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 certain types of io the result
176 may still be skewed a bit, since the speed may be different. It is possible to
177 specify a number of IO's to do before getting a new offset, this is one by
178 appending a `:\fI<nr>\fR to the end of the string given. For a random read, it
179 would look like \fBrw=randread:8\fR for passing in an offset modifier with a
180 value of 8. See the \fBrw_sequencer\fR option.
183 .BI rw_sequencer \fR=\fPstr
184 If an offset modifier is given by appending a number to the \fBrw=<str>\fR line,
185 then this option controls how that number modifies the IO offset being
186 generated. Accepted values are:
191 Generate sequential offset
194 Generate the same offset
197 \fBsequential\fR is only useful for random IO, where fio would normally
198 generate a new random offset for every IO. If you append eg 8 to randread, you
199 would get a new random offset for every 8 IO's. The result would be a seek for
200 only every 8 IO's, instead of for every IO. Use \fBrw=randread:8\fR to specify
201 that. As sequential IO is already sequential, setting \fBsequential\fR for that
202 would not result in any differences. \fBidentical\fR behaves in a similar
203 fashion, except it sends the same offset 8 number of times before generating a
208 .BI kb_base \fR=\fPint
209 The base unit for a kilobyte. The defacto base is 2^10, 1024. Storage
210 manufacturers like to use 10^3 or 1000 as a base ten unit instead, for obvious
211 reasons. Allow values are 1024 or 1000, with 1024 being the default.
213 .BI randrepeat \fR=\fPbool
214 Seed the random number generator in a predictable way so results are repeatable
215 across runs. Default: true.
217 .BI use_os_rand \fR=\fPbool
218 Fio can either use the random generator supplied by the OS to generator random
219 offsets, or it can use it's own internal generator (based on Tausworthe).
220 Default is to use the internal generator, which is often of better quality and
221 faster. Default: false.
223 .BI fallocate \fR=\fPbool
224 By default, fio will use fallocate() to advise the system of the size of the
225 file we are going to write. This can be turned off with fallocate=0. May not
226 be available on all supported platforms.
228 .BI fadvise_hint \fR=\fPbool
229 Disable use of \fIposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
230 are likely to be issued. Default: true.
233 Total size of I/O for this job. \fBfio\fR will run until this many bytes have
234 been transfered, unless limited by other options (\fBruntime\fR, for instance).
235 Unless \fBnrfiles\fR and \fBfilesize\fR options are given, this amount will be
236 divided between the available files for the job. If not set, fio will use the
237 full size of the given files or devices. If the the files do not exist, size
240 .BI fill_device \fR=\fPbool "\fR,\fB fill_fs" \fR=\fPbool
241 Sets size to something really large and waits for ENOSPC (no space left on
242 device) as the terminating condition. Only makes sense with sequential write.
243 For a read workload, the mount point will be filled first then IO started on
244 the result. This option doesn't make sense if operating on a raw device node,
245 since the size of that is already known by the file system. Additionally,
246 writing beyond end-of-device will not return ENOSPC there.
248 .BI filesize \fR=\fPirange
249 Individual file sizes. May be a range, in which case \fBfio\fR will select sizes
250 for files at random within the given range, limited to \fBsize\fR in total (if
251 that is given). If \fBfilesize\fR is not specified, each created file is the
254 .BI blocksize \fR=\fPint[,int] "\fR,\fB bs" \fR=\fPint[,int]
255 Block size for I/O units. Default: 4k. Values for reads and writes can be
256 specified separately in the format \fIread\fR,\fIwrite\fR, either of
257 which may be empty to leave that value at its default.
259 .BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange]
260 Specify a range of I/O block sizes. The issued I/O unit will always be a
261 multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies
262 to both reads and writes if only one range is given, but can be specified
263 separately with a comma seperating the values. Example: bsrange=1k-4k,2k-8k.
264 Also (see \fBblocksize\fR).
266 .BI bssplit \fR=\fPstr
267 This option allows even finer grained control of the block sizes issued,
268 not just even splits between them. With this option, you can weight various
269 block sizes for exact control of the issued IO for a job that has mixed
270 block sizes. The format of the option is bssplit=blocksize/percentage,
271 optionally adding as many definitions as needed seperated by a colon.
272 Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k
273 blocks and 40% 32k blocks. \fBbssplit\fR also supports giving separate
274 splits to reads and writes. The format is identical to what the
275 \fBbs\fR option accepts, the read and write parts are separated with a
278 .B blocksize_unaligned\fR,\fP bs_unaligned
279 If set, any size in \fBblocksize_range\fR may be used. This typically won't
280 work with direct I/O, as that normally requires sector alignment.
282 .BI blockalign \fR=\fPint[,int] "\fR,\fB ba" \fR=\fPint[,int]
283 At what boundary to align random IO offsets. Defaults to the same as 'blocksize'
284 the minimum blocksize given. Minimum alignment is typically 512b
285 for using direct IO, though it usually depends on the hardware block size.
286 This option is mutually exclusive with using a random map for files, so it
287 will turn off that option.
290 Initialise buffers with all zeros. Default: fill buffers with random data.
293 If this option is given, fio will refill the IO buffers on every submit. The
294 default is to only fill it at init time and reuse that data. Only makes sense
295 if zero_buffers isn't specified, naturally. If data verification is enabled,
296 refill_buffers is also automatically enabled.
298 .BI nrfiles \fR=\fPint
299 Number of files to use for this job. Default: 1.
301 .BI openfiles \fR=\fPint
302 Number of files to keep open at the same time. Default: \fBnrfiles\fR.
304 .BI file_service_type \fR=\fPstr
305 Defines how files to service are selected. The following types are defined:
310 Choose a file at random
313 Round robin over open files (default).
315 Do each file in the set sequentially.
318 The number of I/Os to issue before switching a new file can be specified by
319 appending `:\fIint\fR' to the service type.
322 .BI ioengine \fR=\fPstr
323 Defines how the job issues I/O. The following types are defined:
328 Basic \fIread\fR\|(2) or \fIwrite\fR\|(2) I/O. \fIfseek\fR\|(2) is used to
329 position the I/O location.
332 Basic \fIpread\fR\|(2) or \fIpwrite\fR\|(2) I/O.
335 Basic \fIreadv\fR\|(2) or \fIwritev\fR\|(2) I/O. Will emulate queuing by
336 coalescing adjacents IOs into a single submission.
339 Linux native asynchronous I/O.
342 POSIX asynchronous I/O using \fIaio_read\fR\|(3) and \fIaio_write\fR\|(3).
345 Solaris native asynchronous I/O.
348 Windows native asynchronous I/O.
351 File is memory mapped with \fImmap\fR\|(2) and data copied using
355 \fIsplice\fR\|(2) is used to transfer the data and \fIvmsplice\fR\|(2) to
356 transfer data from user-space to the kernel.
359 Use the syslet system calls to make regular read/write asynchronous.
362 SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if
363 the target is an sg character device, we use \fIread\fR\|(2) and
364 \fIwrite\fR\|(2) for asynchronous I/O.
367 Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR
368 itself and for debugging and testing purposes.
371 Transfer over the network. \fBfilename\fR must be set appropriately to
372 `\fIhost\fR/\fIport\fR' regardless of data direction. If receiving, only the
373 \fIport\fR argument is used.
376 Like \fBnet\fR, but uses \fIsplice\fR\|(2) and \fIvmsplice\fR\|(2) to map data
380 Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and
381 \fBcpucycles\fR parameters.
384 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface
385 approach to asycnronous I/O.
387 See <http://www.xmailserver.org/guasi\-lib.html>.
390 Loads an external I/O engine object file. Append the engine filename as
395 .BI iodepth \fR=\fPint
396 Number of I/O units to keep in flight against the file. Note that increasing
397 iodepth beyond 1 will not affect synchronous ioengines (except for small
398 degress when verify_async is in use). Even async engines my impose OS
399 restrictions causing the desired depth not to be achieved. This may happen on
400 Linux when using libaio and not setting \fBdirect\fR=1, since buffered IO is
401 not async on that OS. Keep an eye on the IO depth distribution in the
402 fio output to verify that the achieved depth is as expected. Default: 1.
404 .BI iodepth_batch \fR=\fPint
405 Number of I/Os to submit at once. Default: \fBiodepth\fR.
407 .BI iodepth_batch_complete \fR=\fPint
408 This defines how many pieces of IO to retrieve at once. It defaults to 1 which
409 means that we'll ask for a minimum of 1 IO in the retrieval process from the
410 kernel. The IO retrieval will go on until we hit the limit set by
411 \fBiodepth_low\fR. If this variable is set to 0, then fio will always check for
412 completed events before queuing more IO. This helps reduce IO latency, at the
413 cost of more retrieval system calls.
415 .BI iodepth_low \fR=\fPint
416 Low watermark indicating when to start filling the queue again. Default:
419 .BI direct \fR=\fPbool
420 If true, use non-buffered I/O (usually O_DIRECT). Default: false.
422 .BI buffered \fR=\fPbool
423 If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter.
426 .BI offset \fR=\fPint
427 Offset in the file to start I/O. Data before the offset will not be touched.
430 How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If
431 0, don't sync. Default: 0.
433 .BI fdatasync \fR=\fPint
434 Like \fBfsync\fR, but uses \fBfdatasync\fR\|(2) instead to only sync the
435 data parts of the file. Default: 0.
437 .BI sync_file_range \fR=\fPstr:int
438 Use sync_file_range() for every \fRval\fP number of write operations. Fio will
439 track range of writes that have happened since the last sync_file_range() call.
440 \fRstr\fP can currently be one or more of:
444 SYNC_FILE_RANGE_WAIT_BEFORE
447 SYNC_FILE_RANGE_WRITE
450 SYNC_FILE_RANGE_WRITE
454 So if you do sync_file_range=wait_before,write:8, fio would use
455 \fBSYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE\fP for every 8 writes.
456 Also see the sync_file_range(2) man page. This option is Linux specific.
458 .BI overwrite \fR=\fPbool
459 If writing, setup the file first and do overwrites. Default: false.
461 .BI end_fsync \fR=\fPbool
462 Sync file contents when job exits. Default: false.
464 .BI fsync_on_close \fR=\fPbool
465 If true, sync file contents on close. This differs from \fBend_fsync\fR in that
466 it will happen on every close, not just at the end of the job. Default: false.
468 .BI rwmixcycle \fR=\fPint
469 How many milliseconds before switching between reads and writes for a mixed
470 workload. Default: 500ms.
472 .BI rwmixread \fR=\fPint
473 Percentage of a mixed workload that should be reads. Default: 50.
475 .BI rwmixwrite \fR=\fPint
476 Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and
477 \fBrwmixwrite\fR are given and do not sum to 100%, the latter of the two
478 overrides the first. This may interfere with a given rate setting, if fio is
479 asked to limit reads or writes to a certain rate. If that is the case, then
480 the distribution may be skewed. Default: 50.
483 Normally \fBfio\fR will cover every block of the file when doing random I/O. If
484 this parameter is given, a new offset will be chosen without looking at past
485 I/O history. This parameter is mutually exclusive with \fBverify\fR.
488 See \fBnorandommap\fR. If fio runs with the random block map enabled and it
489 fails to allocate the map, if this option is set it will continue without a
490 random block map. As coverage will not be as complete as with random maps, this
491 option is disabled by default.
494 Run job with given nice value. See \fInice\fR\|(2).
497 Set I/O priority value of this job between 0 (highest) and 7 (lowest). See
500 .BI prioclass \fR=\fPint
501 Set I/O priority class. See \fIionice\fR\|(1).
503 .BI thinktime \fR=\fPint
504 Stall job for given number of microseconds between issuing I/Os.
506 .BI thinktime_spin \fR=\fPint
507 Pretend to spend CPU time for given number of microseconds, sleeping the rest
508 of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set.
510 .BI thinktime_blocks \fR=\fPint
511 Number of blocks to issue before waiting \fBthinktime\fR microseconds.
515 Cap bandwidth used by this job. The number is in bytes/sec, the normal postfix
516 rules apply. You can use \fBrate\fR=500k to limit reads and writes to 500k each,
517 or you can specify read and writes separately. Using \fBrate\fR=1m,500k would
518 limit reads to 1MB/sec and writes to 500KB/sec. Capping only reads or writes
519 can be done with \fBrate\fR=,500k or \fBrate\fR=500k,. The former will only
520 limit writes (to 500KB/sec), the latter will only limit reads.
522 .BI ratemin \fR=\fPint
523 Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth.
524 Failing to meet this requirement will cause the job to exit. The same format
525 as \fBrate\fR is used for read vs write separation.
527 .BI rate_iops \fR=\fPint
528 Cap the bandwidth to this number of IOPS. Basically the same as rate, just
529 specified independently of bandwidth. The same format as \fBrate\fR is used for
530 read vs write seperation. If \fBblocksize\fR is a range, the smallest block
531 size is used as the metric.
533 .BI rate_iops_min \fR=\fPint
534 If this rate of I/O is not met, the job will exit. The same format as \fBrate\fR
535 is used for read vs write seperation.
537 .BI ratecycle \fR=\fPint
538 Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of
539 milliseconds. Default: 1000ms.
541 .BI cpumask \fR=\fPint
542 Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job
543 may run on. See \fBsched_setaffinity\fR\|(2).
545 .BI cpus_allowed \fR=\fPstr
546 Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers.
548 .BI startdelay \fR=\fPint
549 Delay start of job for the specified number of seconds.
551 .BI runtime \fR=\fPint
552 Terminate processing after the specified number of seconds.
555 If given, run for the specified \fBruntime\fR duration even if the files are
556 completely read or written. The same workload will be repeated as many times
557 as \fBruntime\fR allows.
559 .BI ramp_time \fR=\fPint
560 If set, fio will run the specified workload for this amount of time before
561 logging any performance numbers. Useful for letting performance settle before
562 logging results, thus minimizing the runtime required for stable results. Note
563 that the \fBramp_time\fR is considered lead in time for a job, thus it will
564 increase the total runtime if a special timeout or runtime is specified.
566 .BI invalidate \fR=\fPbool
567 Invalidate buffer-cache for the file prior to starting I/O. Default: true.
570 Use synchronous I/O for buffered writes. For the majority of I/O engines,
571 this means using O_SYNC. Default: false.
573 .BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr
574 Allocation method for I/O unit buffer. Allowed values are:
579 Allocate memory with \fImalloc\fR\|(3).
582 Use shared memory buffers allocated through \fIshmget\fR\|(2).
585 Same as \fBshm\fR, but use huge pages as backing.
588 Use \fImmap\fR\|(2) for allocation. Uses anonymous memory unless a filename
589 is given after the option in the format `:\fIfile\fR'.
592 Same as \fBmmap\fR, but use huge files as backing.
595 The amount of memory allocated is the maximum allowed \fBblocksize\fR for the
596 job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work,
597 the system must have free huge pages allocated. \fBmmaphuge\fR also needs to
598 have hugetlbfs mounted, and \fIfile\fR must point there. At least on Linux,
599 huge pages must be manually allocated. See \fB/proc/sys/vm/nr_hugehages\fR
600 and the documentation for that. Normally you just need to echo an appropriate
601 number, eg echoing 8 will ensure that the OS has 8 huge pages ready for
605 .BI iomem_align \fR=\fPint
606 This indiciates the memory alignment of the IO memory buffers. Note that the
607 given alignment is applied to the first IO unit buffer, if using \fBiodepth\fR
608 the alignment of the following buffers are given by the \fBbs\fR used. In
609 other words, if using a \fBbs\fR that is a multiple of the page sized in the
610 system, all buffers will be aligned to this value. If using a \fBbs\fR that
611 is not page aligned, the alignment of subsequent IO memory buffers is the
612 sum of the \fBiomem_align\fR and \fBbs\fR used.
614 .BI hugepage\-size \fR=\fPint
615 Defines the size of a huge page. Must be at least equal to the system setting.
616 Should be a multiple of 1MB. Default: 4MB.
619 Terminate all jobs when one finishes. Default: wait for each job to finish.
621 .BI bwavgtime \fR=\fPint
622 Average bandwidth calculations over the given time in milliseconds. Default:
625 .BI create_serialize \fR=\fPbool
626 If true, serialize file creation for the jobs. Default: true.
628 .BI create_fsync \fR=\fPbool
629 \fIfsync\fR\|(2) data file after creation. Default: true.
631 .BI create_on_open \fR=\fPbool
632 If true, the files are not created until they are opened for IO by the job.
634 .BI pre_read \fR=\fPbool
635 If this is given, files will be pre-read into memory before starting the given
636 IO operation. This will also clear the \fR \fBinvalidate\fR flag, since it is
637 pointless to pre-read and then drop the cache. This will only work for IO
638 engines that are seekable, since they allow you to read the same data
639 multiple times. Thus it will not work on eg network or splice IO.
641 .BI unlink \fR=\fPbool
642 Unlink job files when done. Default: false.
645 Specifies the number of iterations (runs of the same workload) of this job.
648 .BI do_verify \fR=\fPbool
649 Run the verify phase after a write phase. Only valid if \fBverify\fR is set.
652 .BI verify \fR=\fPstr
653 Method of verifying file contents after each iteration of the job. Allowed
658 .B md5 crc16 crc32 crc32c crc32c-intel crc64 crc7 sha256 sha512 sha1
659 Store appropriate checksum in the header of each block. crc32c-intel is
660 hardware accelerated SSE4.2 driven, falls back to regular crc32c if
661 not supported by the system.
664 Write extra information about each I/O (timestamp, block number, etc.). The
665 block number is verified. See \fBverify_pattern\fR as well.
668 Pretend to verify. Used for testing internals.
671 This option can be used for repeated burn-in tests of a system to make sure
672 that the written data is also correctly read back. If the data direction given
673 is a read or random read, fio will assume that it should verify a previously
674 written file. If the data direction includes any form of write, the verify will
675 be of the newly written data.
678 .BI verify_sort \fR=\fPbool
679 If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to
680 read them back in a sorted manner. Default: true.
682 .BI verify_offset \fR=\fPint
683 Swap the verification header with data somewhere else in the block before
684 writing. It is swapped back before verifying.
686 .BI verify_interval \fR=\fPint
687 Write the verification header for this number of bytes, which should divide
688 \fBblocksize\fR. Default: \fBblocksize\fR.
690 .BI verify_pattern \fR=\fPstr
691 If set, fio will fill the io buffers with this pattern. Fio defaults to filling
692 with totally random bytes, but sometimes it's interesting to fill with a known
693 pattern for io verification purposes. Depending on the width of the pattern,
694 fio will fill 1/2/3/4 bytes of the buffer at the time(it can be either a
695 decimal or a hex number). The verify_pattern if larger than a 32-bit quantity
696 has to be a hex number that starts with either "0x" or "0X". Use with
699 .BI verify_fatal \fR=\fPbool
700 If true, exit the job on the first observed verification failure. Default:
703 .BI verify_dump \fR=\fPbool
704 If set, dump the contents of both the original data block and the data block we
705 read off disk to files. This allows later analysis to inspect just what kind of
706 data corruption occurred. On by default.
708 .BI verify_async \fR=\fPint
709 Fio will normally verify IO inline from the submitting thread. This option
710 takes an integer describing how many async offload threads to create for IO
711 verification instead, causing fio to offload the duty of verifying IO contents
712 to one or more separate threads. If using this offload option, even sync IO
713 engines can benefit from using an \fBiodepth\fR setting higher than 1, as it
714 allows them to have IO in flight while verifies are running.
716 .BI verify_async_cpus \fR=\fPstr
717 Tell fio to set the given CPU affinity on the async IO verification threads.
718 See \fBcpus_allowed\fP for the format used.
720 .BI verify_backlog \fR=\fPint
721 Fio will normally verify the written contents of a job that utilizes verify
722 once that job has completed. In other words, everything is written then
723 everything is read back and verified. You may want to verify continually
724 instead for a variety of reasons. Fio stores the meta data associated with an
725 IO block in memory, so for large verify workloads, quite a bit of memory would
726 be used up holding this meta data. If this option is enabled, fio will write
727 only N blocks before verifying these blocks.
729 .BI verify_backlog_batch \fR=\fPint
730 Control how many blocks fio will verify if verify_backlog is set. If not set,
731 will default to the value of \fBverify_backlog\fR (meaning the entire queue is
732 read back and verified). If \fBverify_backlog_batch\fR is less than
733 \fBverify_backlog\fR then not all blocks will be verified, if
734 \fBverify_backlog_batch\fR is larger than \fBverify_backlog\fR, some blocks
735 will be verified more than once.
738 Wait for preceeding jobs in the job file to exit before starting this one.
739 \fBstonewall\fR implies \fBnew_group\fR.
742 Start a new reporting group. If not given, all jobs in a file will be part
743 of the same reporting group, unless separated by a stonewall.
745 .BI numjobs \fR=\fPint
746 Number of clones (processes/threads performing the same workload) of this job.
750 If set, display per-group reports instead of per-job when \fBnumjobs\fR is
754 Use threads created with \fBpthread_create\fR\|(3) instead of processes created
755 with \fBfork\fR\|(2).
757 .BI zonesize \fR=\fPint
758 Divide file into zones of the specified size in bytes. See \fBzoneskip\fR.
760 .BI zoneskip \fR=\fPint
761 Skip the specified number of bytes when \fBzonesize\fR bytes of data have been
764 .BI write_iolog \fR=\fPstr
765 Write the issued I/O patterns to the specified file. Specify a separate file
766 for each job, otherwise the iologs will be interspersed and the file may be
769 .BI read_iolog \fR=\fPstr
770 Replay the I/O patterns contained in the specified file generated by
771 \fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file.
773 .BI replay_no_stall \fR=\fPint
774 While replaying I/O patterns using \fBread_iolog\fR the default behavior
775 attempts to respect timing information between I/Os. Enabling
776 \fBreplay_no_stall\fR causes I/Os to be replayed as fast as possible while
777 still respecting ordering.
779 .BI replay_redirect \fR=\fPstr
780 While replaying I/O patterns using \fBread_iolog\fR the default behavior
781 is to replay the IOPS onto the major/minor device that each IOP was recorded
782 from. Setting \fBreplay_redirect\fR causes all IOPS to be replayed onto the
783 single specified device regardless of the device it was recorded from.
785 .B write_bw_log \fR=\fPstr
786 If given, write a bandwidth log of the jobs in this job file. Can be used to
787 store data of the bandwidth of the jobs in their lifetime. The included
788 fio_generate_plots script uses gnuplot to turn these text files into nice
789 graphs. See \fBwrite_log_log\fR for behaviour of given filename. For this
790 option, the postfix is _bw.log.
792 .B write_lat_log \fR=\fPstr
793 Same as \fBwrite_bw_log\fR, but writes I/O completion latencies. If no
794 filename is given with this option, the default filename of "jobname_type.log"
795 is used. Even if the filename is given, fio will still append the type of log.
797 .B disable_lat \fR=\fPbool
798 Disable measurements of total latency numbers. Useful only for cutting
799 back the number of calls to gettimeofday, as that does impact performance at
800 really high IOPS rates. Note that to really get rid of a large amount of these
801 calls, this option must be used with disable_slat and disable_bw as well.
803 .B disable_clat \fR=\fPbool
804 Disable measurements of submission latency numbers. See \fBdisable_lat\fR.
806 .B disable_slat \fR=\fPbool
807 Disable measurements of submission latency numbers. See \fBdisable_lat\fR.
809 .B disable_bw_measurement \fR=\fPbool
810 Disable measurements of throughput/bandwidth numbers. See \fBdisable_lat\fR.
812 .BI lockmem \fR=\fPint
813 Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to
814 simulate a smaller amount of memory.
816 .BI exec_prerun \fR=\fPstr
817 Before running the job, execute the specified command with \fBsystem\fR\|(3).
819 .BI exec_postrun \fR=\fPstr
820 Same as \fBexec_prerun\fR, but the command is executed after the job completes.
822 .BI ioscheduler \fR=\fPstr
823 Attempt to switch the device hosting the file to the specified I/O scheduler.
825 .BI cpuload \fR=\fPint
826 If the job is a CPU cycle-eater, attempt to use the specified percentage of
829 .BI cpuchunks \fR=\fPint
830 If the job is a CPU cycle-eater, split the load into cycles of the
831 given time in milliseconds.
833 .BI disk_util \fR=\fPbool
834 Generate disk utilization statistics if the platform supports it. Default: true.
836 .BI gtod_reduce \fR=\fPbool
837 Enable all of the gettimeofday() reducing options (disable_clat, disable_slat,
838 disable_bw) plus reduce precision of the timeout somewhat to really shrink the
839 gettimeofday() call count. With this option enabled, we only do about 0.4% of
840 the gtod() calls we would have done if all time keeping was enabled.
842 .BI gtod_cpu \fR=\fPint
843 Sometimes it's cheaper to dedicate a single thread of execution to just getting
844 the current time. Fio (and databases, for instance) are very intensive on
845 gettimeofday() calls. With this option, you can set one CPU aside for doing
846 nothing but logging current time to a shared memory location. Then the other
847 threads/processes that run IO workloads need only copy that segment, instead of
848 entering the kernel with a gettimeofday() call. The CPU set aside for doing
849 these time calls will be excluded from other uses. Fio will manually clear it
850 from the CPU mask of other jobs.
852 .BI cgroup \fR=\fPstr
853 Add job to this control group. If it doesn't exist, it will be created.
854 The system must have a mounted cgroup blkio mount point for this to work. If
855 your system doesn't have it mounted, you can do so with:
857 # mount -t cgroup -o blkio none /cgroup
859 .BI cgroup_weight \fR=\fPint
860 Set the weight of the cgroup to this value. See the documentation that comes
861 with the kernel, allowed values are in the range of 100..1000.
863 .BI cgroup_nodelete \fR=\fPbool
864 Normally fio will delete the cgroups it has created after the job completion.
865 To override this behavior and to leave cgroups around after the job completion,
866 set cgroup_nodelete=1. This can be useful if one wants to inspect various
867 cgroup files after job completion. Default: false
870 Instead of running as the invoking user, set the user ID to this value before
871 the thread/process does any work.
874 Set group ID, see \fBuid\fR.
876 While running, \fBfio\fR will display the status of the created jobs. For
880 Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
883 The characters in the first set of brackets denote the current status of each
884 threads. The possible values are:
890 Setup but not started.
896 Initialized, waiting.
899 Running, doing sequential reads.
902 Running, doing random reads.
905 Running, doing sequential writes.
908 Running, doing random writes.
911 Running, doing mixed sequential reads/writes.
914 Running, doing mixed random reads/writes.
917 Running, currently waiting for \fBfsync\fR\|(2).
920 Running, verifying written data.
923 Exited, not reaped by main thread.
926 Exited, thread reaped.
930 The second set of brackets shows the estimated completion percentage of
931 the current group. The third set shows the read and write I/O rate,
932 respectively. Finally, the estimated run time of the job is displayed.
934 When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data
935 for each thread, each group of threads, and each disk, in that order.
937 Per-thread statistics first show the threads client number, group-id, and
938 error code. The remaining figures are as follows:
942 Number of megabytes of I/O performed.
945 Average data rate (bandwidth).
951 Submission latency minimum, maximum, average and standard deviation. This is
952 the time it took to submit the I/O.
955 Completion latency minimum, maximum, average and standard deviation. This
956 is the time between submission and completion.
959 Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average
960 and standard deviation.
963 CPU usage statistics. Includes user and system time, number of context switches
964 this thread went through and number of major and minor page faults.
967 Distribution of I/O depths. Each depth includes everything less than (or equal)
968 to it, but greater than the previous depth.
971 Number of read/write requests issued, and number of short read/write requests.
974 Distribution of I/O completion latencies. The numbers follow the same pattern
978 The group statistics show:
983 Number of megabytes I/O performed.
986 Aggregate bandwidth of threads in the group.
989 Minimum average bandwidth a thread saw.
992 Maximum average bandwidth a thread saw.
995 Shortest runtime of threads in the group.
998 Longest runtime of threads in the group.
1002 Finally, disk statistics are printed with reads first:
1007 Number of I/Os performed by all groups.
1010 Number of merges in the I/O scheduler.
1013 Number of ticks we kept the disk busy.
1016 Total time spent in the disk queue.
1023 If the \fB\-\-minimal\fR option is given, the results will be printed in a
1024 semicolon-delimited format suitable for scripted use - a job description
1025 (if provided) follows on a new line. Note that the first
1026 number in the line is the version number. If the output has to be changed
1027 for some reason, this number will be incremented by 1 to signify that
1028 change. The fields are:
1031 .B version, jobname, groupid, error
1035 .B KB I/O, bandwidth \fR(KB/s)\fP, runtime \fR(ms)\fP
1039 .B min, max, mean, standard deviation
1043 .B min, max, mean, standard deviation
1047 .B min, max, mean, standard deviation
1051 .B min, max, aggregate percentage of total, mean, standard deviation
1057 .B KB I/O, bandwidth \fR(KB/s)\fP, runtime \fR(ms)\fP
1061 .B min, max, mean, standard deviation
1065 .B min, max, mean, standard deviation
1069 .B min, max, mean, standard deviation
1073 .B min, max, aggregate percentage of total, mean, standard deviation
1079 .B user, system, context switches, major page faults, minor page faults
1082 IO depth distribution:
1084 .B <=1, 2, 4, 8, 16, 32, >=64
1087 IO latency distribution:
1091 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
1095 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
1099 Error Info (dependant on continue_on_error, default off):
1101 .B total # errors, first error code
1104 .B text description (if provided in config - appears on newline)
1108 was written by Jens Axboe <jens.axboe@oracle.com>,
1109 now Jens Axboe <jaxboe@fusionio.com>.
1111 This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based
1112 on documentation by Jens Axboe.
1113 .SH "REPORTING BUGS"
1114 Report bugs to the \fBfio\fR mailing list <fio@vger.kernel.org>.
1117 For further documentation see \fBHOWTO\fR and \fBREADME\fR.
1119 Sample jobfiles are available in the \fBexamples\fR directory.
projects / fio.git / blob