blktrace support update
[fio.git] / HOWTO
CommitLineData
71bfa161
JA
1Table of contents
2-----------------
3
41. Overview
52. How fio works
63. Running fio
74. Job file format
85. Detailed list of parameters
96. Normal output
107. Terse output
11
12
131.0 Overview and history
14------------------------
15fio was originally written to save me the hassle of writing special test
16case programs when I wanted to test a specific workload, either for
17performance reasons or to find/reproduce a bug. The process of writing
18such a test app can be tiresome, especially if you have to do it often.
19Hence I needed a tool that would be able to simulate a given io workload
20without resorting to writing a tailored test case again and again.
21
22A test work load is difficult to define, though. There can be any number
23of processes or threads involved, and they can each be using their own
24way of generating io. You could have someone dirtying large amounts of
25memory in an memory mapped file, or maybe several threads issuing
26reads using asynchronous io. fio needed to be flexible enough to
27simulate both of these cases, and many more.
28
292.0 How fio works
30-----------------
31The first step in getting fio to simulate a desired io workload, is
32writing a job file describing that specific setup. A job file may contain
33any number of threads and/or files - the typical contents of the job file
34is a global section defining shared parameters, and one or more job
35sections describing the jobs involved. When run, fio parses this file
36and sets everything up as described. If we break down a job from top to
37bottom, it contains the following basic parameters:
38
39 IO type Defines the io pattern issued to the file(s).
40 We may only be reading sequentially from this
41 file(s), or we may be writing randomly. Or even
42 mixing reads and writes, sequentially or randomly.
43
44 Block size In how large chunks are we issuing io? This may be
45 a single value, or it may describe a range of
46 block sizes.
47
48 IO size How much data are we going to be reading/writing.
49
50 IO engine How do we issue io? We could be memory mapping the
51 file, we could be using regular read/write, we
d0ff85df 52 could be using splice, async io, syslet, or even
71bfa161
JA
53 SG (SCSI generic sg).
54
6c219763 55 IO depth If the io engine is async, how large a queuing
71bfa161
JA
56 depth do we want to maintain?
57
58 IO type Should we be doing buffered io, or direct/raw io?
59
60 Num files How many files are we spreading the workload over.
61
62 Num threads How many threads or processes should we spread
63 this workload over.
64
65The above are the basic parameters defined for a workload, in addition
66there's a multitude of parameters that modify other aspects of how this
67job behaves.
68
69
703.0 Running fio
71---------------
72See the README file for command line parameters, there are only a few
73of them.
74
75Running fio is normally the easiest part - you just give it the job file
76(or job files) as parameters:
77
78$ fio job_file
79
80and it will start doing what the job_file tells it to do. You can give
81more than one job file on the command line, fio will serialize the running
82of those files. Internally that is the same as using the 'stonewall'
83parameter described the the parameter section.
84
b4692828
JA
85If the job file contains only one job, you may as well just give the
86parameters on the command line. The command line parameters are identical
87to the job parameters, with a few extra that control global parameters
88(see README). For example, for the job file parameter iodepth=2, the
c2b1e753
JA
89mirror command line option would be --iodepth 2 or --iodepth=2. You can
90also use the command line for giving more than one job entry. For each
91--name option that fio sees, it will start a new job with that name.
92Command line entries following a --name entry will apply to that job,
93until there are no more entries or a new --name entry is seen. This is
94similar to the job file options, where each option applies to the current
95job until a new [] job entry is seen.
b4692828 96
71bfa161
JA
97fio does not need to run as root, except if the files or devices specified
98in the job section requires that. Some other options may also be restricted,
6c219763 99such as memory locking, io scheduler switching, and decreasing the nice value.
71bfa161
JA
100
101
1024.0 Job file format
103-------------------
104As previously described, fio accepts one or more job files describing
105what it is supposed to do. The job file format is the classic ini file,
106where the names enclosed in [] brackets define the job name. You are free
107to use any ascii name you want, except 'global' which has special meaning.
108A global section sets defaults for the jobs described in that file. A job
109may override a global section parameter, and a job file may even have
110several global sections if so desired. A job is only affected by a global
65db0851
JA
111section residing above it. If the first character in a line is a ';' or a
112'#', the entire line is discarded as a comment.
71bfa161
JA
113
114So lets look at a really simple job file that define to threads, each
115randomly reading from a 128MiB file.
116
117; -- start job file --
118[global]
119rw=randread
120size=128m
121
122[job1]
123
124[job2]
125
126; -- end job file --
127
128As you can see, the job file sections themselves are empty as all the
129described parameters are shared. As no filename= option is given, fio
c2b1e753
JA
130makes up a filename for each of the jobs as it sees fit. On the command
131line, this job would look as follows:
132
133$ fio --name=global --rw=randread --size=128m --name=job1 --name=job2
134
71bfa161
JA
135
136Lets look at an example that have a number of processes writing randomly
137to files.
138
139; -- start job file --
140[random-writers]
141ioengine=libaio
142iodepth=4
143rw=randwrite
144bs=32k
145direct=0
146size=64m
147numjobs=4
148
149; -- end job file --
150
151Here we have no global section, as we only have one job defined anyway.
152We want to use async io here, with a depth of 4 for each file. We also
153increased the buffer size used to 32KiB and define numjobs to 4 to
154fork 4 identical jobs. The result is 4 processes each randomly writing
b4692828
JA
155to their own 64MiB file. Instead of using the above job file, you could
156have given the parameters on the command line. For this case, you would
157specify:
158
159$ fio --name=random-writers --ioengine=libaio --iodepth=4 --rw=randwrite --bs=32k --direct=0 --size=64m --numjobs=4
71bfa161
JA
160
161fio ships with a few example job files, you can also look there for
162inspiration.
163
164
1655.0 Detailed list of parameters
166-------------------------------
167
168This section describes in details each parameter associated with a job.
169Some parameters take an option of a given type, such as an integer or
170a string. The following types are used:
171
172str String. This is a sequence of alpha characters.
173int Integer. A whole number value, may be negative.
174siint SI integer. A whole number value, which may contain a postfix
175 describing the base of the number. Accepted postfixes are k/m/g,
6c219763 176 meaning kilo, mega, and giga. So if you want to specify 4096,
71bfa161
JA
177 you could either write out '4096' or just give 4k. The postfixes
178 signify base 2 values, so 1024 is 1k and 1024k is 1m and so on.
43159d18
JA
179 If the option accepts an upper and lower range, use a colon ':'
180 or minus '-' to seperate such values. See irange.
71bfa161
JA
181bool Boolean. Usually parsed as an integer, however only defined for
182 true and false (1 and 0).
183irange Integer range with postfix. Allows value range to be given, such
0c9baf91
JA
184 as 1024-4096. A colon may also be used as the seperator, eg
185 1k:4k. If the option allows two sets of ranges, they can be
186 specified with a ',' or '/' delimiter: 1k-4k/8k-32k. Also see
187 siint.
71bfa161
JA
188
189With the above in mind, here follows the complete list of fio job
190parameters.
191
192name=str ASCII name of the job. This may be used to override the
193 name printed by fio for this job. Otherwise the job
c2b1e753 194 name is used. On the command line this parameter has the
6c219763 195 special purpose of also signaling the start of a new
c2b1e753 196 job.
71bfa161 197
61697c37
JA
198description=str Text description of the job. Doesn't do anything except
199 dump this text description when this job is run. It's
200 not parsed.
201
71bfa161
JA
202directory=str Prefix filenames with this directory. Used to places files
203 in a different location than "./".
204
205filename=str Fio normally makes up a filename based on the job name,
206 thread number, and file number. If you want to share
207 files between threads in a job or several jobs, specify
ed92ac0c
JA
208 a filename for each of them to override the default. If
209 the ioengine used is 'net', the filename is the host and
9f9214f2 210 port to connect to in the format of =host/port. If the
af52b345
JA
211 ioengine is file based, you can specify a number of files
212 by seperating the names with a ':' colon. So if you wanted
213 a job to open /dev/sda and /dev/sdb as the two working files,
66159828
JA
214 you would use filename=/dev/sda:/dev/sdb. '-' is a reserved
215 name, meaning stdin or stdout. Which of the two depends
216 on the read/write direction set.
71bfa161 217
bbf6b540
JA
218opendir=str Tell fio to recursively add any file it can find in this
219 directory and down the file system tree.
220
d3aad8f2 221readwrite=str
71bfa161
JA
222rw=str Type of io pattern. Accepted values are:
223
224 read Sequential reads
225 write Sequential writes
226 randwrite Random writes
227 randread Random reads
228 rw Sequential mixed reads and writes
229 randrw Random mixed reads and writes
230
231 For the mixed io types, the default is to split them 50/50.
232 For certain types of io the result may still be skewed a bit,
211097b2
JA
233 since the speed may be different. It is possible to specify
234 a number of IO's to do before getting a new offset - this
235 is only useful for random IO, where fio would normally
236 generate a new random offset for every IO. If you append
237 eg 8 to randread, you would get a new random offset for
238 every 8 IO's. The result would be a seek for only every 8
239 IO's, instead of for every IO. Use rw=randread:8 to specify
240 that.
71bfa161 241
ee738499
JA
242randrepeat=bool For random IO workloads, seed the generator in a predictable
243 way so that results are repeatable across repetitions.
244
d2f3ac35
JA
245fadvise_hint=bool By default, fio will use fadvise() to advise the kernel
246 on what IO patterns it is likely to issue. Sometimes you
247 want to test specific IO patterns without telling the
248 kernel about it, in which case you can disable this option.
249 If set, fio will use POSIX_FADV_SEQUENTIAL for sequential
250 IO and POSIX_FADV_RANDOM for random IO.
251
71bfa161
JA
252size=siint The total size of file io for this job. This may describe
253 the size of the single file the job uses, or it may be
254 divided between the number of files in the job. If the
255 file already exists, the file size will be adjusted to this
256 size if larger than the current file size. If this parameter
257 is not given and the file exists, the file size will be used.
258
9c60ce64
JA
259filesize=siint Individual file sizes. May be a range, in which case fio
260 will select sizes for files at random within the given range
261 and limited to 'size' in total (if that is given). If not
262 given, each created file is the same size.
263
d3aad8f2 264blocksize=siint
f90eff5a
JA
265bs=siint The block size used for the io units. Defaults to 4k. Values
266 can be given for both read and writes. If a single siint is
267 given, it will apply to both. If a second siint is specified
268 after a comma, it will apply to writes only. In other words,
269 the format is either bs=read_and_write or bs=read,write.
270 bs=4k,8k will thus use 4k blocks for reads, and 8k blocks
787f7e95
JA
271 for writes. If you only wish to set the write size, you
272 can do so by passing an empty read size - bs=,8k will set
273 8k for writes and leave the read default value.
a00735e6 274
d3aad8f2 275blocksize_range=irange
71bfa161
JA
276bsrange=irange Instead of giving a single block size, specify a range
277 and fio will mix the issued io block sizes. The issued
278 io unit will always be a multiple of the minimum value
f90eff5a
JA
279 given (also see bs_unaligned). Applies to both reads and
280 writes, however a second range can be given after a comma.
281 See bs=.
a00735e6 282
d3aad8f2 283blocksize_unaligned
690adba3
JA
284bs_unaligned If this option is given, any byte size value within bsrange
285 may be used as a block range. This typically wont work with
286 direct IO, as that normally requires sector alignment.
71bfa161 287
e9459e5a
JA
288zero_buffers If this option is given, fio will init the IO buffers to
289 all zeroes. The default is to fill them with random data.
290
71bfa161
JA
291nrfiles=int Number of files to use for this job. Defaults to 1.
292
390b1537
JA
293openfiles=int Number of files to keep open at the same time. Defaults to
294 the same as nrfiles, can be set smaller to limit the number
295 simultaneous opens.
296
5af1c6f3
JA
297file_service_type=str Defines how fio decides which file from a job to
298 service next. The following types are defined:
299
300 random Just choose a file at random.
301
302 roundrobin Round robin over open files. This
303 is the default.
304
1907dbc6
JA
305 The string can have a number appended, indicating how
306 often to switch to a new file. So if option random:4 is
307 given, fio will switch to a new random file after 4 ios
308 have been issued.
309
71bfa161
JA
310ioengine=str Defines how the job issues io to the file. The following
311 types are defined:
312
313 sync Basic read(2) or write(2) io. lseek(2) is
314 used to position the io location.
315
316 libaio Linux native asynchronous io.
317
318 posixaio glibc posix asynchronous io.
319
320 mmap File is memory mapped and data copied
321 to/from using memcpy(3).
322
323 splice splice(2) is used to transfer the data and
324 vmsplice(2) to transfer data from user
325 space to the kernel.
326
d0ff85df
JA
327 syslet-rw Use the syslet system calls to make
328 regular read/write async.
329
71bfa161 330 sg SCSI generic sg v3 io. May either be
6c219763 331 synchronous using the SG_IO ioctl, or if
71bfa161
JA
332 the target is an sg character device
333 we use read(2) and write(2) for asynchronous
334 io.
335
a94ea28b
JA
336 null Doesn't transfer any data, just pretends
337 to. This is mainly used to exercise fio
338 itself and for debugging/testing purposes.
339
ed92ac0c
JA
340 net Transfer over the network to given host:port.
341 'filename' must be set appropriately to
9f9214f2 342 filename=host/port regardless of send
ed92ac0c
JA
343 or receive, if the latter only the port
344 argument is used.
345
ba0fbe10
JA
346 cpu Doesn't transfer any data, but burns CPU
347 cycles according to the cpuload= and
348 cpucycle= options. Setting cpuload=85
349 will cause that job to do nothing but burn
350 85% of the CPU.
351
e9a1806f
JA
352 guasi The GUASI IO engine is the Generic Userspace
353 Asyncronous Syscall Interface approach
354 to async IO. See
355
356 http://www.xmailserver.org/guasi-lib.html
357
358 for more info on GUASI.
359
8a7bd877
JA
360 external Prefix to specify loading an external
361 IO engine object file. Append the engine
362 filename, eg ioengine=external:/tmp/foo.o
363 to load ioengine foo.o in /tmp.
364
71bfa161
JA
365iodepth=int This defines how many io units to keep in flight against
366 the file. The default is 1 for each file defined in this
367 job, can be overridden with a larger value for higher
368 concurrency.
369
cb5ab512
JA
370iodepth_batch=int This defines how many pieces of IO to submit at once.
371 It defaults to the same as iodepth, but can be set lower
372 if one so desires.
373
e916b390
JA
374iodepth_low=int The low water mark indicating when to start filling
375 the queue again. Defaults to the same as iodepth, meaning
376 that fio will attempt to keep the queue full at all times.
377 If iodepth is set to eg 16 and iodepth_low is set to 4, then
378 after fio has filled the queue of 16 requests, it will let
379 the depth drain down to 4 before starting to fill it again.
380
71bfa161 381direct=bool If value is true, use non-buffered io. This is usually
76a43db4
JA
382 O_DIRECT.
383
384buffered=bool If value is true, use buffered io. This is the opposite
385 of the 'direct' option. Defaults to true.
71bfa161
JA
386
387offset=siint Start io at the given offset in the file. The data before
388 the given offset will not be touched. This effectively
389 caps the file size at real_size - offset.
390
391fsync=int If writing to a file, issue a sync of the dirty data
392 for every number of blocks given. For example, if you give
393 32 as a parameter, fio will sync the file for every 32
394 writes issued. If fio is using non-buffered io, we may
395 not sync the file. The exception is the sg io engine, which
6c219763 396 synchronizes the disk cache anyway.
71bfa161
JA
397
398overwrite=bool If writing to a file, setup the file first and do overwrites.
399
400end_fsync=bool If true, fsync file contents when the job exits.
401
ebb1415f
JA
402fsync_on_close=bool If true, fio will fsync() a dirty file on close.
403 This differs from end_fsync in that it will happen on every
404 file close, not just at the end of the job.
405
6c219763 406rwmixcycle=int Value in milliseconds describing how often to switch between
71bfa161
JA
407 reads and writes for a mixed workload. The default is
408 500 msecs.
409
410rwmixread=int How large a percentage of the mix should be reads.
411
412rwmixwrite=int How large a percentage of the mix should be writes. If both
413 rwmixread and rwmixwrite is given and the values do not add
414 up to 100%, the latter of the two will be used to override
415 the first.
416
bb8895e0
JA
417norandommap Normally fio will cover every block of the file when doing
418 random IO. If this option is given, fio will just get a
419 new random offset without looking at past io history. This
420 means that some blocks may not be read or written, and that
421 some blocks may be read/written more than once. This option
422 is mutually exclusive with verify= for that reason.
423
71bfa161
JA
424nice=int Run the job with the given nice value. See man nice(2).
425
426prio=int Set the io priority value of this job. Linux limits us to
427 a positive value between 0 and 7, with 0 being the highest.
428 See man ionice(1).
429
430prioclass=int Set the io priority class. See man ionice(1).
431
432thinktime=int Stall the job x microseconds after an io has completed before
433 issuing the next. May be used to simulate processing being
48097d5c
JA
434 done by an application. See thinktime_blocks and
435 thinktime_spin.
436
437thinktime_spin=int
438 Only valid if thinktime is set - pretend to spend CPU time
439 doing something with the data received, before falling back
440 to sleeping for the rest of the period specified by
441 thinktime.
9c1f7434
JA
442
443thinktime_blocks
444 Only valid if thinktime is set - control how many blocks
445 to issue, before waiting 'thinktime' usecs. If not set,
446 defaults to 1 which will make fio wait 'thinktime' usecs
447 after every block.
71bfa161
JA
448
449rate=int Cap the bandwidth used by this job to this number of KiB/sec.
450
451ratemin=int Tell fio to do whatever it can to maintain at least this
4e991c23
JA
452 bandwidth. Failing to meet this requirement, will cause
453 the job to exit.
454
455rate_iops=int Cap the bandwidth to this number of IOPS. Basically the same
456 as rate, just specified independently of bandwidth. If the
457 job is given a block size range instead of a fixed value,
458 the smallest block size is used as the metric.
459
460rate_iops_min=int If fio doesn't meet this rate of IO, it will cause
461 the job to exit.
71bfa161
JA
462
463ratecycle=int Average bandwidth for 'rate' and 'ratemin' over this number
6c219763 464 of milliseconds.
71bfa161
JA
465
466cpumask=int Set the CPU affinity of this job. The parameter given is a
467 bitmask of allowed CPU's the job may run on. See man
468 sched_setaffinity(2).
469
470startdelay=int Start this job the specified number of seconds after fio
471 has started. Only useful if the job file contains several
472 jobs, and you want to delay starting some jobs to a certain
473 time.
474
03b74b3e 475runtime=int Tell fio to terminate processing after the specified number
71bfa161
JA
476 of seconds. It can be quite hard to determine for how long
477 a specified job will run, so this parameter is handy to
478 cap the total runtime to a given time.
479
cf4464ca
JA
480time_based If set, fio will run for the duration of the runtime
481 specified even if the file(s) are completey read or
482 written. It will simply loop over the same workload
483 as many times as the runtime allows.
484
71bfa161
JA
485invalidate=bool Invalidate the buffer/page cache parts for this file prior
486 to starting io. Defaults to true.
487
488sync=bool Use sync io for buffered writes. For the majority of the
489 io engines, this means using O_SYNC.
490
d3aad8f2 491iomem=str
71bfa161
JA
492mem=str Fio can use various types of memory as the io unit buffer.
493 The allowed values are:
494
495 malloc Use memory from malloc(3) as the buffers.
496
497 shm Use shared memory as the buffers. Allocated
498 through shmget(2).
499
74b025b0
JA
500 shmhuge Same as shm, but use huge pages as backing.
501
313cb206
JA
502 mmap Use mmap to allocate buffers. May either be
503 anonymous memory, or can be file backed if
504 a filename is given after the option. The
505 format is mem=mmap:/path/to/file.
71bfa161 506
d0bdaf49
JA
507 mmaphuge Use a memory mapped huge file as the buffer
508 backing. Append filename after mmaphuge, ala
509 mem=mmaphuge:/hugetlbfs/file
510
71bfa161 511 The area allocated is a function of the maximum allowed
5394ae5f
JA
512 bs size for the job, multiplied by the io depth given. Note
513 that for shmhuge and mmaphuge to work, the system must have
514 free huge pages allocated. This can normally be checked
515 and set by reading/writing /proc/sys/vm/nr_hugepages on a
516 Linux system. Fio assumes a huge page is 4MiB in size. So
517 to calculate the number of huge pages you need for a given
518 job file, add up the io depth of all jobs (normally one unless
519 iodepth= is used) and multiply by the maximum bs set. Then
520 divide that number by the huge page size. You can see the
521 size of the huge pages in /proc/meminfo. If no huge pages
522 are allocated by having a non-zero number in nr_hugepages,
56bb17f2 523 using mmaphuge or shmhuge will fail. Also see hugepage-size.
5394ae5f
JA
524
525 mmaphuge also needs to have hugetlbfs mounted and the file
526 location should point there. So if it's mounted in /huge,
527 you would use mem=mmaphuge:/huge/somefile.
71bfa161 528
56bb17f2
JA
529hugepage-size=siint
530 Defines the size of a huge page. Must at least be equal
531 to the system setting, see /proc/meminfo. Defaults to 4MiB.
c51074e7
JA
532 Should probably always be a multiple of megabytes, so using
533 hugepage-size=Xm is the preferred way to set this to avoid
534 setting a non-pow-2 bad value.
56bb17f2 535
71bfa161
JA
536exitall When one job finishes, terminate the rest. The default is
537 to wait for each job to finish, sometimes that is not the
538 desired action.
539
540bwavgtime=int Average the calculated bandwidth over the given time. Value
6c219763 541 is specified in milliseconds.
71bfa161
JA
542
543create_serialize=bool If true, serialize the file creating for the jobs.
544 This may be handy to avoid interleaving of data
545 files, which may greatly depend on the filesystem
546 used and even the number of processors in the system.
547
548create_fsync=bool fsync the data file after creation. This is the
549 default.
550
e545a6ce
JA
551unlink=bool Unlink the job files when done. Not the default, as repeated
552 runs of that job would then waste time recreating the fileset
553 again and again.
71bfa161
JA
554
555loops=int Run the specified number of iterations of this job. Used
556 to repeat the same workload a given number of times. Defaults
557 to 1.
558
559verify=str If writing to a file, fio can verify the file contents
560 after each iteration of the job. The allowed values are:
561
562 md5 Use an md5 sum of the data area and store
563 it in the header of each block.
564
565 crc32 Use a crc32 sum of the data area and store
566 it in the header of each block.
567
36690c9b
JA
568 null Only pretend to verify. Useful for testing
569 internals with ioengine=null, not for much
570 else.
571
6c219763 572 This option can be used for repeated burn-in tests of a
71bfa161
JA
573 system to make sure that the written data is also
574 correctly read back.
575
160b966d
JA
576verifysort=bool If set, fio will sort written verify blocks when it deems
577 it faster to read them back in a sorted manner. This is
578 often the case when overwriting an existing file, since
579 the blocks are already laid out in the file system. You
580 can ignore this option unless doing huge amounts of really
581 fast IO where the red-black tree sorting CPU time becomes
582 significant.
583
71bfa161
JA
584stonewall Wait for preceeding jobs in the job file to exit, before
585 starting this one. Can be used to insert serialization
b3d62a75
JA
586 points in the job file. A stone wall also implies starting
587 a new reporting group.
588
589new_group Start a new reporting group. If this option isn't given,
590 jobs in a file will be part of the same reporting group
591 unless seperated by a stone wall (or if it's a group
592 by itself, with the numjobs option).
71bfa161
JA
593
594numjobs=int Create the specified number of clones of this job. May be
595 used to setup a larger number of threads/processes doing
fa28c85a
JA
596 the same thing. We regard that grouping of jobs as a
597 specific group.
598
599group_reporting If 'numjobs' is set, it may be interesting to display
600 statistics for the group as a whole instead of for each
601 individual job. This is especially true of 'numjobs' is
602 large, looking at individual thread/process output quickly
603 becomes unwieldy. If 'group_reporting' is specified, fio
604 will show the final report per-group instead of per-job.
71bfa161
JA
605
606thread fio defaults to forking jobs, however if this option is
607 given, fio will use pthread_create(3) to create threads
608 instead.
609
610zonesize=siint Divide a file into zones of the specified size. See zoneskip.
611
612zoneskip=siint Skip the specified number of bytes when zonesize data has
613 been read. The two zone options can be used to only do
614 io on zones of a file.
615
076efc7c
JA
616write_iolog=str Write the issued io patterns to the specified file. See
617 read_iolog.
71bfa161 618
076efc7c 619read_iolog=str Open an iolog with the specified file name and replay the
71bfa161
JA
620 io patterns it contains. This can be used to store a
621 workload and replay it sometime later.
622
623write_bw_log If given, write a bandwidth log of the jobs in this job
624 file. Can be used to store data of the bandwidth of the
e0da9bc2
JA
625 jobs in their lifetime. The included fio_generate_plots
626 script uses gnuplot to turn these text files into nice
627 graphs.
71bfa161
JA
628
629write_lat_log Same as write_bw_log, except that this option stores io
630 completion latencies instead.
631
632lockmem=siint Pin down the specified amount of memory with mlock(2). Can
633 potentially be used instead of removing memory or booting
634 with less memory to simulate a smaller amount of memory.
635
636exec_prerun=str Before running this job, issue the command specified
637 through system(3).
638
639exec_postrun=str After the job completes, issue the command specified
640 though system(3).
641
642ioscheduler=str Attempt to switch the device hosting the file to the specified
643 io scheduler before running.
644
645cpuload=int If the job is a CPU cycle eater, attempt to use the specified
646 percentage of CPU cycles.
647
648cpuchunks=int If the job is a CPU cycle eater, split the load into
6c219763 649 cycles of the given time. In milliseconds.
71bfa161 650
0a839f30
JA
651disk_util=bool Generate disk utilization statistics, if the platform
652 supports it. Defaults to on.
653
71bfa161
JA
654
6556.0 Interpreting the output
656---------------------------
657
658fio spits out a lot of output. While running, fio will display the
659status of the jobs created. An example of that would be:
660
73c8b082 661Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
71bfa161
JA
662
663The characters inside the square brackets denote the current status of
664each thread. The possible values (in typical life cycle order) are:
665
666Idle Run
667---- ---
668P Thread setup, but not started.
669C Thread created.
670I Thread initialized, waiting.
671 R Running, doing sequential reads.
672 r Running, doing random reads.
673 W Running, doing sequential writes.
674 w Running, doing random writes.
675 M Running, doing mixed sequential reads/writes.
676 m Running, doing mixed random reads/writes.
677 F Running, currently waiting for fsync()
678V Running, doing verification of written data.
679E Thread exited, not reaped by main thread yet.
680_ Thread reaped.
681
682The other values are fairly self explanatory - number of threads
6043c579
JA
683currently running and doing io, rate of io since last check, and the estimated
684completion percentage and time for the running group. It's impossible to
685estimate runtime of the following groups (if any).
71bfa161
JA
686
687When fio is done (or interrupted by ctrl-c), it will show the data for
688each thread, group of threads, and disks in that order. For each data
689direction, the output looks like:
690
691Client1 (g=0): err= 0:
692 write: io= 32MiB, bw= 666KiB/s, runt= 50320msec
6104ddb6
JA
693 slat (msec): min= 0, max= 136, avg= 0.03, stdev= 1.92
694 clat (msec): min= 0, max= 631, avg=48.50, stdev=86.82
695 bw (KiB/s) : min= 0, max= 1196, per=51.00%, avg=664.02, stdev=681.68
71bfa161 696 cpu : usr=1.49%, sys=0.25%, ctx=7969
71619dc2 697 IO depths : 1=0.1%, 2=0.3%, 4=0.5%, 8=99.0%, 16=0.0%, 32=0.0%, >32=0.0%
30061b97 698 issued r/w: total=0/32768, short=0/0
8abdce66
JA
699 lat (msec): 2=1.6%, 4=0.0%, 10=3.2%, 20=12.8%, 50=38.4%, 100=24.8%,
700 lat (msec): 250=15.2%, 500=0.0%, 750=0.0%, 1000=0.0%, >=2048=0.0%
71bfa161
JA
701
702The client number is printed, along with the group id and error of that
703thread. Below is the io statistics, here for writes. In the order listed,
704they denote:
705
706io= Number of megabytes io performed
707bw= Average bandwidth rate
708runt= The runtime of that thread
72fbda2a 709 slat= Submission latency (avg being the average, stdev being the
71bfa161
JA
710 standard deviation). This is the time it took to submit
711 the io. For sync io, the slat is really the completion
712 latency, since queue/complete is one operation there.
713 clat= Completion latency. Same names as slat, this denotes the
714 time from submission to completion of the io pieces. For
715 sync io, clat will usually be equal (or very close) to 0,
716 as the time from submit to complete is basically just
717 CPU time (io has already been done, see slat explanation).
718 bw= Bandwidth. Same names as the xlat stats, but also includes
719 an approximate percentage of total aggregate bandwidth
720 this thread received in this group. This last value is
721 only really useful if the threads in this group are on the
722 same disk, since they are then competing for disk access.
723cpu= CPU usage. User and system time, along with the number
724 of context switches this thread went through.
71619dc2
JA
725IO depths= The distribution of io depths over the job life time. The
726 numbers are divided into powers of 2, so for example the
727 16= entries includes depths up to that value but higher
728 than the previous entry. In other words, it covers the
729 range from 16 to 31.
30061b97
JA
730IO issued= The number of read/write requests issued, and how many
731 of them were short.
ec118304
JA
732IO latencies= The distribution of IO completion latencies. This is the
733 time from when IO leaves fio and when it gets completed.
734 The numbers follow the same pattern as the IO depths,
735 meaning that 2=1.6% means that 1.6% of the IO completed
8abdce66
JA
736 within 2 msecs, 20=12.8% means that 12.8% of the IO
737 took more than 10 msecs, but less than (or equal to) 20 msecs.
71bfa161
JA
738
739After each client has been listed, the group statistics are printed. They
740will look like this:
741
742Run status group 0 (all jobs):
743 READ: io=64MiB, aggrb=22178, minb=11355, maxb=11814, mint=2840msec, maxt=2955msec
744 WRITE: io=64MiB, aggrb=1302, minb=666, maxb=669, mint=50093msec, maxt=50320msec
745
746For each data direction, it prints:
747
748io= Number of megabytes io performed.
749aggrb= Aggregate bandwidth of threads in this group.
750minb= The minimum average bandwidth a thread saw.
751maxb= The maximum average bandwidth a thread saw.
752mint= The smallest runtime of the threads in that group.
753maxt= The longest runtime of the threads in that group.
754
755And finally, the disk statistics are printed. They will look like this:
756
757Disk stats (read/write):
758 sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00%
759
760Each value is printed for both reads and writes, with reads first. The
761numbers denote:
762
763ios= Number of ios performed by all groups.
764merge= Number of merges io the io scheduler.
765ticks= Number of ticks we kept the disk busy.
766io_queue= Total time spent in the disk queue.
767util= The disk utilization. A value of 100% means we kept the disk
768 busy constantly, 50% would be a disk idling half of the time.
769
770
7717.0 Terse output
772----------------
773
774For scripted usage where you typically want to generate tables or graphs
6af019c9 775of the results, fio can output the results in a semicolon separated format.
71bfa161
JA
776The format is one long line of values, such as:
777
6af019c9
JA
778client1;0;0;1906777;1090804;1790;0;0;0.000000;0.000000;0;0;0.000000;0.000000;929380;1152890;25.510151%;1078276.333333;128948.113404;0;0;0;0;0;0.000000;0.000000;0;0;0.000000;0.000000;0;0;0.000000%;0.000000;0.000000;100.000000%;0.000000%;324;100.0%;0.0%;0.0%;0.0%;0.0%;0.0%;0.0%;100.0%;0.0%;0.0%;0.0%;0.0%;0.0%
779;0.0%;0.0%;0.0%;0.0%;0.0%
71bfa161
JA
780
781Split up, the format is as follows:
782
783 jobname, groupid, error
784 READ status:
785 KiB IO, bandwidth (KiB/sec), runtime (msec)
786 Submission latency: min, max, mean, deviation
787 Completion latency: min, max, mean, deviation
6c219763 788 Bw: min, max, aggregate percentage of total, mean, deviation
71bfa161
JA
789 WRITE status:
790 KiB IO, bandwidth (KiB/sec), runtime (msec)
791 Submission latency: min, max, mean, deviation
792 Completion latency: min, max, mean, deviation
6c219763 793 Bw: min, max, aggregate percentage of total, mean, deviation
71bfa161 794 CPU usage: user, system, context switches
2270890c
JA
795 IO depths: <=1, 2, 4, 8, 16, 32, >=64
796 IO latencies: <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, >=2000
797 Text description
71bfa161 798