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 \-\-debug \fR=\fPtype
16 Enable verbose tracing of various fio actions. May be `all' for all types
17 or individual types separated by a comma (eg \-\-debug=io,file). `help' will
18 list all available tracing options.
20 .BI \-\-output \fR=\fPfilename
21 Write output to \fIfilename\fR.
23 .BI \-\-runtime \fR=\fPruntime
24 Limit run time to \fIruntime\fR seconds.
27 Generate per-job latency logs.
30 Generate per-job bandwidth logs.
33 Print statistics in a terse, semicolon-delimited format.
36 Display version information and exit.
38 .BI \-\-terse\-version \fR=\fPversion
39 Set terse version output format (Current version 3, or older version 2).
42 Display usage information and exit.
44 .BI \-\-cmdhelp \fR=\fPcommand
45 Print help information for \fIcommand\fR. May be `all' for all commands.
47 .BI \-\-enghelp \fR=\fPioengine[,command]
48 List all commands defined by \fIioengine\fR, or print help for \fIcommand\fR defined by \fIioengine\fR.
50 .BI \-\-showcmd \fR=\fPjobfile
51 Convert \fIjobfile\fR to a set of command-line options.
53 .BI \-\-eta \fR=\fPwhen
54 Specifies when real-time ETA estimate should be printed. \fIwhen\fR may
55 be one of `always', `never' or `auto'.
58 Turn on safety read-only checks, preventing any attempted write.
60 .BI \-\-section \fR=\fPsec
61 Only run section \fIsec\fR from job file. Multiple of these options can be given, adding more sections to run.
63 .BI \-\-alloc\-size \fR=\fPkb
64 Set the internal smalloc pool size to \fIkb\fP kilobytes.
66 .BI \-\-warnings\-fatal
67 All fio parser warnings are fatal, causing fio to exit with an error.
69 .BI \-\-max\-jobs \fR=\fPnr
70 Set the maximum allowed number of jobs (threads/processes) to support.
72 .BI \-\-server \fR=\fPargs
73 Start a backend server, with \fIargs\fP specifying what to listen to. See client/server section.
75 .BI \-\-daemonize \fR=\fPpidfile
76 Background a fio server, writing the pid to the given pid file.
78 .BI \-\-client \fR=\fPhost
79 Instead of running the jobs locally, send and run them on the given host.
81 Job files are in `ini' format. They consist of one or more
82 job definitions, which begin with a job name in square brackets and
83 extend to the next job name. The job name can be any ASCII string
84 except `global', which has a special meaning. Following the job name is
85 a sequence of zero or more parameters, one per line, that define the
86 behavior of the job. Any line starting with a `;' or `#' character is
87 considered a comment and ignored.
89 If \fIjobfile\fR is specified as `-', the job file will be read from
92 The global section contains default parameters for jobs specified in the
93 job file. A job is only affected by global sections residing above it,
94 and there may be any number of global sections. Specific job definitions
95 may override any parameter set in global sections.
98 Some parameters may take arguments of a specific type. The types used are:
101 String: a sequence of alphanumeric characters.
104 SI integer: a whole number, possibly containing a suffix denoting the base unit
105 of the value. Accepted suffixes are `k', 'M', 'G', 'T', and 'P', denoting
106 kilo (1024), mega (1024^2), giga (1024^3), tera (1024^4), and peta (1024^5)
107 respectively. The suffix is not case sensitive. If prefixed with '0x', the
108 value is assumed to be base 16 (hexadecimal). A suffix may include a trailing 'b',
109 for instance 'kb' is identical to 'k'. You can specify a base 10 value
110 by using 'KiB', 'MiB', 'GiB', etc. This is useful for disk drives where
111 values are often given in base 10 values. Specifying '30GiB' will get you
115 Boolean: a true or false value. `0' denotes false, `1' denotes true.
118 Integer range: a range of integers specified in the format
119 \fIlower\fR:\fIupper\fR or \fIlower\fR\-\fIupper\fR. \fIlower\fR and
120 \fIupper\fR may contain a suffix as described above. If an option allows two
121 sets of ranges, they are separated with a `,' or `/' character. For example:
125 List of floating numbers: A list of floating numbers, separated by
130 May be used to override the job name. On the command line, this parameter
131 has the special purpose of signalling the start of a new job.
133 .BI description \fR=\fPstr
134 Human-readable description of the job. It is printed when the job is run, but
135 otherwise has no special purpose.
137 .BI directory \fR=\fPstr
138 Prefix filenames with this directory. Used to place files in a location other
141 .BI filename \fR=\fPstr
143 normally makes up a file name based on the job name, thread number, and file
144 number. If you want to share files between threads in a job or several jobs,
145 specify a \fIfilename\fR for each of them to override the default.
146 If the I/O engine is file-based, you can specify
147 a number of files by separating the names with a `:' character. `\-' is a
148 reserved name, meaning stdin or stdout, depending on the read/write direction
151 .BI lockfile \fR=\fPstr
152 Fio defaults to not locking any files before it does IO to them. If a file or
153 file descriptor is shared, fio can serialize IO to that file to make the end
154 result consistent. This is usual for emulating real workloads that share files.
160 No locking. This is the default.
163 Only one thread or process may do IO at the time, excluding all others.
166 Read-write locking on the file. Many readers may access the file at the same
167 time, but writes get exclusive access.
170 The option may be post-fixed with a lock batch number. If set, then each
171 thread/process may do that amount of IOs to the file before giving up the lock.
172 Since lock acquisition is expensive, batching the lock/unlocks will speed up IO.
175 .BI opendir \fR=\fPstr
176 Recursively open any files below directory \fIstr\fR.
178 .BI readwrite \fR=\fPstr "\fR,\fP rw" \fR=\fPstr
179 Type of I/O pattern. Accepted values are:
196 Mixed sequential reads and writes.
199 Mixed random reads and writes.
202 For mixed I/O, the default split is 50/50. For certain types of io the result
203 may still be skewed a bit, since the speed may be different. It is possible to
204 specify a number of IO's to do before getting a new offset, this is done by
205 appending a `:\fI<nr>\fR to the end of the string given. For a random read, it
206 would look like \fBrw=randread:8\fR for passing in an offset modifier with a
207 value of 8. If the postfix is used with a sequential IO pattern, then the value
208 specified will be added to the generated offset for each IO. For instance,
209 using \fBrw=write:4k\fR will skip 4k for every write. It turns sequential IO
210 into sequential IO with holes. See the \fBrw_sequencer\fR option.
213 .BI rw_sequencer \fR=\fPstr
214 If an offset modifier is given by appending a number to the \fBrw=<str>\fR line,
215 then this option controls how that number modifies the IO offset being
216 generated. Accepted values are:
221 Generate sequential offset
224 Generate the same offset
227 \fBsequential\fR is only useful for random IO, where fio would normally
228 generate a new random offset for every IO. If you append eg 8 to randread, you
229 would get a new random offset for every 8 IO's. The result would be a seek for
230 only every 8 IO's, instead of for every IO. Use \fBrw=randread:8\fR to specify
231 that. As sequential IO is already sequential, setting \fBsequential\fR for that
232 would not result in any differences. \fBidentical\fR behaves in a similar
233 fashion, except it sends the same offset 8 number of times before generating a
238 .BI kb_base \fR=\fPint
239 The base unit for a kilobyte. The defacto base is 2^10, 1024. Storage
240 manufacturers like to use 10^3 or 1000 as a base ten unit instead, for obvious
241 reasons. Allow values are 1024 or 1000, with 1024 being the default.
243 .BI randrepeat \fR=\fPbool
244 Seed the random number generator in a predictable way so results are repeatable
245 across runs. Default: true.
247 .BI use_os_rand \fR=\fPbool
248 Fio can either use the random generator supplied by the OS to generator random
249 offsets, or it can use it's own internal generator (based on Tausworthe).
250 Default is to use the internal generator, which is often of better quality and
251 faster. Default: false.
253 .BI fallocate \fR=\fPstr
254 Whether pre-allocation is performed when laying down files. Accepted values
260 Do not pre-allocate space.
263 Pre-allocate via posix_fallocate().
266 Pre-allocate via fallocate() with FALLOC_FL_KEEP_SIZE set.
269 Backward-compatible alias for 'none'.
272 Backward-compatible alias for 'posix'.
275 May not be available on all supported platforms. 'keep' is only
276 available on Linux. If using ZFS on Solaris this must be set to 'none'
277 because ZFS doesn't support it. Default: 'posix'.
280 .BI fadvise_hint \fR=\fPbool
281 Use of \fIposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
282 are likely to be issued. Default: true.
285 Total size of I/O for this job. \fBfio\fR will run until this many bytes have
286 been transfered, unless limited by other options (\fBruntime\fR, for instance).
287 Unless \fBnrfiles\fR and \fBfilesize\fR options are given, this amount will be
288 divided between the available files for the job. If not set, fio will use the
289 full size of the given files or devices. If the the files do not exist, size
290 must be given. It is also possible to give size as a percentage between 1 and
291 100. If size=20% is given, fio will use 20% of the full size of the given files
294 .BI fill_device \fR=\fPbool "\fR,\fB fill_fs" \fR=\fPbool
295 Sets size to something really large and waits for ENOSPC (no space left on
296 device) as the terminating condition. Only makes sense with sequential write.
297 For a read workload, the mount point will be filled first then IO started on
298 the result. This option doesn't make sense if operating on a raw device node,
299 since the size of that is already known by the file system. Additionally,
300 writing beyond end-of-device will not return ENOSPC there.
302 .BI filesize \fR=\fPirange
303 Individual file sizes. May be a range, in which case \fBfio\fR will select sizes
304 for files at random within the given range, limited to \fBsize\fR in total (if
305 that is given). If \fBfilesize\fR is not specified, each created file is the
308 .BI blocksize \fR=\fPint[,int] "\fR,\fB bs" \fR=\fPint[,int]
309 Block size for I/O units. Default: 4k. Values for reads and writes can be
310 specified separately in the format \fIread\fR,\fIwrite\fR, either of
311 which may be empty to leave that value at its default.
313 .BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange]
314 Specify a range of I/O block sizes. The issued I/O unit will always be a
315 multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies
316 to both reads and writes if only one range is given, but can be specified
317 separately with a comma seperating the values. Example: bsrange=1k-4k,2k-8k.
318 Also (see \fBblocksize\fR).
320 .BI bssplit \fR=\fPstr
321 This option allows even finer grained control of the block sizes issued,
322 not just even splits between them. With this option, you can weight various
323 block sizes for exact control of the issued IO for a job that has mixed
324 block sizes. The format of the option is bssplit=blocksize/percentage,
325 optionally adding as many definitions as needed separated by a colon.
326 Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k
327 blocks and 40% 32k blocks. \fBbssplit\fR also supports giving separate
328 splits to reads and writes. The format is identical to what the
329 \fBbs\fR option accepts, the read and write parts are separated with a
332 .B blocksize_unaligned\fR,\fP bs_unaligned
333 If set, any size in \fBblocksize_range\fR may be used. This typically won't
334 work with direct I/O, as that normally requires sector alignment.
336 .BI blockalign \fR=\fPint[,int] "\fR,\fB ba" \fR=\fPint[,int]
337 At what boundary to align random IO offsets. Defaults to the same as 'blocksize'
338 the minimum blocksize given. Minimum alignment is typically 512b
339 for using direct IO, though it usually depends on the hardware block size.
340 This option is mutually exclusive with using a random map for files, so it
341 will turn off that option.
344 Initialise buffers with all zeros. Default: fill buffers with random data.
347 If this option is given, fio will refill the IO buffers on every submit. The
348 default is to only fill it at init time and reuse that data. Only makes sense
349 if zero_buffers isn't specified, naturally. If data verification is enabled,
350 refill_buffers is also automatically enabled.
352 .BI scramble_buffers \fR=\fPbool
353 If \fBrefill_buffers\fR is too costly and the target is using data
354 deduplication, then setting this option will slightly modify the IO buffer
355 contents to defeat normal de-dupe attempts. This is not enough to defeat
356 more clever block compression attempts, but it will stop naive dedupe
357 of blocks. Default: true.
359 .BI buffer_compress_percentage \fR=\fPint
360 If this is set, then fio will attempt to provide IO buffer content (on WRITEs)
361 that compress to the specified level. Fio does this by providing a mix of
362 random data and zeroes. Note that this is per block size unit, for file/disk
363 wide compression level that matches this setting, you'll also want to set
364 \fBrefill_buffers\fR.
366 .BI buffer_compress_chunk \fR=\fPint
367 See \fBbuffer_compress_percentage\fR. This setting allows fio to manage how
368 big the ranges of random data and zeroed data is. Without this set, fio will
369 provide \fBbuffer_compress_percentage\fR of blocksize random data, followed by
370 the remaining zeroed. With this set to some chunk size smaller than the block
371 size, fio can alternate random and zeroed data throughout the IO buffer.
373 .BI nrfiles \fR=\fPint
374 Number of files to use for this job. Default: 1.
376 .BI openfiles \fR=\fPint
377 Number of files to keep open at the same time. Default: \fBnrfiles\fR.
379 .BI file_service_type \fR=\fPstr
380 Defines how files to service are selected. The following types are defined:
385 Choose a file at random
388 Round robin over open files (default).
390 Do each file in the set sequentially.
393 The number of I/Os to issue before switching a new file can be specified by
394 appending `:\fIint\fR' to the service type.
397 .BI ioengine \fR=\fPstr
398 Defines how the job issues I/O. The following types are defined:
403 Basic \fIread\fR\|(2) or \fIwrite\fR\|(2) I/O. \fIfseek\fR\|(2) is used to
404 position the I/O location.
407 Basic \fIpread\fR\|(2) or \fIpwrite\fR\|(2) I/O.
410 Basic \fIreadv\fR\|(2) or \fIwritev\fR\|(2) I/O. Will emulate queuing by
411 coalescing adjacents IOs into a single submission.
414 Linux native asynchronous I/O. This ioengine defines engine specific options.
417 POSIX asynchronous I/O using \fIaio_read\fR\|(3) and \fIaio_write\fR\|(3).
420 Solaris native asynchronous I/O.
423 Windows native asynchronous I/O.
426 File is memory mapped with \fImmap\fR\|(2) and data copied using
430 \fIsplice\fR\|(2) is used to transfer the data and \fIvmsplice\fR\|(2) to
431 transfer data from user-space to the kernel.
434 Use the syslet system calls to make regular read/write asynchronous.
437 SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if
438 the target is an sg character device, we use \fIread\fR\|(2) and
439 \fIwrite\fR\|(2) for asynchronous I/O.
442 Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR
443 itself and for debugging and testing purposes.
446 Transfer over the network. The protocol to be used can be defined with the
447 \fBprotocol\fR parameter. Depending on the protocol, \fBfilename\fR,
448 \fBhostname\fR, \fBport\fR, or \fBlisten\fR must be specified.
449 This ioengine defines engine specific options.
452 Like \fBnet\fR, but uses \fIsplice\fR\|(2) and \fIvmsplice\fR\|(2) to map data
453 and send/receive. This ioengine defines engine specific options.
456 Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and
457 \fBcpucycles\fR parameters.
460 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface
461 approach to asycnronous I/O.
463 See <http://www.xmailserver.org/guasi\-lib.html>.
466 The RDMA I/O engine supports both RDMA memory semantics (RDMA_WRITE/RDMA_READ)
467 and channel semantics (Send/Recv) for the InfiniBand, RoCE and iWARP protocols.
470 Loads an external I/O engine object file. Append the engine filename as
474 IO engine that does regular linux native fallocate callt to simulate data
475 transfer as fio ioengine
477 DDIR_READ does fallocate(,mode = FALLOC_FL_KEEP_SIZE,)
479 DIR_WRITE does fallocate(,mode = 0)
481 DDIR_TRIM does fallocate(,mode = FALLOC_FL_KEEP_SIZE|FALLOC_FL_PUNCH_HOLE)
484 IO engine that does regular EXT4_IOC_MOVE_EXT ioctls to simulate defragment activity
485 request to DDIR_WRITE event
490 .BI iodepth \fR=\fPint
491 Number of I/O units to keep in flight against the file. Note that increasing
492 iodepth beyond 1 will not affect synchronous ioengines (except for small
493 degress when verify_async is in use). Even async engines my impose OS
494 restrictions causing the desired depth not to be achieved. This may happen on
495 Linux when using libaio and not setting \fBdirect\fR=1, since buffered IO is
496 not async on that OS. Keep an eye on the IO depth distribution in the
497 fio output to verify that the achieved depth is as expected. Default: 1.
499 .BI iodepth_batch \fR=\fPint
500 Number of I/Os to submit at once. Default: \fBiodepth\fR.
502 .BI iodepth_batch_complete \fR=\fPint
503 This defines how many pieces of IO to retrieve at once. It defaults to 1 which
504 means that we'll ask for a minimum of 1 IO in the retrieval process from the
505 kernel. The IO retrieval will go on until we hit the limit set by
506 \fBiodepth_low\fR. If this variable is set to 0, then fio will always check for
507 completed events before queuing more IO. This helps reduce IO latency, at the
508 cost of more retrieval system calls.
510 .BI iodepth_low \fR=\fPint
511 Low watermark indicating when to start filling the queue again. Default:
514 .BI direct \fR=\fPbool
515 If true, use non-buffered I/O (usually O_DIRECT). Default: false.
517 .BI buffered \fR=\fPbool
518 If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter.
521 .BI offset \fR=\fPint
522 Offset in the file to start I/O. Data before the offset will not be touched.
524 .BI offset_increment \fR=\fPint
525 If this is provided, then the real offset becomes the
526 offset + offset_increment * thread_number, where the thread number is a counter
527 that starts at 0 and is incremented for each job. This option is useful if
528 there are several jobs which are intended to operate on a file in parallel in
529 disjoint segments, with even spacing between the starting points.
532 How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If
533 0, don't sync. Default: 0.
535 .BI fdatasync \fR=\fPint
536 Like \fBfsync\fR, but uses \fBfdatasync\fR\|(2) instead to only sync the
537 data parts of the file. Default: 0.
539 .BI sync_file_range \fR=\fPstr:int
540 Use sync_file_range() for every \fRval\fP number of write operations. Fio will
541 track range of writes that have happened since the last sync_file_range() call.
542 \fRstr\fP can currently be one or more of:
546 SYNC_FILE_RANGE_WAIT_BEFORE
549 SYNC_FILE_RANGE_WRITE
552 SYNC_FILE_RANGE_WRITE
556 So if you do sync_file_range=wait_before,write:8, fio would use
557 \fBSYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE\fP for every 8 writes.
558 Also see the sync_file_range(2) man page. This option is Linux specific.
560 .BI overwrite \fR=\fPbool
561 If writing, setup the file first and do overwrites. Default: false.
563 .BI end_fsync \fR=\fPbool
564 Sync file contents when a write stage has completed. Default: false.
566 .BI fsync_on_close \fR=\fPbool
567 If true, sync file contents on close. This differs from \fBend_fsync\fR in that
568 it will happen on every close, not just at the end of the job. Default: false.
570 .BI rwmixread \fR=\fPint
571 Percentage of a mixed workload that should be reads. Default: 50.
573 .BI rwmixwrite \fR=\fPint
574 Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and
575 \fBrwmixwrite\fR are given and do not sum to 100%, the latter of the two
576 overrides the first. This may interfere with a given rate setting, if fio is
577 asked to limit reads or writes to a certain rate. If that is the case, then
578 the distribution may be skewed. Default: 50.
580 .BI random_distribution \fR=\fPstr:float
581 By default, fio will use a completely uniform random distribution when asked
582 to perform random IO. Sometimes it is useful to skew the distribution in
583 specific ways, ensuring that some parts of the data is more hot than others.
584 Fio includes the following distribution models:
588 Uniform random distribution
598 When using a zipf or pareto distribution, an input value is also needed to
599 define the access pattern. For zipf, this is the zipf theta. For pareto,
600 it's the pareto power. Fio includes a test program, genzipf, that can be
601 used visualize what the given input values will yield in terms of hit rates.
602 If you wanted to use zipf with a theta of 1.2, you would use
603 random_distribution=zipf:1.2 as the option. If a non-uniform model is used,
604 fio will disable use of the random map.
607 Normally \fBfio\fR will cover every block of the file when doing random I/O. If
608 this parameter is given, a new offset will be chosen without looking at past
609 I/O history. This parameter is mutually exclusive with \fBverify\fR.
611 .BI softrandommap \fR=\fPbool
612 See \fBnorandommap\fR. If fio runs with the random block map enabled and it
613 fails to allocate the map, if this option is set it will continue without a
614 random block map. As coverage will not be as complete as with random maps, this
615 option is disabled by default.
617 .BI random_generator \fR=\fPstr
618 Fio supports the following engines for generating IO offsets for random IO:
622 Strong 2^88 cycle random number generator
625 Linear feedback shift register generator
629 Tausworthe is a strong random number generator, but it requires tracking on the
630 side if we want to ensure that blocks are only read or written once. LFSR
631 guarantees that we never generate the same offset twice, and it's also less
632 computationally expensive. It's not a true random generator, however, though
633 for IO purposes it's typically good enough. LFSR only works with single block
634 sizes, not with workloads that use multiple block sizes. If used with such a
635 workload, fio may read or write some blocks multiple times.
638 Run job with given nice value. See \fInice\fR\|(2).
641 Set I/O priority value of this job between 0 (highest) and 7 (lowest). See
644 .BI prioclass \fR=\fPint
645 Set I/O priority class. See \fIionice\fR\|(1).
647 .BI thinktime \fR=\fPint
648 Stall job for given number of microseconds between issuing I/Os.
650 .BI thinktime_spin \fR=\fPint
651 Pretend to spend CPU time for given number of microseconds, sleeping the rest
652 of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set.
654 .BI thinktime_blocks \fR=\fPint
655 Number of blocks to issue before waiting \fBthinktime\fR microseconds.
659 Cap bandwidth used by this job. The number is in bytes/sec, the normal postfix
660 rules apply. You can use \fBrate\fR=500k to limit reads and writes to 500k each,
661 or you can specify read and writes separately. Using \fBrate\fR=1m,500k would
662 limit reads to 1MB/sec and writes to 500KB/sec. Capping only reads or writes
663 can be done with \fBrate\fR=,500k or \fBrate\fR=500k,. The former will only
664 limit writes (to 500KB/sec), the latter will only limit reads.
666 .BI ratemin \fR=\fPint
667 Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth.
668 Failing to meet this requirement will cause the job to exit. The same format
669 as \fBrate\fR is used for read vs write separation.
671 .BI rate_iops \fR=\fPint
672 Cap the bandwidth to this number of IOPS. Basically the same as rate, just
673 specified independently of bandwidth. The same format as \fBrate\fR is used for
674 read vs write seperation. If \fBblocksize\fR is a range, the smallest block
675 size is used as the metric.
677 .BI rate_iops_min \fR=\fPint
678 If this rate of I/O is not met, the job will exit. The same format as \fBrate\fR
679 is used for read vs write seperation.
681 .BI ratecycle \fR=\fPint
682 Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of
683 milliseconds. Default: 1000ms.
685 .BI max_latency \fR=\fPint
686 If set, fio will exit the job if it exceeds this maximum latency. It will exit
689 .BI cpumask \fR=\fPint
690 Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job
691 may run on. See \fBsched_setaffinity\fR\|(2).
693 .BI cpus_allowed \fR=\fPstr
694 Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers.
696 .BI numa_cpu_nodes \fR=\fPstr
697 Set this job running on spcified NUMA nodes' CPUs. The arguments allow
698 comma delimited list of cpu numbers, A-B ranges, or 'all'.
700 .BI numa_mem_policy \fR=\fPstr
701 Set this job's memory policy and corresponding NUMA nodes. Format of
705 .B <mode>[:<nodelist>]
708 is one of the following memory policy:
710 .B default, prefer, bind, interleave, local
713 For \fBdefault\fR and \fBlocal\fR memory policy, no \fBnodelist\fR is
714 needed to be specified. For \fBprefer\fR, only one node is
715 allowed. For \fBbind\fR and \fBinterleave\fR, \fBnodelist\fR allows
716 comma delimited list of numbers, A-B ranges, or 'all'.
718 .BI startdelay \fR=\fPint
719 Delay start of job for the specified number of seconds.
721 .BI runtime \fR=\fPint
722 Terminate processing after the specified number of seconds.
725 If given, run for the specified \fBruntime\fR duration even if the files are
726 completely read or written. The same workload will be repeated as many times
727 as \fBruntime\fR allows.
729 .BI ramp_time \fR=\fPint
730 If set, fio will run the specified workload for this amount of time before
731 logging any performance numbers. Useful for letting performance settle before
732 logging results, thus minimizing the runtime required for stable results. Note
733 that the \fBramp_time\fR is considered lead in time for a job, thus it will
734 increase the total runtime if a special timeout or runtime is specified.
736 .BI invalidate \fR=\fPbool
737 Invalidate buffer-cache for the file prior to starting I/O. Default: true.
740 Use synchronous I/O for buffered writes. For the majority of I/O engines,
741 this means using O_SYNC. Default: false.
743 .BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr
744 Allocation method for I/O unit buffer. Allowed values are:
749 Allocate memory with \fImalloc\fR\|(3).
752 Use shared memory buffers allocated through \fIshmget\fR\|(2).
755 Same as \fBshm\fR, but use huge pages as backing.
758 Use \fImmap\fR\|(2) for allocation. Uses anonymous memory unless a filename
759 is given after the option in the format `:\fIfile\fR'.
762 Same as \fBmmap\fR, but use huge files as backing.
765 The amount of memory allocated is the maximum allowed \fBblocksize\fR for the
766 job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work,
767 the system must have free huge pages allocated. \fBmmaphuge\fR also needs to
768 have hugetlbfs mounted, and \fIfile\fR must point there. At least on Linux,
769 huge pages must be manually allocated. See \fB/proc/sys/vm/nr_hugehages\fR
770 and the documentation for that. Normally you just need to echo an appropriate
771 number, eg echoing 8 will ensure that the OS has 8 huge pages ready for
775 .BI iomem_align \fR=\fPint "\fR,\fP mem_align" \fR=\fPint
776 This indiciates the memory alignment of the IO memory buffers. Note that the
777 given alignment is applied to the first IO unit buffer, if using \fBiodepth\fR
778 the alignment of the following buffers are given by the \fBbs\fR used. In
779 other words, if using a \fBbs\fR that is a multiple of the page sized in the
780 system, all buffers will be aligned to this value. If using a \fBbs\fR that
781 is not page aligned, the alignment of subsequent IO memory buffers is the
782 sum of the \fBiomem_align\fR and \fBbs\fR used.
784 .BI hugepage\-size \fR=\fPint
785 Defines the size of a huge page. Must be at least equal to the system setting.
786 Should be a multiple of 1MB. Default: 4MB.
789 Terminate all jobs when one finishes. Default: wait for each job to finish.
791 .BI bwavgtime \fR=\fPint
792 Average bandwidth calculations over the given time in milliseconds. Default:
795 .BI iopsavgtime \fR=\fPint
796 Average IOPS calculations over the given time in milliseconds. Default:
799 .BI create_serialize \fR=\fPbool
800 If true, serialize file creation for the jobs. Default: true.
802 .BI create_fsync \fR=\fPbool
803 \fIfsync\fR\|(2) data file after creation. Default: true.
805 .BI create_on_open \fR=\fPbool
806 If true, the files are not created until they are opened for IO by the job.
808 .BI create_only \fR=\fPbool
809 If true, fio will only run the setup phase of the job. If files need to be
810 laid out or updated on disk, only that will be done. The actual job contents
813 .BI pre_read \fR=\fPbool
814 If this is given, files will be pre-read into memory before starting the given
815 IO operation. This will also clear the \fR \fBinvalidate\fR flag, since it is
816 pointless to pre-read and then drop the cache. This will only work for IO
817 engines that are seekable, since they allow you to read the same data
818 multiple times. Thus it will not work on eg network or splice IO.
820 .BI unlink \fR=\fPbool
821 Unlink job files when done. Default: false.
824 Specifies the number of iterations (runs of the same workload) of this job.
827 .BI do_verify \fR=\fPbool
828 Run the verify phase after a write phase. Only valid if \fBverify\fR is set.
831 .BI verify \fR=\fPstr
832 Method of verifying file contents after each iteration of the job. Allowed
837 .B md5 crc16 crc32 crc32c crc32c-intel crc64 crc7 sha256 sha512 sha1
838 Store appropriate checksum in the header of each block. crc32c-intel is
839 hardware accelerated SSE4.2 driven, falls back to regular crc32c if
840 not supported by the system.
843 Write extra information about each I/O (timestamp, block number, etc.). The
844 block number is verified. See \fBverify_pattern\fR as well.
847 Pretend to verify. Used for testing internals.
850 This option can be used for repeated burn-in tests of a system to make sure
851 that the written data is also correctly read back. If the data direction given
852 is a read or random read, fio will assume that it should verify a previously
853 written file. If the data direction includes any form of write, the verify will
854 be of the newly written data.
857 .BI verify_sort \fR=\fPbool
858 If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to
859 read them back in a sorted manner. Default: true.
861 .BI verify_offset \fR=\fPint
862 Swap the verification header with data somewhere else in the block before
863 writing. It is swapped back before verifying.
865 .BI verify_interval \fR=\fPint
866 Write the verification header for this number of bytes, which should divide
867 \fBblocksize\fR. Default: \fBblocksize\fR.
869 .BI verify_pattern \fR=\fPstr
870 If set, fio will fill the io buffers with this pattern. Fio defaults to filling
871 with totally random bytes, but sometimes it's interesting to fill with a known
872 pattern for io verification purposes. Depending on the width of the pattern,
873 fio will fill 1/2/3/4 bytes of the buffer at the time(it can be either a
874 decimal or a hex number). The verify_pattern if larger than a 32-bit quantity
875 has to be a hex number that starts with either "0x" or "0X". Use with
878 .BI verify_fatal \fR=\fPbool
879 If true, exit the job on the first observed verification failure. Default:
882 .BI verify_dump \fR=\fPbool
883 If set, dump the contents of both the original data block and the data block we
884 read off disk to files. This allows later analysis to inspect just what kind of
885 data corruption occurred. Off by default.
887 .BI verify_async \fR=\fPint
888 Fio will normally verify IO inline from the submitting thread. This option
889 takes an integer describing how many async offload threads to create for IO
890 verification instead, causing fio to offload the duty of verifying IO contents
891 to one or more separate threads. If using this offload option, even sync IO
892 engines can benefit from using an \fBiodepth\fR setting higher than 1, as it
893 allows them to have IO in flight while verifies are running.
895 .BI verify_async_cpus \fR=\fPstr
896 Tell fio to set the given CPU affinity on the async IO verification threads.
897 See \fBcpus_allowed\fP for the format used.
899 .BI verify_backlog \fR=\fPint
900 Fio will normally verify the written contents of a job that utilizes verify
901 once that job has completed. In other words, everything is written then
902 everything is read back and verified. You may want to verify continually
903 instead for a variety of reasons. Fio stores the meta data associated with an
904 IO block in memory, so for large verify workloads, quite a bit of memory would
905 be used up holding this meta data. If this option is enabled, fio will write
906 only N blocks before verifying these blocks.
908 .BI verify_backlog_batch \fR=\fPint
909 Control how many blocks fio will verify if verify_backlog is set. If not set,
910 will default to the value of \fBverify_backlog\fR (meaning the entire queue is
911 read back and verified). If \fBverify_backlog_batch\fR is less than
912 \fBverify_backlog\fR then not all blocks will be verified, if
913 \fBverify_backlog_batch\fR is larger than \fBverify_backlog\fR, some blocks
914 will be verified more than once.
916 .B stonewall "\fR,\fP wait_for_previous"
917 Wait for preceding jobs in the job file to exit before starting this one.
918 \fBstonewall\fR implies \fBnew_group\fR.
921 Start a new reporting group. If not given, all jobs in a file will be part
922 of the same reporting group, unless separated by a stonewall.
924 .BI numjobs \fR=\fPint
925 Number of clones (processes/threads performing the same workload) of this job.
929 If set, display per-group reports instead of per-job when \fBnumjobs\fR is
933 Use threads created with \fBpthread_create\fR\|(3) instead of processes created
934 with \fBfork\fR\|(2).
936 .BI zonesize \fR=\fPint
937 Divide file into zones of the specified size in bytes. See \fBzoneskip\fR.
939 .BI zoneskip \fR=\fPint
940 Skip the specified number of bytes when \fBzonesize\fR bytes of data have been
943 .BI write_iolog \fR=\fPstr
944 Write the issued I/O patterns to the specified file. Specify a separate file
945 for each job, otherwise the iologs will be interspersed and the file may be
948 .BI read_iolog \fR=\fPstr
949 Replay the I/O patterns contained in the specified file generated by
950 \fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file.
952 .BI replay_no_stall \fR=\fPint
953 While replaying I/O patterns using \fBread_iolog\fR the default behavior
954 attempts to respect timing information between I/Os. Enabling
955 \fBreplay_no_stall\fR causes I/Os to be replayed as fast as possible while
956 still respecting ordering.
958 .BI replay_redirect \fR=\fPstr
959 While replaying I/O patterns using \fBread_iolog\fR the default behavior
960 is to replay the IOPS onto the major/minor device that each IOP was recorded
961 from. Setting \fBreplay_redirect\fR causes all IOPS to be replayed onto the
962 single specified device regardless of the device it was recorded from.
964 .BI write_bw_log \fR=\fPstr
965 If given, write a bandwidth log of the jobs in this job file. Can be used to
966 store data of the bandwidth of the jobs in their lifetime. The included
967 fio_generate_plots script uses gnuplot to turn these text files into nice
968 graphs. See \fBwrite_log_log\fR for behaviour of given filename. For this
969 option, the postfix is _bw.log.
971 .BI write_lat_log \fR=\fPstr
972 Same as \fBwrite_bw_log\fR, but writes I/O completion latencies. If no
973 filename is given with this option, the default filename of "jobname_type.log"
974 is used. Even if the filename is given, fio will still append the type of log.
976 .BI write_iops_log \fR=\fPstr
977 Same as \fBwrite_bw_log\fR, but writes IOPS. If no filename is given with this
978 option, the default filename of "jobname_type.log" is used. Even if the
979 filename is given, fio will still append the type of log.
981 .BI log_avg_msec \fR=\fPint
982 By default, fio will log an entry in the iops, latency, or bw log for every
983 IO that completes. When writing to the disk log, that can quickly grow to a
984 very large size. Setting this option makes fio average the each log entry
985 over the specified period of time, reducing the resolution of the log.
988 .BI disable_lat \fR=\fPbool
989 Disable measurements of total latency numbers. Useful only for cutting
990 back the number of calls to gettimeofday, as that does impact performance at
991 really high IOPS rates. Note that to really get rid of a large amount of these
992 calls, this option must be used with disable_slat and disable_bw as well.
994 .BI disable_clat \fR=\fPbool
995 Disable measurements of completion latency numbers. See \fBdisable_lat\fR.
997 .BI disable_slat \fR=\fPbool
998 Disable measurements of submission latency numbers. See \fBdisable_lat\fR.
1000 .BI disable_bw_measurement \fR=\fPbool
1001 Disable measurements of throughput/bandwidth numbers. See \fBdisable_lat\fR.
1003 .BI lockmem \fR=\fPint
1004 Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to
1005 simulate a smaller amount of memory.
1007 .BI exec_prerun \fR=\fPstr
1008 Before running the job, execute the specified command with \fBsystem\fR\|(3).
1010 .BI exec_postrun \fR=\fPstr
1011 Same as \fBexec_prerun\fR, but the command is executed after the job completes.
1013 .BI ioscheduler \fR=\fPstr
1014 Attempt to switch the device hosting the file to the specified I/O scheduler.
1016 .BI cpuload \fR=\fPint
1017 If the job is a CPU cycle-eater, attempt to use the specified percentage of
1020 .BI cpuchunks \fR=\fPint
1021 If the job is a CPU cycle-eater, split the load into cycles of the
1022 given time in milliseconds.
1024 .BI disk_util \fR=\fPbool
1025 Generate disk utilization statistics if the platform supports it. Default: true.
1027 .BI clocksource \fR=\fPstr
1028 Use the given clocksource as the base of timing. The supported options are:
1038 Internal CPU clock source
1042 \fBcpu\fR is the preferred clocksource if it is reliable, as it is very fast
1043 (and fio is heavy on time calls). Fio will automatically use this clocksource
1044 if it's supported and considered reliable on the system it is running on,
1045 unless another clocksource is specifically set. For x86/x86-64 CPUs, this
1046 means supporting TSC Invariant.
1048 .BI gtod_reduce \fR=\fPbool
1049 Enable all of the gettimeofday() reducing options (disable_clat, disable_slat,
1050 disable_bw) plus reduce precision of the timeout somewhat to really shrink the
1051 gettimeofday() call count. With this option enabled, we only do about 0.4% of
1052 the gtod() calls we would have done if all time keeping was enabled.
1054 .BI gtod_cpu \fR=\fPint
1055 Sometimes it's cheaper to dedicate a single thread of execution to just getting
1056 the current time. Fio (and databases, for instance) are very intensive on
1057 gettimeofday() calls. With this option, you can set one CPU aside for doing
1058 nothing but logging current time to a shared memory location. Then the other
1059 threads/processes that run IO workloads need only copy that segment, instead of
1060 entering the kernel with a gettimeofday() call. The CPU set aside for doing
1061 these time calls will be excluded from other uses. Fio will manually clear it
1062 from the CPU mask of other jobs.
1064 .BI ignore_error \fR=\fPstr
1065 Sometimes you want to ignore some errors during test in that case you can specify
1066 error list for each error type.
1068 ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST
1070 errors for given error type is separated with ':'.
1071 Error may be symbol ('ENOSPC', 'ENOMEM') or an integer.
1073 Example: ignore_error=EAGAIN,ENOSPC:122 .
1075 This option will ignore EAGAIN from READ, and ENOSPC and 122(EDQUOT) from WRITE.
1077 .BI error_dump \fR=\fPbool
1078 If set dump every error even if it is non fatal, true by default. If disabled
1079 only fatal error will be dumped
1081 .BI cgroup \fR=\fPstr
1082 Add job to this control group. If it doesn't exist, it will be created.
1083 The system must have a mounted cgroup blkio mount point for this to work. If
1084 your system doesn't have it mounted, you can do so with:
1086 # mount \-t cgroup \-o blkio none /cgroup
1088 .BI cgroup_weight \fR=\fPint
1089 Set the weight of the cgroup to this value. See the documentation that comes
1090 with the kernel, allowed values are in the range of 100..1000.
1092 .BI cgroup_nodelete \fR=\fPbool
1093 Normally fio will delete the cgroups it has created after the job completion.
1094 To override this behavior and to leave cgroups around after the job completion,
1095 set cgroup_nodelete=1. This can be useful if one wants to inspect various
1096 cgroup files after job completion. Default: false
1099 Instead of running as the invoking user, set the user ID to this value before
1100 the thread/process does any work.
1103 Set group ID, see \fBuid\fR.
1105 .BI flow_id \fR=\fPint
1106 The ID of the flow. If not specified, it defaults to being a global flow. See
1110 Weight in token-based flow control. If this value is used, then there is a
1111 \fBflow counter\fR which is used to regulate the proportion of activity between
1112 two or more jobs. fio attempts to keep this flow counter near zero. The
1113 \fBflow\fR parameter stands for how much should be added or subtracted to the
1114 flow counter on each iteration of the main I/O loop. That is, if one job has
1115 \fBflow=8\fR and another job has \fBflow=-1\fR, then there will be a roughly
1116 1:8 ratio in how much one runs vs the other.
1118 .BI flow_watermark \fR=\fPint
1119 The maximum value that the absolute value of the flow counter is allowed to
1120 reach before the job must wait for a lower value of the counter.
1122 .BI flow_sleep \fR=\fPint
1123 The period of time, in microseconds, to wait after the flow watermark has been
1124 exceeded before retrying operations
1126 .BI clat_percentiles \fR=\fPbool
1127 Enable the reporting of percentiles of completion latencies.
1129 .BI percentile_list \fR=\fPfloat_list
1130 Overwrite the default list of percentiles for completion
1131 latencies. Each number is a floating number in the range (0,100], and
1132 the maximum length of the list is 20. Use ':' to separate the
1133 numbers. For example, \-\-percentile_list=99.5:99.9 will cause fio to
1134 report the values of completion latency below which 99.5% and 99.9% of
1135 the observed latencies fell, respectively.
1136 .SS "Ioengine Parameters List"
1137 Some parameters are only valid when a specific ioengine is in use. These are
1138 used identically to normal parameters, with the caveat that when used on the
1139 command line, the must come after the ioengine that defines them is selected.
1141 .BI (libaio)userspace_reap
1142 Normally, with the libaio engine in use, fio will use
1143 the io_getevents system call to reap newly returned events.
1144 With this flag turned on, the AIO ring will be read directly
1145 from user-space to reap events. The reaping mode is only
1146 enabled when polling for a minimum of 0 events (eg when
1147 iodepth_batch_complete=0).
1149 .BI (net,netsplice)hostname \fR=\fPstr
1150 The host name or IP address to use for TCP or UDP based IO.
1151 If the job is a TCP listener or UDP reader, the hostname is not
1152 used and must be omitted.
1154 .BI (net,netsplice)port \fR=\fPint
1155 The TCP or UDP port to bind to or connect to.
1157 .BI (net,netsplice)protocol \fR=\fPstr "\fR,\fP proto" \fR=\fPstr
1158 The network protocol to use. Accepted values are:
1163 Transmission control protocol
1166 User datagram protocol
1172 When the protocol is TCP or UDP, the port must also be given,
1173 as well as the hostname if the job is a TCP listener or UDP
1174 reader. For unix sockets, the normal filename option should be
1175 used and the port is invalid.
1178 .BI (net,netsplice)listen
1179 For TCP network connections, tell fio to listen for incoming
1180 connections rather than initiating an outgoing connection. The
1181 hostname must be omitted if this option is used.
1183 .BI (net, pingpong) \fR=\fPbool
1184 Normal a network writer will just continue writing data, and a network reader
1185 will just consume packages. If pingpong=1 is set, a writer will send its normal
1186 payload to the reader, then wait for the reader to send the same payload back.
1187 This allows fio to measure network latencies. The submission and completion
1188 latencies then measure local time spent sending or receiving, and the
1189 completion latency measures how long it took for the other end to receive and
1192 .BI (e4defrag,donorname) \fR=\fPstr
1193 File will be used as a block donor (swap extents between files)
1195 .BI (e4defrag,inplace) \fR=\fPint
1196 Configure donor file block allocation strategy
1199 Preallocate donor's file on init
1202 allocate space immidietly inside defragment event, and free right after event
1206 While running, \fBfio\fR will display the status of the created jobs. For
1210 Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
1213 The characters in the first set of brackets denote the current status of each
1214 threads. The possible values are:
1220 Setup but not started.
1226 Initialized, waiting.
1229 Running, doing sequential reads.
1232 Running, doing random reads.
1235 Running, doing sequential writes.
1238 Running, doing random writes.
1241 Running, doing mixed sequential reads/writes.
1244 Running, doing mixed random reads/writes.
1247 Running, currently waiting for \fBfsync\fR\|(2).
1250 Running, verifying written data.
1253 Exited, not reaped by main thread.
1256 Exited, thread reaped.
1260 The second set of brackets shows the estimated completion percentage of
1261 the current group. The third set shows the read and write I/O rate,
1262 respectively. Finally, the estimated run time of the job is displayed.
1264 When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data
1265 for each thread, each group of threads, and each disk, in that order.
1267 Per-thread statistics first show the threads client number, group-id, and
1268 error code. The remaining figures are as follows:
1272 Number of megabytes of I/O performed.
1275 Average data rate (bandwidth).
1281 Submission latency minimum, maximum, average and standard deviation. This is
1282 the time it took to submit the I/O.
1285 Completion latency minimum, maximum, average and standard deviation. This
1286 is the time between submission and completion.
1289 Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average
1290 and standard deviation.
1293 CPU usage statistics. Includes user and system time, number of context switches
1294 this thread went through and number of major and minor page faults.
1297 Distribution of I/O depths. Each depth includes everything less than (or equal)
1298 to it, but greater than the previous depth.
1301 Number of read/write requests issued, and number of short read/write requests.
1304 Distribution of I/O completion latencies. The numbers follow the same pattern
1308 The group statistics show:
1313 Number of megabytes I/O performed.
1316 Aggregate bandwidth of threads in the group.
1319 Minimum average bandwidth a thread saw.
1322 Maximum average bandwidth a thread saw.
1325 Shortest runtime of threads in the group.
1328 Longest runtime of threads in the group.
1332 Finally, disk statistics are printed with reads first:
1337 Number of I/Os performed by all groups.
1340 Number of merges in the I/O scheduler.
1343 Number of ticks we kept the disk busy.
1346 Total time spent in the disk queue.
1353 It is also possible to get fio to dump the current output while it is
1354 running, without terminating the job. To do that, send fio the \fBUSR1\fR
1357 If the \fB\-\-minimal\fR option is given, the results will be printed in a
1358 semicolon-delimited format suitable for scripted use - a job description
1359 (if provided) follows on a new line. Note that the first
1360 number in the line is the version number. If the output has to be changed
1361 for some reason, this number will be incremented by 1 to signify that
1362 change. The fields are:
1365 .B terse version, fio version, jobname, groupid, error
1369 .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP
1373 .B min, max, mean, standard deviation
1377 .B min, max, mean, standard deviation
1379 Completion latency percentiles (20 fields):
1381 .B Xth percentile=usec
1385 .B min, max, mean, standard deviation
1389 .B min, max, aggregate percentage of total, mean, standard deviation
1395 .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP
1399 .B min, max, mean, standard deviation
1403 .B min, max, mean, standard deviation
1405 Completion latency percentiles (20 fields):
1407 .B Xth percentile=usec
1411 .B min, max, mean, standard deviation
1415 .B min, max, aggregate percentage of total, mean, standard deviation
1421 .B user, system, context switches, major page faults, minor page faults
1424 IO depth distribution:
1426 .B <=1, 2, 4, 8, 16, 32, >=64
1429 IO latency distribution:
1433 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
1437 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
1441 Disk utilization (1 for each disk used):
1443 .B name, read ios, write ios, read merges, write merges, read ticks, write ticks, read in-queue time, write in-queue time, disk utilization percentage
1446 Error Info (dependent on continue_on_error, default off):
1448 .B total # errors, first error code
1451 .B text description (if provided in config - appears on newline)
1454 Normally you would run fio as a stand-alone application on the machine
1455 where the IO workload should be generated. However, it is also possible to
1456 run the frontend and backend of fio separately. This makes it possible to
1457 have a fio server running on the machine(s) where the IO workload should
1458 be running, while controlling it from another machine.
1460 To start the server, you would do:
1462 \fBfio \-\-server=args\fR
1464 on that machine, where args defines what fio listens to. The arguments
1465 are of the form 'type:hostname or IP:port'. 'type' is either 'ip' (or ip4)
1466 for TCP/IP v4, 'ip6' for TCP/IP v6, or 'sock' for a local unix domain
1467 socket. 'hostname' is either a hostname or IP address, and 'port' is the port to
1468 listen to (only valid for TCP/IP, not a local socket). Some examples:
1472 Start a fio server, listening on all interfaces on the default port (8765).
1474 2) fio \-\-server=ip:hostname,4444
1476 Start a fio server, listening on IP belonging to hostname and on port 4444.
1478 3) fio \-\-server=ip6:::1,4444
1480 Start a fio server, listening on IPv6 localhost ::1 and on port 4444.
1482 4) fio \-\-server=,4444
1484 Start a fio server, listening on all interfaces on port 4444.
1486 5) fio \-\-server=1.2.3.4
1488 Start a fio server, listening on IP 1.2.3.4 on the default port.
1490 6) fio \-\-server=sock:/tmp/fio.sock
1492 Start a fio server, listening on the local socket /tmp/fio.sock.
1494 When a server is running, you can connect to it from a client. The client
1497 fio \-\-local-args \-\-client=server \-\-remote-args <job file(s)>
1499 where \-\-local-args are arguments that are local to the client where it is
1500 running, 'server' is the connect string, and \-\-remote-args and <job file(s)>
1501 are sent to the server. The 'server' string follows the same format as it
1502 does on the server side, to allow IP/hostname/socket and port strings.
1503 You can connect to multiple clients as well, to do that you could run:
1505 fio \-\-client=server2 \-\-client=server2 <job file(s)>
1509 was written by Jens Axboe <jens.axboe@oracle.com>,
1510 now Jens Axboe <jaxboe@fusionio.com>.
1512 This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based
1513 on documentation by Jens Axboe.
1514 .SH "REPORTING BUGS"
1515 Report bugs to the \fBfio\fR mailing list <fio@vger.kernel.org>.
1518 For further documentation see \fBHOWTO\fR and \fBREADME\fR.
1520 Sample jobfiles are available in the \fBexamples\fR directory.
projects / fio.git / blob