cycle-eater, split the load into cycles of the
560given time in milliseconds.
561.TP
562.BI disk_util \fR=\fPbool
d1429b5c 563Generate disk utilization statistics if the platform supports it. Default: true.
d60e92d1 564.SH OUTPUT
d1429b5c
AC
565While running, \fBfio\fR will display the status of the created jobs. For
566example:
d60e92d1 567.RS
d1429b5c 568.P
d60e92d1
AC
569Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
570.RE
571.P
d1429b5c
AC
572The characters in the first set of brackets denote the current status of each
573threads. The possible values are:
574.P
575.PD 0
d60e92d1
AC
576.RS
577.TP
578.B P
579Setup but not started.
580.TP
581.B C
582Thread created.
583.TP
584.B I
585Initialized, waiting.
586.TP
587.B R
588Running, doing sequential reads.
589.TP
590.B r
591Running, doing random reads.
592.TP
593.B W
594Running, doing sequential writes.
595.TP
596.B w
597Running, doing random writes.
598.TP
599.B M
600Running, doing mixed sequential reads/writes.
601.TP
602.B m
603Running, doing mixed random reads/writes.
604.TP
605.B F
606Running, currently waiting for \fBfsync\fR\|(2).
607.TP
608.B V
609Running, verifying written data.
610.TP
611.B E
612Exited, not reaped by main thread.
613.TP
614.B \-
615Exited, thread reaped.
616.RE
d1429b5c 617.PD
d60e92d1
AC
618.P
619The second set of brackets shows the estimated completion percentage of
620the current group. The third set shows the read and write I/O rate,
621respectively. Finally, the estimated run time of the job is displayed.
622.P
623When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data
624for each thread, each group of threads, and each disk, in that order.
625.P
626Per-thread statistics first show the threads client number, group-id, and
627error code. The remaining figures are as follows:
628.RS
d60e92d1
AC
629.TP
630.B io
631Number of megabytes of I/O performed.
632.TP
633.B bw
634Average data rate (bandwidth).
635.TP
636.B runt
637Threads run time.
638.TP
639.B slat
640Submission latency minimum, maximum, average and standard deviation. This is
641the time it took to submit the I/O.
642.TP
643.B clat
644Completion latency minimum, maximum, average and standard deviation. This
645is the time between submission and completion.
646.TP
647.B bw
648Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average
649and standard deviation.
650.TP
651.B cpu
652CPU usage statistics. Includes user and system time, number of context switches
653this thread went through and number of major and minor page faults.
654.TP
655.B IO depths
656Distribution of I/O depths. Each depth includes everything less than (or equal)
657to it, but greater than the previous depth.
658.TP
659.B IO issued
660Number of read/write requests issued, and number of short read/write requests.
661.TP
662.B IO latencies
663Distribution of I/O completion latencies. The numbers follow the same pattern
664as \fBIO depths\fR.
665.RE
d60e92d1
AC
666.P
667The group statistics show:
d1429b5c 668.PD 0
d60e92d1
AC
669.RS
670.TP
671.B io
672Number of megabytes I/O performed.
673.TP
674.B aggrb
675Aggregate bandwidth of threads in the group.
676.TP
677.B minb
678Minimum average bandwidth a thread saw.
679.TP
680.B maxb
681Maximum average bandwidth a thread saw.
682.TP
683.B mint
d1429b5c 684Shortest runtime of threads in the group.
d60e92d1
AC
685.TP
686.B maxt
687Longest runtime of threads in the group.
688.RE
d1429b5c 689.PD
d60e92d1
AC
690.P
691Finally, disk statistics are printed with reads first:
d1429b5c 692.PD 0
d60e92d1
AC
693.RS
694.TP
695.B ios
696Number of I/Os performed by all groups.
697.TP
698.B merge
699Number of merges in the I/O scheduler.
700.TP
701.B ticks
702Number of ticks we kept the disk busy.
703.TP
704.B io_queue
705Total time spent in the disk queue.
706.TP
707.B util
708Disk utilization.
709.RE
d1429b5c 710.PD
d60e92d1
AC
711.SH TERSE OUTPUT
712If the \fB\-\-minimal\fR option is given, the results will be printed in a
713semicolon-delimited format suitable for scripted use. The fields are:
714.P
715.RS
716.B jobname, groupid, error
717.P
718Read status:
719.RS
720.B KiB I/O, bandwidth \fR(KiB/s)\fP, runtime \fR(ms)\fP
721.P
722Submission latency:
723.RS
724.B min, max, mean, standard deviation
725.RE
726Completion latency:
727.RS
728.B min, max, mean, standard deviation
729.RE
730Bandwidth:
731.RS
732.B min, max, aggregate percentage of total, mean, standard deviation
733.RE
734.RE
735.P
736Write status:
737.RS
738.B KiB I/O, bandwidth \fR(KiB/s)\fP, runtime \fR(ms)\fP
739.P
740Submission latency:
741.RS
742.B min, max, mean, standard deviation
743.RE
744Completion latency:
745.RS
746.B min, max, mean, standard deviation
747.RE
748Bandwidth:
749.RS
750.B min, max, aggregate percentage of total, mean, standard deviation
751.RE
752.RE
753.P
d1429b5c 754CPU usage:
d60e92d1
AC
755.RS
756.B user, system, context switches
757.RE
758.P
759IO depth distribution:
760.RS
761.B <=1, 2, 4, 8, 16, 32, >=64
762.RE
763.P
764IO latency distribution (ms):
765.RS
766.B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, >=2000
767.RE
768.P
769.B text description
770.RE
771.SH AUTHORS
772.B fio
773was written by Jens Axboe <jens.axboe@oracle.com>.
d1429b5c
AC
774.br
775This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based
d60e92d1
AC
776on documentation by Jens Axboe.
777.SH "REPORTING BUGS"
d1429b5c
AC
778Report bugs to the \fBfio\fR mailing list <fio-devel@kernel.dk>.
779See \fBREADME\fR.
d60e92d1 780.SH "SEE ALSO"
d1429b5c
AC
781For further documentation see \fBHOWTO\fR and \fBREADME\fR.
782.br
783Sample jobfiles are available in the \fBexamples\fR directory.
d60e92d1 784