8 5. Detailed list of parameters
12 9. CPU idleness profiling
14 1.0 Overview and history
15 ------------------------
16 fio was originally written to save me the hassle of writing special test
17 case programs when I wanted to test a specific workload, either for
18 performance reasons or to find/reproduce a bug. The process of writing
19 such a test app can be tiresome, especially if you have to do it often.
20 Hence I needed a tool that would be able to simulate a given io workload
21 without resorting to writing a tailored test case again and again.
23 A test work load is difficult to define, though. There can be any number
24 of processes or threads involved, and they can each be using their own
25 way of generating io. You could have someone dirtying large amounts of
26 memory in an memory mapped file, or maybe several threads issuing
27 reads using asynchronous io. fio needed to be flexible enough to
28 simulate both of these cases, and many more.
32 The first step in getting fio to simulate a desired io workload, is
33 writing a job file describing that specific setup. A job file may contain
34 any number of threads and/or files - the typical contents of the job file
35 is a global section defining shared parameters, and one or more job
36 sections describing the jobs involved. When run, fio parses this file
37 and sets everything up as described. If we break down a job from top to
38 bottom, it contains the following basic parameters:
40 IO type Defines the io pattern issued to the file(s).
41 We may only be reading sequentially from this
42 file(s), or we may be writing randomly. Or even
43 mixing reads and writes, sequentially or randomly.
45 Block size In how large chunks are we issuing io? This may be
46 a single value, or it may describe a range of
49 IO size How much data are we going to be reading/writing.
51 IO engine How do we issue io? We could be memory mapping the
52 file, we could be using regular read/write, we
53 could be using splice, async io, syslet, or even
56 IO depth If the io engine is async, how large a queuing
57 depth do we want to maintain?
59 IO type Should we be doing buffered io, or direct/raw io?
61 Num files How many files are we spreading the workload over.
63 Num threads How many threads or processes should we spread
66 The above are the basic parameters defined for a workload, in addition
67 there's a multitude of parameters that modify other aspects of how this
73 See the README file for command line parameters, there are only a few
76 Running fio is normally the easiest part - you just give it the job file
77 (or job files) as parameters:
81 and it will start doing what the job_file tells it to do. You can give
82 more than one job file on the command line, fio will serialize the running
83 of those files. Internally that is the same as using the 'stonewall'
84 parameter described in the parameter section.
86 If the job file contains only one job, you may as well just give the
87 parameters on the command line. The command line parameters are identical
88 to the job parameters, with a few extra that control global parameters
89 (see README). For example, for the job file parameter iodepth=2, the
90 mirror command line option would be --iodepth 2 or --iodepth=2. You can
91 also use the command line for giving more than one job entry. For each
92 --name option that fio sees, it will start a new job with that name.
93 Command line entries following a --name entry will apply to that job,
94 until there are no more entries or a new --name entry is seen. This is
95 similar to the job file options, where each option applies to the current
96 job until a new [] job entry is seen.
98 fio does not need to run as root, except if the files or devices specified
99 in the job section requires that. Some other options may also be restricted,
100 such as memory locking, io scheduler switching, and decreasing the nice value.
105 As previously described, fio accepts one or more job files describing
106 what it is supposed to do. The job file format is the classic ini file,
107 where the names enclosed in [] brackets define the job name. You are free
108 to use any ascii name you want, except 'global' which has special meaning.
109 A global section sets defaults for the jobs described in that file. A job
110 may override a global section parameter, and a job file may even have
111 several global sections if so desired. A job is only affected by a global
112 section residing above it. If the first character in a line is a ';' or a
113 '#', the entire line is discarded as a comment.
115 So let's look at a really simple job file that defines two processes, each
116 randomly reading from a 128MB file.
118 ; -- start job file --
129 As you can see, the job file sections themselves are empty as all the
130 described parameters are shared. As no filename= option is given, fio
131 makes up a filename for each of the jobs as it sees fit. On the command
132 line, this job would look as follows:
134 $ fio --name=global --rw=randread --size=128m --name=job1 --name=job2
137 Let's look at an example that has a number of processes writing randomly
140 ; -- start job file --
152 Here we have no global section, as we only have one job defined anyway.
153 We want to use async io here, with a depth of 4 for each file. We also
154 increased the buffer size used to 32KB and define numjobs to 4 to
155 fork 4 identical jobs. The result is 4 processes each randomly writing
156 to their own 64MB file. Instead of using the above job file, you could
157 have given the parameters on the command line. For this case, you would
160 $ fio --name=random-writers --ioengine=libaio --iodepth=4 --rw=randwrite --bs=32k --direct=0 --size=64m --numjobs=4
162 When fio is utilized as a basis of any reasonably large test suite, it might be
163 desirable to share a set of standardized settings across multiple job files.
164 Instead of copy/pasting such settings, any section may pull in an external
165 .fio file with 'include filename' directive, as in the following example:
167 ; -- start job file including.fio --
171 include glob-include.fio
178 include test-include.fio
179 ; -- end job file including.fio --
181 ; -- start job file glob-include.fio --
184 ; -- end job file glob-include.fio --
186 ; -- start job file test-include.fio --
189 ; -- end job file test-include.fio --
191 Settings pulled into a section apply to that section only (except global
192 section). Include directives may be nested in that any included file may
193 contain further include directive(s). Include files may not contain []
197 4.1 Environment variables
198 -------------------------
200 fio also supports environment variable expansion in job files. Any
201 substring of the form "${VARNAME}" as part of an option value (in other
202 words, on the right of the `='), will be expanded to the value of the
203 environment variable called VARNAME. If no such environment variable
204 is defined, or VARNAME is the empty string, the empty string will be
207 As an example, let's look at a sample fio invocation and job file:
209 $ SIZE=64m NUMJOBS=4 fio jobfile.fio
211 ; -- start job file --
218 This will expand to the following equivalent job file at runtime:
220 ; -- start job file --
227 fio ships with a few example job files, you can also look there for
230 4.2 Reserved keywords
231 ---------------------
233 Additionally, fio has a set of reserved keywords that will be replaced
234 internally with the appropriate value. Those keywords are:
236 $pagesize The architecture page size of the running system
237 $mb_memory Megabytes of total memory in the system
238 $ncpus Number of online available CPUs
240 These can be used on the command line or in the job file, and will be
241 automatically substituted with the current system values when the job
242 is run. Simple math is also supported on these keywords, so you can
243 perform actions like:
247 and get that properly expanded to 8 times the size of memory in the
251 5.0 Detailed list of parameters
252 -------------------------------
254 This section describes in details each parameter associated with a job.
255 Some parameters take an option of a given type, such as an integer or
256 a string. Anywhere a numeric value is required, an arithmetic expression
257 may be used, provided it is surrounded by parentheses. Supported operators
267 For time values in expressions, units are microseconds by default. This is
268 different than for time values not in expressions (not enclosed in
269 parentheses). The following types are used:
271 str String. This is a sequence of alpha characters.
272 time Integer with possible time suffix. In seconds unless otherwise
273 specified, use eg 10m for 10 minutes. Accepts s/m/h for seconds,
274 minutes, and hours, and accepts 'ms' (or 'msec') for milliseconds,
275 and 'us' (or 'usec') for microseconds.
276 int SI integer. A whole number value, which may contain a suffix
277 describing the base of the number. Accepted suffixes are k/m/g/t/p,
278 meaning kilo, mega, giga, tera, and peta. The suffix is not case
279 sensitive, and you may also include trailing 'b' (eg 'kb' is the same
280 as 'k'). So if you want to specify 4096, you could either write
281 out '4096' or just give 4k. The suffixes signify base 2 values, so
282 1024 is 1k and 1024k is 1m and so on, unless the suffix is explicitly
283 set to a base 10 value using 'kib', 'mib', 'gib', etc. If that is the
284 case, then 1000 is used as the multiplier. This can be handy for
285 disks, since manufacturers generally use base 10 values when listing
286 the capacity of a drive. If the option accepts an upper and lower
287 range, use a colon ':' or minus '-' to separate such values. May also
288 include a prefix to indicate numbers base. If 0x is used, the number
289 is assumed to be hexadecimal. See irange.
290 bool Boolean. Usually parsed as an integer, however only defined for
291 true and false (1 and 0).
292 irange Integer range with suffix. Allows value range to be given, such
293 as 1024-4096. A colon may also be used as the separator, eg
294 1k:4k. If the option allows two sets of ranges, they can be
295 specified with a ',' or '/' delimiter: 1k-4k/8k-32k. Also see
297 float_list A list of floating numbers, separated by a ':' character.
299 With the above in mind, here follows the complete list of fio job
302 name=str ASCII name of the job. This may be used to override the
303 name printed by fio for this job. Otherwise the job
304 name is used. On the command line this parameter has the
305 special purpose of also signaling the start of a new
308 description=str Text description of the job. Doesn't do anything except
309 dump this text description when this job is run. It's
312 directory=str Prefix filenames with this directory. Used to place files
313 in a different location than "./". See the 'filename' option
314 for escaping certain characters.
316 filename=str Fio normally makes up a filename based on the job name,
317 thread number, and file number. If you want to share
318 files between threads in a job or several jobs, specify
319 a filename for each of them to override the default. If
320 the ioengine used is 'net', the filename is the host, port,
321 and protocol to use in the format of =host,port,protocol.
322 See ioengine=net for more. If the ioengine is file based, you
323 can specify a number of files by separating the names with a
324 ':' colon. So if you wanted a job to open /dev/sda and /dev/sdb
325 as the two working files, you would use
326 filename=/dev/sda:/dev/sdb. On Windows, disk devices are
327 accessed as \\.\PhysicalDrive0 for the first device,
328 \\.\PhysicalDrive1 for the second etc. Note: Windows and
329 FreeBSD prevent write access to areas of the disk containing
330 in-use data (e.g. filesystems).
331 If the wanted filename does need to include a colon, then
332 escape that with a '\' character. For instance, if the filename
333 is "/dev/dsk/foo@3,0:c", then you would use
334 filename="/dev/dsk/foo@3,0\:c". '-' is a reserved name, meaning
335 stdin or stdout. Which of the two depends on the read/write
339 If sharing multiple files between jobs, it is usually necessary
340 to have fio generate the exact names that you want. By default,
341 fio will name a file based on the default file format
342 specification of jobname.jobnumber.filenumber. With this
343 option, that can be customized. Fio will recognize and replace
344 the following keywords in this string:
347 The name of the worker thread or process.
350 The incremental number of the worker thread or
354 The incremental number of the file for that worker
357 To have dependent jobs share a set of files, this option can
358 be set to have fio generate filenames that are shared between
359 the two. For instance, if testfiles.$filenum is specified,
360 file number 4 for any job will be named testfiles.4. The
361 default of $jobname.$jobnum.$filenum will be used if
362 no other format specifier is given.
364 opendir=str Tell fio to recursively add any file it can find in this
365 directory and down the file system tree.
367 lockfile=str Fio defaults to not locking any files before it does
368 IO to them. If a file or file descriptor is shared, fio
369 can serialize IO to that file to make the end result
370 consistent. This is usual for emulating real workloads that
371 share files. The lock modes are:
373 none No locking. The default.
374 exclusive Only one thread/process may do IO,
375 excluding all others.
376 readwrite Read-write locking on the file. Many
377 readers may access the file at the
378 same time, but writes get exclusive
382 rw=str Type of io pattern. Accepted values are:
384 read Sequential reads
385 write Sequential writes
386 randwrite Random writes
387 randread Random reads
388 rw,readwrite Sequential mixed reads and writes
389 randrw Random mixed reads and writes
390 trimwrite Mixed trims and writes. Blocks will be
391 trimmed first, then written to.
393 For the mixed io types, the default is to split them 50/50.
394 For certain types of io the result may still be skewed a bit,
395 since the speed may be different. It is possible to specify
396 a number of IO's to do before getting a new offset, this is
397 done by appending a ':<nr>' to the end of the string given.
398 For a random read, it would look like 'rw=randread:8' for
399 passing in an offset modifier with a value of 8. If the
400 suffix is used with a sequential IO pattern, then the value
401 specified will be added to the generated offset for each IO.
402 For instance, using rw=write:4k will skip 4k for every
403 write. It turns sequential IO into sequential IO with holes.
404 See the 'rw_sequencer' option.
406 rw_sequencer=str If an offset modifier is given by appending a number to
407 the rw=<str> line, then this option controls how that
408 number modifies the IO offset being generated. Accepted
411 sequential Generate sequential offset
412 identical Generate the same offset
414 'sequential' is only useful for random IO, where fio would
415 normally generate a new random offset for every IO. If you
416 append eg 8 to randread, you would get a new random offset for
417 every 8 IO's. The result would be a seek for only every 8
418 IO's, instead of for every IO. Use rw=randread:8 to specify
419 that. As sequential IO is already sequential, setting
420 'sequential' for that would not result in any differences.
421 'identical' behaves in a similar fashion, except it sends
422 the same offset 8 number of times before generating a new
425 kb_base=int The base unit for a kilobyte. The defacto base is 2^10, 1024.
426 Storage manufacturers like to use 10^3 or 1000 as a base
427 ten unit instead, for obvious reasons. Allow values are
428 1024 or 1000, with 1024 being the default.
430 unified_rw_reporting=bool Fio normally reports statistics on a per
431 data direction basis, meaning that read, write, and trim are
432 accounted and reported separately. If this option is set,
433 the fio will sum the results and report them as "mixed"
436 randrepeat=bool For random IO workloads, seed the generator in a predictable
437 way so that results are repeatable across repetitions.
439 randseed=int Seed the random number generators based on this seed value, to
440 be able to control what sequence of output is being generated.
441 If not set, the random sequence depends on the randrepeat
444 fallocate=str Whether pre-allocation is performed when laying down files.
447 none Do not pre-allocate space
448 posix Pre-allocate via posix_fallocate()
449 keep Pre-allocate via fallocate() with
450 FALLOC_FL_KEEP_SIZE set
451 0 Backward-compatible alias for 'none'
452 1 Backward-compatible alias for 'posix'
454 May not be available on all supported platforms. 'keep' is only
455 available on Linux.If using ZFS on Solaris this must be set to
456 'none' because ZFS doesn't support it. Default: 'posix'.
458 fadvise_hint=bool By default, fio will use fadvise() to advise the kernel
459 on what IO patterns it is likely to issue. Sometimes you
460 want to test specific IO patterns without telling the
461 kernel about it, in which case you can disable this option.
462 If set, fio will use POSIX_FADV_SEQUENTIAL for sequential
463 IO and POSIX_FADV_RANDOM for random IO.
465 fadvise_stream=int Notify the kernel what write stream ID to place these
466 writes under. Only supported on Linux. Note, this option
467 may change going forward.
469 size=int The total size of file io for this job. Fio will run until
470 this many bytes has been transferred, unless runtime is
471 limited by other options (such as 'runtime', for instance,
472 or increased/decreased by 'io_size'). Unless specific nrfiles
473 and filesize options are given, fio will divide this size
474 between the available files specified by the job. If not set,
475 fio will use the full size of the given files or devices.
476 If the files do not exist, size must be given. It is also
477 possible to give size as a percentage between 1 and 100. If
478 size=20% is given, fio will use 20% of the full size of the
479 given files or devices.
482 io_limit=int Normally fio operates within the region set by 'size', which
483 means that the 'size' option sets both the region and size of
484 IO to be performed. Sometimes that is not what you want. With
485 this option, it is possible to define just the amount of IO
486 that fio should do. For instance, if 'size' is set to 20G and
487 'io_size' is set to 5G, fio will perform IO within the first
488 20G but exit when 5G have been done. The opposite is also
489 possible - if 'size' is set to 20G, and 'io_size' is set to
490 40G, then fio will do 40G of IO within the 0..20G region.
492 filesize=int Individual file sizes. May be a range, in which case fio
493 will select sizes for files at random within the given range
494 and limited to 'size' in total (if that is given). If not
495 given, each created file is the same size.
497 file_append=bool Perform IO after the end of the file. Normally fio will
498 operate within the size of a file. If this option is set, then
499 fio will append to the file instead. This has identical
500 behavior to setting offset to the size of a file. This option
501 is ignored on non-regular files.
504 fill_fs=bool Sets size to something really large and waits for ENOSPC (no
505 space left on device) as the terminating condition. Only makes
506 sense with sequential write. For a read workload, the mount
507 point will be filled first then IO started on the result. This
508 option doesn't make sense if operating on a raw device node,
509 since the size of that is already known by the file system.
510 Additionally, writing beyond end-of-device will not return
514 bs=int The block size used for the io units. Defaults to 4k. Values
515 can be given for both read and writes. If a single int is
516 given, it will apply to both. If a second int is specified
517 after a comma, it will apply to writes only. In other words,
518 the format is either bs=read_and_write or bs=read,write,trim.
519 bs=4k,8k will thus use 4k blocks for reads, 8k blocks for
520 writes, and 8k for trims. You can terminate the list with
521 a trailing comma. bs=4k,8k, would use the default value for
522 trims.. If you only wish to set the write size, you
523 can do so by passing an empty read size - bs=,8k will set
524 8k for writes and leave the read default value.
527 ba=int At what boundary to align random IO offsets. Defaults to
528 the same as 'blocksize' the minimum blocksize given.
529 Minimum alignment is typically 512b for using direct IO,
530 though it usually depends on the hardware block size. This
531 option is mutually exclusive with using a random map for
532 files, so it will turn off that option.
534 blocksize_range=irange
535 bsrange=irange Instead of giving a single block size, specify a range
536 and fio will mix the issued io block sizes. The issued
537 io unit will always be a multiple of the minimum value
538 given (also see bs_unaligned). Applies to both reads and
539 writes, however a second range can be given after a comma.
542 bssplit=str Sometimes you want even finer grained control of the
543 block sizes issued, not just an even split between them.
544 This option allows you to weight various block sizes,
545 so that you are able to define a specific amount of
546 block sizes issued. The format for this option is:
548 bssplit=blocksize/percentage:blocksize/percentage
550 for as many block sizes as needed. So if you want to define
551 a workload that has 50% 64k blocks, 10% 4k blocks, and
552 40% 32k blocks, you would write:
554 bssplit=4k/10:64k/50:32k/40
556 Ordering does not matter. If the percentage is left blank,
557 fio will fill in the remaining values evenly. So a bssplit
558 option like this one:
560 bssplit=4k/50:1k/:32k/
562 would have 50% 4k ios, and 25% 1k and 32k ios. The percentages
563 always add up to 100, if bssplit is given a range that adds
564 up to more, it will error out.
566 bssplit also supports giving separate splits to reads and
567 writes. The format is identical to what bs= accepts. You
568 have to separate the read and write parts with a comma. So
569 if you want a workload that has 50% 2k reads and 50% 4k reads,
570 while having 90% 4k writes and 10% 8k writes, you would
573 bssplit=2k/50:4k/50,4k/90:8k/10
576 bs_unaligned If this option is given, any byte size value within bsrange
577 may be used as a block range. This typically wont work with
578 direct IO, as that normally requires sector alignment.
580 bs_is_seq_rand If this option is set, fio will use the normal read,write
581 blocksize settings as sequential,random instead. Any random
582 read or write will use the WRITE blocksize settings, and any
583 sequential read or write will use the READ blocksize setting.
585 zero_buffers If this option is given, fio will init the IO buffers to
586 all zeroes. The default is to fill them with random data.
588 refill_buffers If this option is given, fio will refill the IO buffers
589 on every submit. The default is to only fill it at init
590 time and reuse that data. Only makes sense if zero_buffers
591 isn't specified, naturally. If data verification is enabled,
592 refill_buffers is also automatically enabled.
594 scramble_buffers=bool If refill_buffers is too costly and the target is
595 using data deduplication, then setting this option will
596 slightly modify the IO buffer contents to defeat normal
597 de-dupe attempts. This is not enough to defeat more clever
598 block compression attempts, but it will stop naive dedupe of
599 blocks. Default: true.
601 buffer_compress_percentage=int If this is set, then fio will attempt to
602 provide IO buffer content (on WRITEs) that compress to
603 the specified level. Fio does this by providing a mix of
604 random data and a fixed pattern. The fixed pattern is either
605 zeroes, or the pattern specified by buffer_pattern. If the
606 pattern option is used, it might skew the compression ratio
607 slightly. Note that this is per block size unit, for file/disk
608 wide compression level that matches this setting, you'll also
609 want to set refill_buffers.
611 buffer_compress_chunk=int See buffer_compress_percentage. This
612 setting allows fio to manage how big the ranges of random
613 data and zeroed data is. Without this set, fio will
614 provide buffer_compress_percentage of blocksize random
615 data, followed by the remaining zeroed. With this set
616 to some chunk size smaller than the block size, fio can
617 alternate random and zeroed data throughout the IO
620 buffer_pattern=str If set, fio will fill the io buffers with this
621 pattern. If not set, the contents of io buffers is defined by
622 the other options related to buffer contents. The setting can
623 be any pattern of bytes, and can be prefixed with 0x for hex
624 values. It may also be a string, where the string must then
627 dedupe_percentage=int If set, fio will generate this percentage of
628 identical buffers when writing. These buffers will be
629 naturally dedupable. The contents of the buffers depend on
630 what other buffer compression settings have been set. It's
631 possible to have the individual buffers either fully
632 compressible, or not at all. This option only controls the
633 distribution of unique buffers.
635 nrfiles=int Number of files to use for this job. Defaults to 1.
637 openfiles=int Number of files to keep open at the same time. Defaults to
638 the same as nrfiles, can be set smaller to limit the number
641 file_service_type=str Defines how fio decides which file from a job to
642 service next. The following types are defined:
644 random Just choose a file at random.
646 roundrobin Round robin over open files. This
649 sequential Finish one file before moving on to
650 the next. Multiple files can still be
651 open depending on 'openfiles'.
653 The string can have a number appended, indicating how
654 often to switch to a new file. So if option random:4 is
655 given, fio will switch to a new random file after 4 ios
658 ioengine=str Defines how the job issues io to the file. The following
661 sync Basic read(2) or write(2) io. lseek(2) is
662 used to position the io location.
664 psync Basic pread(2) or pwrite(2) io.
666 vsync Basic readv(2) or writev(2) IO.
668 psyncv Basic preadv(2) or pwritev(2) IO.
670 libaio Linux native asynchronous io. Note that Linux
671 may only support queued behaviour with
672 non-buffered IO (set direct=1 or buffered=0).
673 This engine defines engine specific options.
675 posixaio glibc posix asynchronous io.
677 solarisaio Solaris native asynchronous io.
679 windowsaio Windows native asynchronous io.
681 mmap File is memory mapped and data copied
682 to/from using memcpy(3).
684 splice splice(2) is used to transfer the data and
685 vmsplice(2) to transfer data from user
688 syslet-rw Use the syslet system calls to make
689 regular read/write async.
691 sg SCSI generic sg v3 io. May either be
692 synchronous using the SG_IO ioctl, or if
693 the target is an sg character device
694 we use read(2) and write(2) for asynchronous
697 null Doesn't transfer any data, just pretends
698 to. This is mainly used to exercise fio
699 itself and for debugging/testing purposes.
701 net Transfer over the network to given host:port.
702 Depending on the protocol used, the hostname,
703 port, listen and filename options are used to
704 specify what sort of connection to make, while
705 the protocol option determines which protocol
707 This engine defines engine specific options.
709 netsplice Like net, but uses splice/vmsplice to
710 map data and send/receive.
711 This engine defines engine specific options.
713 cpuio Doesn't transfer any data, but burns CPU
714 cycles according to the cpuload= and
715 cpucycle= options. Setting cpuload=85
716 will cause that job to do nothing but burn
717 85% of the CPU. In case of SMP machines,
718 use numjobs=<no_of_cpu> to get desired CPU
719 usage, as the cpuload only loads a single
720 CPU at the desired rate.
722 guasi The GUASI IO engine is the Generic Userspace
723 Asyncronous Syscall Interface approach
726 http://www.xmailserver.org/guasi-lib.html
728 for more info on GUASI.
730 rdma The RDMA I/O engine supports both RDMA
731 memory semantics (RDMA_WRITE/RDMA_READ) and
732 channel semantics (Send/Recv) for the
733 InfiniBand, RoCE and iWARP protocols.
735 falloc IO engine that does regular fallocate to
736 simulate data transfer as fio ioengine.
737 DDIR_READ does fallocate(,mode = keep_size,)
738 DDIR_WRITE does fallocate(,mode = 0)
739 DDIR_TRIM does fallocate(,mode = punch_hole)
741 e4defrag IO engine that does regular EXT4_IOC_MOVE_EXT
742 ioctls to simulate defragment activity in
743 request to DDIR_WRITE event
745 rbd IO engine supporting direct access to Ceph
746 Rados Block Devices (RBD) via librbd without
747 the need to use the kernel rbd driver. This
748 ioengine defines engine specific options.
750 gfapi Using Glusterfs libgfapi sync interface to
751 direct access to Glusterfs volumes without
754 gfapi_async Using Glusterfs libgfapi async interface
755 to direct access to Glusterfs volumes without
756 having to go through FUSE. This ioengine
757 defines engine specific options.
759 libhdfs Read and write through Hadoop (HDFS).
760 The 'filename' option is used to specify host,
761 port of the hdfs name-node to connect. This
762 engine interprets offsets a little
763 differently. In HDFS, files once created
764 cannot be modified. So random writes are not
765 possible. To imitate this, libhdfs engine
766 expects bunch of small files to be created
767 over HDFS, and engine will randomly pick a
768 file out of those files based on the offset
769 generated by fio backend. (see the example
770 job file to create such files, use rw=write
771 option). Please note, you might want to set
772 necessary environment variables to work with
773 hdfs/libhdfs properly.
775 mtd Read, write and erase an MTD character device
776 (e.g., /dev/mtd0). Discards are treated as
777 erases. Depending on the underlying device
778 type, the I/O may have to go in a certain
779 pattern, e.g., on NAND, writing sequentially
780 to erase blocks and discarding before
781 overwriting. The writetrim mode works well
784 external Prefix to specify loading an external
785 IO engine object file. Append the engine
786 filename, eg ioengine=external:/tmp/foo.o
787 to load ioengine foo.o in /tmp.
789 iodepth=int This defines how many io units to keep in flight against
790 the file. The default is 1 for each file defined in this
791 job, can be overridden with a larger value for higher
792 concurrency. Note that increasing iodepth beyond 1 will not
793 affect synchronous ioengines (except for small degress when
794 verify_async is in use). Even async engines may impose OS
795 restrictions causing the desired depth not to be achieved.
796 This may happen on Linux when using libaio and not setting
797 direct=1, since buffered IO is not async on that OS. Keep an
798 eye on the IO depth distribution in the fio output to verify
799 that the achieved depth is as expected. Default: 1.
801 iodepth_batch_submit=int
802 iodepth_batch=int This defines how many pieces of IO to submit at once.
803 It defaults to 1 which means that we submit each IO
804 as soon as it is available, but can be raised to submit
805 bigger batches of IO at the time.
807 iodepth_batch_complete=int This defines how many pieces of IO to retrieve
808 at once. It defaults to 1 which means that we'll ask
809 for a minimum of 1 IO in the retrieval process from
810 the kernel. The IO retrieval will go on until we
811 hit the limit set by iodepth_low. If this variable is
812 set to 0, then fio will always check for completed
813 events before queuing more IO. This helps reduce
814 IO latency, at the cost of more retrieval system calls.
816 iodepth_low=int The low water mark indicating when to start filling
817 the queue again. Defaults to the same as iodepth, meaning
818 that fio will attempt to keep the queue full at all times.
819 If iodepth is set to eg 16 and iodepth_low is set to 4, then
820 after fio has filled the queue of 16 requests, it will let
821 the depth drain down to 4 before starting to fill it again.
823 direct=bool If value is true, use non-buffered io. This is usually
824 O_DIRECT. Note that ZFS on Solaris doesn't support direct io.
825 On Windows the synchronous ioengines don't support direct io.
827 atomic=bool If value is true, attempt to use atomic direct IO. Atomic
828 writes are guaranteed to be stable once acknowledged by
829 the operating system. Only Linux supports O_ATOMIC right
832 buffered=bool If value is true, use buffered io. This is the opposite
833 of the 'direct' option. Defaults to true.
835 offset=int Start io at the given offset in the file. The data before
836 the given offset will not be touched. This effectively
837 caps the file size at real_size - offset.
839 offset_increment=int If this is provided, then the real offset becomes
840 offset + offset_increment * thread_number, where the thread
841 number is a counter that starts at 0 and is incremented for
842 each sub-job (i.e. when numjobs option is specified). This
843 option is useful if there are several jobs which are intended
844 to operate on a file in parallel disjoint segments, with
845 even spacing between the starting points.
847 number_ios=int Fio will normally perform IOs until it has exhausted the size
848 of the region set by size=, or if it exhaust the allocated
849 time (or hits an error condition). With this setting, the
850 range/size can be set independently of the number of IOs to
851 perform. When fio reaches this number, it will exit normally
852 and report status. Note that this does not extend the amount
853 of IO that will be done, it will only stop fio if this
854 condition is met before other end-of-job criteria.
856 fsync=int If writing to a file, issue a sync of the dirty data
857 for every number of blocks given. For example, if you give
858 32 as a parameter, fio will sync the file for every 32
859 writes issued. If fio is using non-buffered io, we may
860 not sync the file. The exception is the sg io engine, which
861 synchronizes the disk cache anyway.
863 fdatasync=int Like fsync= but uses fdatasync() to only sync data and not
865 In FreeBSD and Windows there is no fdatasync(), this falls back to
868 sync_file_range=str:val Use sync_file_range() for every 'val' number of
869 write operations. Fio will track range of writes that
870 have happened since the last sync_file_range() call. 'str'
871 can currently be one or more of:
873 wait_before SYNC_FILE_RANGE_WAIT_BEFORE
874 write SYNC_FILE_RANGE_WRITE
875 wait_after SYNC_FILE_RANGE_WAIT_AFTER
877 So if you do sync_file_range=wait_before,write:8, fio would
878 use SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE for
879 every 8 writes. Also see the sync_file_range(2) man page.
880 This option is Linux specific.
882 overwrite=bool If true, writes to a file will always overwrite existing
883 data. If the file doesn't already exist, it will be
884 created before the write phase begins. If the file exists
885 and is large enough for the specified write phase, nothing
888 end_fsync=bool If true, fsync file contents when a write stage has completed.
890 fsync_on_close=bool If true, fio will fsync() a dirty file on close.
891 This differs from end_fsync in that it will happen on every
892 file close, not just at the end of the job.
894 rwmixread=int How large a percentage of the mix should be reads.
896 rwmixwrite=int How large a percentage of the mix should be writes. If both
897 rwmixread and rwmixwrite is given and the values do not add
898 up to 100%, the latter of the two will be used to override
899 the first. This may interfere with a given rate setting,
900 if fio is asked to limit reads or writes to a certain rate.
901 If that is the case, then the distribution may be skewed.
903 random_distribution=str:float By default, fio will use a completely uniform
904 random distribution when asked to perform random IO. Sometimes
905 it is useful to skew the distribution in specific ways,
906 ensuring that some parts of the data is more hot than others.
907 fio includes the following distribution models:
909 random Uniform random distribution
910 zipf Zipf distribution
911 pareto Pareto distribution
913 When using a zipf or pareto distribution, an input value
914 is also needed to define the access pattern. For zipf, this
915 is the zipf theta. For pareto, it's the pareto power. Fio
916 includes a test program, genzipf, that can be used visualize
917 what the given input values will yield in terms of hit rates.
918 If you wanted to use zipf with a theta of 1.2, you would use
919 random_distribution=zipf:1.2 as the option. If a non-uniform
920 model is used, fio will disable use of the random map.
922 percentage_random=int For a random workload, set how big a percentage should
923 be random. This defaults to 100%, in which case the workload
924 is fully random. It can be set from anywhere from 0 to 100.
925 Setting it to 0 would make the workload fully sequential. Any
926 setting in between will result in a random mix of sequential
927 and random IO, at the given percentages. It is possible to
928 set different values for reads, writes, and trim. To do so,
929 simply use a comma separated list. See blocksize.
931 norandommap Normally fio will cover every block of the file when doing
932 random IO. If this option is given, fio will just get a
933 new random offset without looking at past io history. This
934 means that some blocks may not be read or written, and that
935 some blocks may be read/written more than once. If this option
936 is used with verify= and multiple blocksizes (via bsrange=),
937 only intact blocks are verified, i.e., partially-overwritten
940 softrandommap=bool See norandommap. If fio runs with the random block map
941 enabled and it fails to allocate the map, if this option is
942 set it will continue without a random block map. As coverage
943 will not be as complete as with random maps, this option is
946 random_generator=str Fio supports the following engines for generating
947 IO offsets for random IO:
949 tausworthe Strong 2^88 cycle random number generator
950 lfsr Linear feedback shift register generator
952 Tausworthe is a strong random number generator, but it
953 requires tracking on the side if we want to ensure that
954 blocks are only read or written once. LFSR guarantees
955 that we never generate the same offset twice, and it's
956 also less computationally expensive. It's not a true
957 random generator, however, though for IO purposes it's
958 typically good enough. LFSR only works with single
959 block sizes, not with workloads that use multiple block
960 sizes. If used with such a workload, fio may read or write
961 some blocks multiple times.
963 nice=int Run the job with the given nice value. See man nice(2).
965 prio=int Set the io priority value of this job. Linux limits us to
966 a positive value between 0 and 7, with 0 being the highest.
969 prioclass=int Set the io priority class. See man ionice(1).
971 thinktime=int Stall the job x microseconds after an io has completed before
972 issuing the next. May be used to simulate processing being
973 done by an application. See thinktime_blocks and
977 Only valid if thinktime is set - pretend to spend CPU time
978 doing something with the data received, before falling back
979 to sleeping for the rest of the period specified by
983 Only valid if thinktime is set - control how many blocks
984 to issue, before waiting 'thinktime' usecs. If not set,
985 defaults to 1 which will make fio wait 'thinktime' usecs
986 after every block. This effectively makes any queue depth
987 setting redundant, since no more than 1 IO will be queued
988 before we have to complete it and do our thinktime. In
989 other words, this setting effectively caps the queue depth
990 if the latter is larger.
992 rate=int Cap the bandwidth used by this job. The number is in bytes/sec,
993 the normal suffix rules apply. You can use rate=500k to limit
994 reads and writes to 500k each, or you can specify read and
995 writes separately. Using rate=1m,500k would limit reads to
996 1MB/sec and writes to 500KB/sec. Capping only reads or
997 writes can be done with rate=,500k or rate=500k,. The former
998 will only limit writes (to 500KB/sec), the latter will only
1001 ratemin=int Tell fio to do whatever it can to maintain at least this
1002 bandwidth. Failing to meet this requirement, will cause
1003 the job to exit. The same format as rate is used for
1004 read vs write separation.
1006 rate_iops=int Cap the bandwidth to this number of IOPS. Basically the same
1007 as rate, just specified independently of bandwidth. If the
1008 job is given a block size range instead of a fixed value,
1009 the smallest block size is used as the metric. The same format
1010 as rate is used for read vs write separation.
1012 rate_iops_min=int If fio doesn't meet this rate of IO, it will cause
1013 the job to exit. The same format as rate is used for read vs
1016 latency_target=int If set, fio will attempt to find the max performance
1017 point that the given workload will run at while maintaining a
1018 latency below this target. The values is given in microseconds.
1019 See latency_window and latency_percentile
1021 latency_window=int Used with latency_target to specify the sample window
1022 that the job is run at varying queue depths to test the
1023 performance. The value is given in microseconds.
1025 latency_percentile=float The percentage of IOs that must fall within the
1026 criteria specified by latency_target and latency_window. If not
1027 set, this defaults to 100.0, meaning that all IOs must be equal
1028 or below to the value set by latency_target.
1030 max_latency=int If set, fio will exit the job if it exceeds this maximum
1031 latency. It will exit with an ETIME error.
1033 ratecycle=int Average bandwidth for 'rate' and 'ratemin' over this number
1036 cpumask=int Set the CPU affinity of this job. The parameter given is a
1037 bitmask of allowed CPU's the job may run on. So if you want
1038 the allowed CPUs to be 1 and 5, you would pass the decimal
1039 value of (1 << 1 | 1 << 5), or 34. See man
1040 sched_setaffinity(2). This may not work on all supported
1041 operating systems or kernel versions. This option doesn't
1042 work well for a higher CPU count than what you can store in
1043 an integer mask, so it can only control cpus 1-32. For
1044 boxes with larger CPU counts, use cpus_allowed.
1046 cpus_allowed=str Controls the same options as cpumask, but it allows a text
1047 setting of the permitted CPUs instead. So to use CPUs 1 and
1048 5, you would specify cpus_allowed=1,5. This options also
1049 allows a range of CPUs. Say you wanted a binding to CPUs
1050 1, 5, and 8-15, you would set cpus_allowed=1,5,8-15.
1052 cpus_allowed_policy=str Set the policy of how fio distributes the CPUs
1053 specified by cpus_allowed or cpumask. Two policies are
1056 shared All jobs will share the CPU set specified.
1057 split Each job will get a unique CPU from the CPU set.
1059 'shared' is the default behaviour, if the option isn't
1060 specified. If split is specified, then fio will will assign
1061 one cpu per job. If not enough CPUs are given for the jobs
1062 listed, then fio will roundrobin the CPUs in the set.
1064 numa_cpu_nodes=str Set this job running on spcified NUMA nodes' CPUs. The
1065 arguments allow comma delimited list of cpu numbers,
1066 A-B ranges, or 'all'. Note, to enable numa options support,
1067 fio must be built on a system with libnuma-dev(el) installed.
1069 numa_mem_policy=str Set this job's memory policy and corresponding NUMA
1070 nodes. Format of the argements:
1072 `mode' is one of the following memory policy:
1073 default, prefer, bind, interleave, local
1074 For `default' and `local' memory policy, no node is
1075 needed to be specified.
1076 For `prefer', only one node is allowed.
1077 For `bind' and `interleave', it allow comma delimited
1078 list of numbers, A-B ranges, or 'all'.
1080 startdelay=time Start this job the specified number of seconds after fio
1081 has started. Only useful if the job file contains several
1082 jobs, and you want to delay starting some jobs to a certain
1085 runtime=time Tell fio to terminate processing after the specified number
1086 of seconds. It can be quite hard to determine for how long
1087 a specified job will run, so this parameter is handy to
1088 cap the total runtime to a given time.
1090 time_based If set, fio will run for the duration of the runtime
1091 specified even if the file(s) are completely read or
1092 written. It will simply loop over the same workload
1093 as many times as the runtime allows.
1095 ramp_time=time If set, fio will run the specified workload for this amount
1096 of time before logging any performance numbers. Useful for
1097 letting performance settle before logging results, thus
1098 minimizing the runtime required for stable results. Note
1099 that the ramp_time is considered lead in time for a job,
1100 thus it will increase the total runtime if a special timeout
1101 or runtime is specified.
1103 invalidate=bool Invalidate the buffer/page cache parts for this file prior
1104 to starting io. Defaults to true.
1106 sync=bool Use sync io for buffered writes. For the majority of the
1107 io engines, this means using O_SYNC.
1110 mem=str Fio can use various types of memory as the io unit buffer.
1111 The allowed values are:
1113 malloc Use memory from malloc(3) as the buffers.
1115 shm Use shared memory as the buffers. Allocated
1118 shmhuge Same as shm, but use huge pages as backing.
1120 mmap Use mmap to allocate buffers. May either be
1121 anonymous memory, or can be file backed if
1122 a filename is given after the option. The
1123 format is mem=mmap:/path/to/file.
1125 mmaphuge Use a memory mapped huge file as the buffer
1126 backing. Append filename after mmaphuge, ala
1127 mem=mmaphuge:/hugetlbfs/file
1129 The area allocated is a function of the maximum allowed
1130 bs size for the job, multiplied by the io depth given. Note
1131 that for shmhuge and mmaphuge to work, the system must have
1132 free huge pages allocated. This can normally be checked
1133 and set by reading/writing /proc/sys/vm/nr_hugepages on a
1134 Linux system. Fio assumes a huge page is 4MB in size. So
1135 to calculate the number of huge pages you need for a given
1136 job file, add up the io depth of all jobs (normally one unless
1137 iodepth= is used) and multiply by the maximum bs set. Then
1138 divide that number by the huge page size. You can see the
1139 size of the huge pages in /proc/meminfo. If no huge pages
1140 are allocated by having a non-zero number in nr_hugepages,
1141 using mmaphuge or shmhuge will fail. Also see hugepage-size.
1143 mmaphuge also needs to have hugetlbfs mounted and the file
1144 location should point there. So if it's mounted in /huge,
1145 you would use mem=mmaphuge:/huge/somefile.
1147 iomem_align=int This indiciates the memory alignment of the IO memory buffers.
1148 Note that the given alignment is applied to the first IO unit
1149 buffer, if using iodepth the alignment of the following buffers
1150 are given by the bs used. In other words, if using a bs that is
1151 a multiple of the page sized in the system, all buffers will
1152 be aligned to this value. If using a bs that is not page
1153 aligned, the alignment of subsequent IO memory buffers is the
1154 sum of the iomem_align and bs used.
1157 Defines the size of a huge page. Must at least be equal
1158 to the system setting, see /proc/meminfo. Defaults to 4MB.
1159 Should probably always be a multiple of megabytes, so using
1160 hugepage-size=Xm is the preferred way to set this to avoid
1161 setting a non-pow-2 bad value.
1163 exitall When one job finishes, terminate the rest. The default is
1164 to wait for each job to finish, sometimes that is not the
1167 bwavgtime=int Average the calculated bandwidth over the given time. Value
1168 is specified in milliseconds.
1170 iopsavgtime=int Average the calculated IOPS over the given time. Value
1171 is specified in milliseconds.
1173 create_serialize=bool If true, serialize the file creating for the jobs.
1174 This may be handy to avoid interleaving of data
1175 files, which may greatly depend on the filesystem
1176 used and even the number of processors in the system.
1178 create_fsync=bool fsync the data file after creation. This is the
1181 create_on_open=bool Don't pre-setup the files for IO, just create open()
1182 when it's time to do IO to that file.
1184 create_only=bool If true, fio will only run the setup phase of the job.
1185 If files need to be laid out or updated on disk, only
1186 that will be done. The actual job contents are not
1189 pre_read=bool If this is given, files will be pre-read into memory before
1190 starting the given IO operation. This will also clear
1191 the 'invalidate' flag, since it is pointless to pre-read
1192 and then drop the cache. This will only work for IO engines
1193 that are seekable, since they allow you to read the same data
1194 multiple times. Thus it will not work on eg network or splice
1197 unlink=bool Unlink the job files when done. Not the default, as repeated
1198 runs of that job would then waste time recreating the file
1199 set again and again.
1201 loops=int Run the specified number of iterations of this job. Used
1202 to repeat the same workload a given number of times. Defaults
1205 verify_only Do not perform specified workload---only verify data still
1206 matches previous invocation of this workload. This option
1207 allows one to check data multiple times at a later date
1208 without overwriting it. This option makes sense only for
1209 workloads that write data, and does not support workloads
1210 with the time_based option set.
1212 do_verify=bool Run the verify phase after a write phase. Only makes sense if
1213 verify is set. Defaults to 1.
1215 verify=str If writing to a file, fio can verify the file contents
1216 after each iteration of the job. The allowed values are:
1218 md5 Use an md5 sum of the data area and store
1219 it in the header of each block.
1221 crc64 Use an experimental crc64 sum of the data
1222 area and store it in the header of each
1225 crc32c Use a crc32c sum of the data area and store
1226 it in the header of each block.
1228 crc32c-intel Use hardware assisted crc32c calcuation
1229 provided on SSE4.2 enabled processors. Falls
1230 back to regular software crc32c, if not
1231 supported by the system.
1233 crc32 Use a crc32 sum of the data area and store
1234 it in the header of each block.
1236 crc16 Use a crc16 sum of the data area and store
1237 it in the header of each block.
1239 crc7 Use a crc7 sum of the data area and store
1240 it in the header of each block.
1242 xxhash Use xxhash as the checksum function. Generally
1243 the fastest software checksum that fio
1246 sha512 Use sha512 as the checksum function.
1248 sha256 Use sha256 as the checksum function.
1250 sha1 Use optimized sha1 as the checksum function.
1252 meta Write extra information about each io
1253 (timestamp, block number etc.). The block
1254 number is verified. The io sequence number is
1255 verified for workloads that write data.
1256 See also verify_pattern.
1258 null Only pretend to verify. Useful for testing
1259 internals with ioengine=null, not for much
1262 This option can be used for repeated burn-in tests of a
1263 system to make sure that the written data is also
1264 correctly read back. If the data direction given is
1265 a read or random read, fio will assume that it should
1266 verify a previously written file. If the data direction
1267 includes any form of write, the verify will be of the
1270 verifysort=bool If set, fio will sort written verify blocks when it deems
1271 it faster to read them back in a sorted manner. This is
1272 often the case when overwriting an existing file, since
1273 the blocks are already laid out in the file system. You
1274 can ignore this option unless doing huge amounts of really
1275 fast IO where the red-black tree sorting CPU time becomes
1278 verify_offset=int Swap the verification header with data somewhere else
1279 in the block before writing. Its swapped back before
1282 verify_interval=int Write the verification header at a finer granularity
1283 than the blocksize. It will be written for chunks the
1284 size of header_interval. blocksize should divide this
1287 verify_pattern=str If set, fio will fill the io buffers with this
1288 pattern. Fio defaults to filling with totally random
1289 bytes, but sometimes it's interesting to fill with a known
1290 pattern for io verification purposes. Depending on the
1291 width of the pattern, fio will fill 1/2/3/4 bytes of the
1292 buffer at the time(it can be either a decimal or a hex number).
1293 The verify_pattern if larger than a 32-bit quantity has to
1294 be a hex number that starts with either "0x" or "0X". Use
1297 verify_fatal=bool Normally fio will keep checking the entire contents
1298 before quitting on a block verification failure. If this
1299 option is set, fio will exit the job on the first observed
1302 verify_dump=bool If set, dump the contents of both the original data
1303 block and the data block we read off disk to files. This
1304 allows later analysis to inspect just what kind of data
1305 corruption occurred. Off by default.
1307 verify_async=int Fio will normally verify IO inline from the submitting
1308 thread. This option takes an integer describing how many
1309 async offload threads to create for IO verification instead,
1310 causing fio to offload the duty of verifying IO contents
1311 to one or more separate threads. If using this offload
1312 option, even sync IO engines can benefit from using an
1313 iodepth setting higher than 1, as it allows them to have
1314 IO in flight while verifies are running.
1316 verify_async_cpus=str Tell fio to set the given CPU affinity on the
1317 async IO verification threads. See cpus_allowed for the
1320 verify_backlog=int Fio will normally verify the written contents of a
1321 job that utilizes verify once that job has completed. In
1322 other words, everything is written then everything is read
1323 back and verified. You may want to verify continually
1324 instead for a variety of reasons. Fio stores the meta data
1325 associated with an IO block in memory, so for large
1326 verify workloads, quite a bit of memory would be used up
1327 holding this meta data. If this option is enabled, fio
1328 will write only N blocks before verifying these blocks.
1330 verify_backlog_batch=int Control how many blocks fio will verify
1331 if verify_backlog is set. If not set, will default to
1332 the value of verify_backlog (meaning the entire queue
1333 is read back and verified). If verify_backlog_batch is
1334 less than verify_backlog then not all blocks will be verified,
1335 if verify_backlog_batch is larger than verify_backlog, some
1336 blocks will be verified more than once.
1338 verify_state_save=bool When a job exits during the write phase of a verify
1339 workload, save its current state. This allows fio to replay
1340 up until that point, if the verify state is loaded for the
1341 verify read phase. The format of the filename is, roughly,
1342 <type>-<jobname>-<jobindex>-verify.state. <type> is "local"
1343 for a local run, "sock" for a client/server socket connection,
1344 and "ip" (192.168.0.1, for instance) for a networked
1345 client/server connection.
1347 verify_state_load=bool If a verify termination trigger was used, fio stores
1348 the current write state of each thread. This can be used at
1349 verification time so that fio knows how far it should verify.
1350 Without this information, fio will run a full verification
1351 pass, according to the settings in the job file used.
1354 wait_for_previous Wait for preceding jobs in the job file to exit, before
1355 starting this one. Can be used to insert serialization
1356 points in the job file. A stone wall also implies starting
1357 a new reporting group.
1359 new_group Start a new reporting group. See: group_reporting.
1361 numjobs=int Create the specified number of clones of this job. May be
1362 used to setup a larger number of threads/processes doing
1363 the same thing. Each thread is reported separately; to see
1364 statistics for all clones as a whole, use group_reporting in
1365 conjunction with new_group.
1367 group_reporting It may sometimes be interesting to display statistics for
1368 groups of jobs as a whole instead of for each individual job.
1369 This is especially true if 'numjobs' is used; looking at
1370 individual thread/process output quickly becomes unwieldy.
1371 To see the final report per-group instead of per-job, use
1372 'group_reporting'. Jobs in a file will be part of the same
1373 reporting group, unless if separated by a stonewall, or by
1376 thread fio defaults to forking jobs, however if this option is
1377 given, fio will use pthread_create(3) to create threads
1380 zonesize=int Divide a file into zones of the specified size. See zoneskip.
1382 zoneskip=int Skip the specified number of bytes when zonesize data has
1383 been read. The two zone options can be used to only do
1384 io on zones of a file.
1386 write_iolog=str Write the issued io patterns to the specified file. See
1387 read_iolog. Specify a separate file for each job, otherwise
1388 the iologs will be interspersed and the file may be corrupt.
1390 read_iolog=str Open an iolog with the specified file name and replay the
1391 io patterns it contains. This can be used to store a
1392 workload and replay it sometime later. The iolog given
1393 may also be a blktrace binary file, which allows fio
1394 to replay a workload captured by blktrace. See blktrace
1395 for how to capture such logging data. For blktrace replay,
1396 the file needs to be turned into a blkparse binary data
1397 file first (blkparse <device> -o /dev/null -d file_for_fio.bin).
1399 replay_no_stall=int When replaying I/O with read_iolog the default behavior
1400 is to attempt to respect the time stamps within the log and
1401 replay them with the appropriate delay between IOPS. By
1402 setting this variable fio will not respect the timestamps and
1403 attempt to replay them as fast as possible while still
1404 respecting ordering. The result is the same I/O pattern to a
1405 given device, but different timings.
1407 replay_redirect=str While replaying I/O patterns using read_iolog the
1408 default behavior is to replay the IOPS onto the major/minor
1409 device that each IOP was recorded from. This is sometimes
1410 undesirable because on a different machine those major/minor
1411 numbers can map to a different device. Changing hardware on
1412 the same system can also result in a different major/minor
1413 mapping. Replay_redirect causes all IOPS to be replayed onto
1414 the single specified device regardless of the device it was
1415 recorded from. i.e. replay_redirect=/dev/sdc would cause all
1416 IO in the blktrace to be replayed onto /dev/sdc. This means
1417 multiple devices will be replayed onto a single, if the trace
1418 contains multiple devices. If you want multiple devices to be
1419 replayed concurrently to multiple redirected devices you must
1420 blkparse your trace into separate traces and replay them with
1421 independent fio invocations. Unfortuantely this also breaks
1422 the strict time ordering between multiple device accesses.
1424 write_bw_log=str If given, write a bandwidth log of the jobs in this job
1425 file. Can be used to store data of the bandwidth of the
1426 jobs in their lifetime. The included fio_generate_plots
1427 script uses gnuplot to turn these text files into nice
1428 graphs. See write_lat_log for behaviour of given
1429 filename. For this option, the suffix is _bw.x.log, where
1430 x is the index of the job (1..N, where N is the number of
1433 write_lat_log=str Same as write_bw_log, except that this option stores io
1434 submission, completion, and total latencies instead. If no
1435 filename is given with this option, the default filename of
1436 "jobname_type.log" is used. Even if the filename is given,
1437 fio will still append the type of log. So if one specifies
1441 The actual log names will be foo_slat.x.log, foo_clat.x.log,
1442 and foo_lat.x.log, where x is the index of the job (1..N,
1443 where N is the number of jobs). This helps fio_generate_plot
1444 fine the logs automatically.
1446 write_iops_log=str Same as write_bw_log, but writes IOPS. If no filename is
1447 given with this option, the default filename of
1448 "jobname_type.x.log" is used,where x is the index of the job
1449 (1..N, where N is the number of jobs). Even if the filename
1450 is given, fio will still append the type of log.
1452 log_avg_msec=int By default, fio will log an entry in the iops, latency,
1453 or bw log for every IO that completes. When writing to the
1454 disk log, that can quickly grow to a very large size. Setting
1455 this option makes fio average the each log entry over the
1456 specified period of time, reducing the resolution of the log.
1459 log_offset=int If this is set, the iolog options will include the byte
1460 offset for the IO entry as well as the other data values.
1462 log_compression=int If this is set, fio will compress the IO logs as
1463 it goes, to keep the memory footprint lower. When a log
1464 reaches the specified size, that chunk is removed and
1465 compressed in the background. Given that IO logs are
1466 fairly highly compressible, this yields a nice memory
1467 savings for longer runs. The downside is that the
1468 compression will consume some background CPU cycles, so
1469 it may impact the run. This, however, is also true if
1470 the logging ends up consuming most of the system memory.
1471 So pick your poison. The IO logs are saved normally at the
1472 end of a run, by decompressing the chunks and storing them
1473 in the specified log file. This feature depends on the
1474 availability of zlib.
1476 log_store_compressed=bool If set, and log_compression is also set,
1477 fio will store the log files in a compressed format. They
1478 can be decompressed with fio, using the --inflate-log
1479 command line parameter. The files will be stored with a
1482 block_error_percentiles=bool If set, record errors in trim block-sized
1483 units from writes and trims and output a histogram of
1484 how many trims it took to get to errors, and what kind
1485 of error was encountered.
1487 lockmem=int Pin down the specified amount of memory with mlock(2). Can
1488 potentially be used instead of removing memory or booting
1489 with less memory to simulate a smaller amount of memory.
1490 The amount specified is per worker.
1492 exec_prerun=str Before running this job, issue the command specified
1493 through system(3). Output is redirected in a file called
1496 exec_postrun=str After the job completes, issue the command specified
1497 though system(3). Output is redirected in a file called
1498 jobname.postrun.txt.
1500 ioscheduler=str Attempt to switch the device hosting the file to the specified
1501 io scheduler before running.
1503 disk_util=bool Generate disk utilization statistics, if the platform
1504 supports it. Defaults to on.
1506 disable_lat=bool Disable measurements of total latency numbers. Useful
1507 only for cutting back the number of calls to gettimeofday,
1508 as that does impact performance at really high IOPS rates.
1509 Note that to really get rid of a large amount of these
1510 calls, this option must be used with disable_slat and
1513 disable_clat=bool Disable measurements of completion latency numbers. See
1516 disable_slat=bool Disable measurements of submission latency numbers. See
1519 disable_bw=bool Disable measurements of throughput/bandwidth numbers. See
1522 clat_percentiles=bool Enable the reporting of percentiles of
1523 completion latencies.
1525 percentile_list=float_list Overwrite the default list of percentiles
1526 for completion latencies and the block error histogram.
1527 Each number is a floating number in the range (0,100],
1528 and the maximum length of the list is 20. Use ':'
1529 to separate the numbers, and list the numbers in ascending
1530 order. For example, --percentile_list=99.5:99.9 will cause
1531 fio to report the values of completion latency below which
1532 99.5% and 99.9% of the observed latencies fell, respectively.
1534 clocksource=str Use the given clocksource as the base of timing. The
1535 supported options are:
1537 gettimeofday gettimeofday(2)
1539 clock_gettime clock_gettime(2)
1541 cpu Internal CPU clock source
1543 cpu is the preferred clocksource if it is reliable, as it
1544 is very fast (and fio is heavy on time calls). Fio will
1545 automatically use this clocksource if it's supported and
1546 considered reliable on the system it is running on, unless
1547 another clocksource is specifically set. For x86/x86-64 CPUs,
1548 this means supporting TSC Invariant.
1550 gtod_reduce=bool Enable all of the gettimeofday() reducing options
1551 (disable_clat, disable_slat, disable_bw) plus reduce
1552 precision of the timeout somewhat to really shrink
1553 the gettimeofday() call count. With this option enabled,
1554 we only do about 0.4% of the gtod() calls we would have
1555 done if all time keeping was enabled.
1557 gtod_cpu=int Sometimes it's cheaper to dedicate a single thread of
1558 execution to just getting the current time. Fio (and
1559 databases, for instance) are very intensive on gettimeofday()
1560 calls. With this option, you can set one CPU aside for
1561 doing nothing but logging current time to a shared memory
1562 location. Then the other threads/processes that run IO
1563 workloads need only copy that segment, instead of entering
1564 the kernel with a gettimeofday() call. The CPU set aside
1565 for doing these time calls will be excluded from other
1566 uses. Fio will manually clear it from the CPU mask of other
1569 continue_on_error=str Normally fio will exit the job on the first observed
1570 failure. If this option is set, fio will continue the job when
1571 there is a 'non-fatal error' (EIO or EILSEQ) until the runtime
1572 is exceeded or the I/O size specified is completed. If this
1573 option is used, there are two more stats that are appended,
1574 the total error count and the first error. The error field
1575 given in the stats is the first error that was hit during the
1578 The allowed values are:
1580 none Exit on any IO or verify errors.
1582 read Continue on read errors, exit on all others.
1584 write Continue on write errors, exit on all others.
1586 io Continue on any IO error, exit on all others.
1588 verify Continue on verify errors, exit on all others.
1590 all Continue on all errors.
1592 0 Backward-compatible alias for 'none'.
1594 1 Backward-compatible alias for 'all'.
1596 ignore_error=str Sometimes you want to ignore some errors during test
1597 in that case you can specify error list for each error type.
1598 ignore_error=READ_ERR_LIST,WRITE_ERR_LIST,VERIFY_ERR_LIST
1599 errors for given error type is separated with ':'. Error
1600 may be symbol ('ENOSPC', 'ENOMEM') or integer.
1602 ignore_error=EAGAIN,ENOSPC:122
1603 This option will ignore EAGAIN from READ, and ENOSPC and
1604 122(EDQUOT) from WRITE.
1606 error_dump=bool If set dump every error even if it is non fatal, true
1607 by default. If disabled only fatal error will be dumped
1609 cgroup=str Add job to this control group. If it doesn't exist, it will
1610 be created. The system must have a mounted cgroup blkio
1611 mount point for this to work. If your system doesn't have it
1612 mounted, you can do so with:
1614 # mount -t cgroup -o blkio none /cgroup
1616 cgroup_weight=int Set the weight of the cgroup to this value. See
1617 the documentation that comes with the kernel, allowed values
1618 are in the range of 100..1000.
1620 cgroup_nodelete=bool Normally fio will delete the cgroups it has created after
1621 the job completion. To override this behavior and to leave
1622 cgroups around after the job completion, set cgroup_nodelete=1.
1623 This can be useful if one wants to inspect various cgroup
1624 files after job completion. Default: false
1626 uid=int Instead of running as the invoking user, set the user ID to
1627 this value before the thread/process does any work.
1629 gid=int Set group ID, see uid.
1631 flow_id=int The ID of the flow. If not specified, it defaults to being a
1632 global flow. See flow.
1634 flow=int Weight in token-based flow control. If this value is used, then
1635 there is a 'flow counter' which is used to regulate the
1636 proportion of activity between two or more jobs. fio attempts
1637 to keep this flow counter near zero. The 'flow' parameter
1638 stands for how much should be added or subtracted to the flow
1639 counter on each iteration of the main I/O loop. That is, if
1640 one job has flow=8 and another job has flow=-1, then there
1641 will be a roughly 1:8 ratio in how much one runs vs the other.
1643 flow_watermark=int The maximum value that the absolute value of the flow
1644 counter is allowed to reach before the job must wait for a
1645 lower value of the counter.
1647 flow_sleep=int The period of time, in microseconds, to wait after the flow
1648 watermark has been exceeded before retrying operations
1650 In addition, there are some parameters which are only valid when a specific
1651 ioengine is in use. These are used identically to normal parameters, with the
1652 caveat that when used on the command line, they must come after the ioengine
1653 that defines them is selected.
1655 [libaio] userspace_reap Normally, with the libaio engine in use, fio will use
1656 the io_getevents system call to reap newly returned events.
1657 With this flag turned on, the AIO ring will be read directly
1658 from user-space to reap events. The reaping mode is only
1659 enabled when polling for a minimum of 0 events (eg when
1660 iodepth_batch_complete=0).
1662 [cpu] cpuload=int Attempt to use the specified percentage of CPU cycles.
1664 [cpu] cpuchunks=int Split the load into cycles of the given time. In
1667 [cpu] exit_on_io_done=bool Detect when IO threads are done, then exit.
1669 [netsplice] hostname=str
1670 [net] hostname=str The host name or IP address to use for TCP or UDP based IO.
1671 If the job is a TCP listener or UDP reader, the hostname is not
1672 used and must be omitted unless it is a valid UDP multicast
1675 [netsplice] port=int
1676 [net] port=int The TCP or UDP port to bind to or connect to. If this is used
1677 with numjobs to spawn multiple instances of the same job type, then this will
1678 be the starting port number since fio will use a range of ports.
1680 [netsplice] interface=str
1681 [net] interface=str The IP address of the network interface used to send or
1682 receive UDP multicast
1685 [net] ttl=int Time-to-live value for outgoing UDP multicast packets.
1688 [netsplice] nodelay=bool
1689 [net] nodelay=bool Set TCP_NODELAY on TCP connections.
1691 [netsplice] protocol=str
1692 [netsplice] proto=str
1694 [net] proto=str The network protocol to use. Accepted values are:
1696 tcp Transmission control protocol
1697 tcpv6 Transmission control protocol V6
1698 udp User datagram protocol
1699 udpv6 User datagram protocol V6
1700 unix UNIX domain socket
1702 When the protocol is TCP or UDP, the port must also be given,
1703 as well as the hostname if the job is a TCP listener or UDP
1704 reader. For unix sockets, the normal filename option should be
1705 used and the port is invalid.
1707 [net] listen For TCP network connections, tell fio to listen for incoming
1708 connections rather than initiating an outgoing connection. The
1709 hostname must be omitted if this option is used.
1711 [net] pingpong Normaly a network writer will just continue writing data, and
1712 a network reader will just consume packages. If pingpong=1
1713 is set, a writer will send its normal payload to the reader,
1714 then wait for the reader to send the same payload back. This
1715 allows fio to measure network latencies. The submission
1716 and completion latencies then measure local time spent
1717 sending or receiving, and the completion latency measures
1718 how long it took for the other end to receive and send back.
1719 For UDP multicast traffic pingpong=1 should only be set for a
1720 single reader when multiple readers are listening to the same
1723 [net] window_size Set the desired socket buffer size for the connection.
1725 [net] mss Set the TCP maximum segment size (TCP_MAXSEG).
1727 [e4defrag] donorname=str
1728 File will be used as a block donor(swap extents between files)
1729 [e4defrag] inplace=int
1730 Configure donor file blocks allocation strategy
1731 0(default): Preallocate donor's file on init
1732 1 : allocate space immidietly inside defragment event,
1733 and free right after event
1735 [mtd] skip_bad=bool Skip operations against known bad blocks.
1738 6.0 Interpreting the output
1739 ---------------------------
1741 fio spits out a lot of output. While running, fio will display the
1742 status of the jobs created. An example of that would be:
1744 Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
1746 The characters inside the square brackets denote the current status of
1747 each thread. The possible values (in typical life cycle order) are:
1751 P Thread setup, but not started.
1753 I Thread initialized, waiting or generating necessary data.
1754 p Thread running pre-reading file(s).
1755 R Running, doing sequential reads.
1756 r Running, doing random reads.
1757 W Running, doing sequential writes.
1758 w Running, doing random writes.
1759 M Running, doing mixed sequential reads/writes.
1760 m Running, doing mixed random reads/writes.
1761 F Running, currently waiting for fsync()
1762 f Running, finishing up (writing IO logs, etc)
1763 V Running, doing verification of written data.
1764 E Thread exited, not reaped by main thread yet.
1766 X Thread reaped, exited with an error.
1767 K Thread reaped, exited due to signal.
1769 Fio will condense the thread string as not to take up more space on the
1770 command line as is needed. For instance, if you have 10 readers and 10
1771 writers running, the output would look like this:
1773 Jobs: 20 (f=20): [R(10),W(10)] [4.0% done] [2103MB/0KB/0KB /s] [538K/0/0 iops] [eta 57m:36s]
1775 Fio will still maintain the ordering, though. So the above means that jobs
1776 1..10 are readers, and 11..20 are writers.
1778 The other values are fairly self explanatory - number of threads
1779 currently running and doing io, rate of io since last check (read speed
1780 listed first, then write speed), and the estimated completion percentage
1781 and time for the running group. It's impossible to estimate runtime of
1782 the following groups (if any). Note that the string is displayed in order,
1783 so it's possible to tell which of the jobs are currently doing what. The
1784 first character is the first job defined in the job file, and so forth.
1786 When fio is done (or interrupted by ctrl-c), it will show the data for
1787 each thread, group of threads, and disks in that order. For each data
1788 direction, the output looks like:
1790 Client1 (g=0): err= 0:
1791 write: io= 32MB, bw= 666KB/s, iops=89 , runt= 50320msec
1792 slat (msec): min= 0, max= 136, avg= 0.03, stdev= 1.92
1793 clat (msec): min= 0, max= 631, avg=48.50, stdev=86.82
1794 bw (KB/s) : min= 0, max= 1196, per=51.00%, avg=664.02, stdev=681.68
1795 cpu : usr=1.49%, sys=0.25%, ctx=7969, majf=0, minf=17
1796 IO depths : 1=0.1%, 2=0.3%, 4=0.5%, 8=99.0%, 16=0.0%, 32=0.0%, >32=0.0%
1797 submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
1798 complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
1799 issued r/w: total=0/32768, short=0/0
1800 lat (msec): 2=1.6%, 4=0.0%, 10=3.2%, 20=12.8%, 50=38.4%, 100=24.8%,
1801 lat (msec): 250=15.2%, 500=0.0%, 750=0.0%, 1000=0.0%, >=2048=0.0%
1803 The client number is printed, along with the group id and error of that
1804 thread. Below is the io statistics, here for writes. In the order listed,
1807 io= Number of megabytes io performed
1808 bw= Average bandwidth rate
1809 iops= Average IOs performed per second
1810 runt= The runtime of that thread
1811 slat= Submission latency (avg being the average, stdev being the
1812 standard deviation). This is the time it took to submit
1813 the io. For sync io, the slat is really the completion
1814 latency, since queue/complete is one operation there. This
1815 value can be in milliseconds or microseconds, fio will choose
1816 the most appropriate base and print that. In the example
1817 above, milliseconds is the best scale. Note: in --minimal mode
1818 latencies are always expressed in microseconds.
1819 clat= Completion latency. Same names as slat, this denotes the
1820 time from submission to completion of the io pieces. For
1821 sync io, clat will usually be equal (or very close) to 0,
1822 as the time from submit to complete is basically just
1823 CPU time (io has already been done, see slat explanation).
1824 bw= Bandwidth. Same names as the xlat stats, but also includes
1825 an approximate percentage of total aggregate bandwidth
1826 this thread received in this group. This last value is
1827 only really useful if the threads in this group are on the
1828 same disk, since they are then competing for disk access.
1829 cpu= CPU usage. User and system time, along with the number
1830 of context switches this thread went through, usage of
1831 system and user time, and finally the number of major
1832 and minor page faults.
1833 IO depths= The distribution of io depths over the job life time. The
1834 numbers are divided into powers of 2, so for example the
1835 16= entries includes depths up to that value but higher
1836 than the previous entry. In other words, it covers the
1837 range from 16 to 31.
1838 IO submit= How many pieces of IO were submitting in a single submit
1839 call. Each entry denotes that amount and below, until
1840 the previous entry - eg, 8=100% mean that we submitted
1841 anywhere in between 5-8 ios per submit call.
1842 IO complete= Like the above submit number, but for completions instead.
1843 IO issued= The number of read/write requests issued, and how many
1845 IO latencies= The distribution of IO completion latencies. This is the
1846 time from when IO leaves fio and when it gets completed.
1847 The numbers follow the same pattern as the IO depths,
1848 meaning that 2=1.6% means that 1.6% of the IO completed
1849 within 2 msecs, 20=12.8% means that 12.8% of the IO
1850 took more than 10 msecs, but less than (or equal to) 20 msecs.
1852 After each client has been listed, the group statistics are printed. They
1853 will look like this:
1855 Run status group 0 (all jobs):
1856 READ: io=64MB, aggrb=22178, minb=11355, maxb=11814, mint=2840msec, maxt=2955msec
1857 WRITE: io=64MB, aggrb=1302, minb=666, maxb=669, mint=50093msec, maxt=50320msec
1859 For each data direction, it prints:
1861 io= Number of megabytes io performed.
1862 aggrb= Aggregate bandwidth of threads in this group.
1863 minb= The minimum average bandwidth a thread saw.
1864 maxb= The maximum average bandwidth a thread saw.
1865 mint= The smallest runtime of the threads in that group.
1866 maxt= The longest runtime of the threads in that group.
1868 And finally, the disk statistics are printed. They will look like this:
1870 Disk stats (read/write):
1871 sda: ios=16398/16511, merge=30/162, ticks=6853/819634, in_queue=826487, util=100.00%
1873 Each value is printed for both reads and writes, with reads first. The
1876 ios= Number of ios performed by all groups.
1877 merge= Number of merges io the io scheduler.
1878 ticks= Number of ticks we kept the disk busy.
1879 io_queue= Total time spent in the disk queue.
1880 util= The disk utilization. A value of 100% means we kept the disk
1881 busy constantly, 50% would be a disk idling half of the time.
1883 It is also possible to get fio to dump the current output while it is
1884 running, without terminating the job. To do that, send fio the USR1 signal.
1885 You can also get regularly timed dumps by using the --status-interval
1886 parameter, or by creating a file in /tmp named fio-dump-status. If fio
1887 sees this file, it will unlink it and dump the current output status.
1893 For scripted usage where you typically want to generate tables or graphs
1894 of the results, fio can output the results in a semicolon separated format.
1895 The format is one long line of values, such as:
1897 2;card0;0;0;7139336;121836;60004;1;10109;27.932460;116.933948;220;126861;3495.446807;1085.368601;226;126864;3523.635629;1089.012448;24063;99944;50.275485%;59818.274627;5540.657370;7155060;122104;60004;1;8338;29.086342;117.839068;388;128077;5032.488518;1234.785715;391;128085;5061.839412;1236.909129;23436;100928;50.287926%;59964.832030;5644.844189;14.595833%;19.394167%;123706;0;7313;0.1%;0.1%;0.1%;0.1%;0.1%;0.1%;100.0%;0.00%;0.00%;0.00%;0.00%;0.00%;0.00%;0.01%;0.02%;0.05%;0.16%;6.04%;40.40%;52.68%;0.64%;0.01%;0.00%;0.01%;0.00%;0.00%;0.00%;0.00%;0.00%
1898 A description of this job goes here.
1900 The job description (if provided) follows on a second line.
1902 To enable terse output, use the --minimal command line option. The first
1903 value is the version of the terse output format. If the output has to
1904 be changed for some reason, this number will be incremented by 1 to
1905 signify that change.
1907 Split up, the format is as follows:
1909 terse version, fio version, jobname, groupid, error
1911 Total IO (KB), bandwidth (KB/sec), IOPS, runtime (msec)
1912 Submission latency: min, max, mean, deviation (usec)
1913 Completion latency: min, max, mean, deviation (usec)
1914 Completion latency percentiles: 20 fields (see below)
1915 Total latency: min, max, mean, deviation (usec)
1916 Bw (KB/s): min, max, aggregate percentage of total, mean, deviation
1918 Total IO (KB), bandwidth (KB/sec), IOPS, runtime (msec)
1919 Submission latency: min, max, mean, deviation (usec)
1920 Completion latency: min, max, mean, deviation (usec)
1921 Completion latency percentiles: 20 fields (see below)
1922 Total latency: min, max, mean, deviation (usec)
1923 Bw (KB/s): min, max, aggregate percentage of total, mean, deviation
1924 CPU usage: user, system, context switches, major faults, minor faults
1925 IO depths: <=1, 2, 4, 8, 16, 32, >=64
1926 IO latencies microseconds: <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000
1927 IO latencies milliseconds: <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000
1928 Disk utilization: Disk name, Read ios, write ios,
1929 Read merges, write merges,
1930 Read ticks, write ticks,
1931 Time spent in queue, disk utilization percentage
1932 Additional Info (dependent on continue_on_error, default off): total # errors, first error code
1934 Additional Info (dependent on description being set): Text description
1936 Completion latency percentiles can be a grouping of up to 20 sets, so
1937 for the terse output fio writes all of them. Each field will look like this:
1941 which is the Xth percentile, and the usec latency associated with it.
1943 For disk utilization, all disks used by fio are shown. So for each disk
1944 there will be a disk utilization section.
1947 8.0 Trace file format
1948 ---------------------
1949 There are two trace file format that you can encounter. The older (v1) format
1950 is unsupported since version 1.20-rc3 (March 2008). It will still be described
1951 below in case that you get an old trace and want to understand it.
1953 In any case the trace is a simple text file with a single action per line.
1956 8.1 Trace file format v1
1957 ------------------------
1958 Each line represents a single io action in the following format:
1962 where rw=0/1 for read/write, and the offset and length entries being in bytes.
1964 This format is not supported in Fio versions => 1.20-rc3.
1967 8.2 Trace file format v2
1968 ------------------------
1969 The second version of the trace file format was added in Fio version 1.17.
1970 It allows to access more then one file per trace and has a bigger set of
1971 possible file actions.
1973 The first line of the trace file has to be:
1977 Following this can be lines in two different formats, which are described below.
1979 The file management format:
1983 The filename is given as an absolute path. The action can be one of these:
1985 add Add the given filename to the trace
1986 open Open the file with the given filename. The filename has to have
1987 been added with the add action before.
1988 close Close the file with the given filename. The file has to have been
1992 The file io action format:
1994 filename action offset length
1996 The filename is given as an absolute path, and has to have been added and opened
1997 before it can be used with this format. The offset and length are given in
1998 bytes. The action can be one of these:
2000 wait Wait for 'offset' microseconds. Everything below 100 is discarded.
2001 read Read 'length' bytes beginning from 'offset'
2002 write Write 'length' bytes beginning from 'offset'
2003 sync fsync() the file
2004 datasync fdatasync() the file
2005 trim trim the given file from the given 'offset' for 'length' bytes
2008 9.0 CPU idleness profiling
2009 --------------------------
2010 In some cases, we want to understand CPU overhead in a test. For example,
2011 we test patches for the specific goodness of whether they reduce CPU usage.
2012 fio implements a balloon approach to create a thread per CPU that runs at
2013 idle priority, meaning that it only runs when nobody else needs the cpu.
2014 By measuring the amount of work completed by the thread, idleness of each
2015 CPU can be derived accordingly.
2017 An unit work is defined as touching a full page of unsigned characters. Mean
2018 and standard deviation of time to complete an unit work is reported in "unit
2019 work" section. Options can be chosen to report detailed percpu idleness or
2020 overall system idleness by aggregating percpu stats.
2023 10.0 Verification and triggers
2024 ------------------------------
2025 Fio is usually run in one of two ways, when data verification is done. The
2026 first is a normal write job of some sort with verify enabled. When the
2027 write phase has completed, fio switches to reads and verifies everything
2028 it wrote. The second model is running just the write phase, and then later
2029 on running the same job (but with reads instead of writes) to repeat the
2030 same IO patterns and verify the contents. Both of these methods depend
2031 on the write phase being completed, as fio otherwise has no idea how much
2034 With verification triggers, fio supports dumping the current write state
2035 to local files. Then a subsequent read verify workload can load this state
2036 and know exactly where to stop. This is useful for testing cases where
2037 power is cut to a server in a managed fashion, for instance.
2039 A verification trigger consists of two things:
2041 1) Storing the write state of each job
2042 2) Executing a trigger command
2044 The write state is relatively small, on the order of hundreds of bytes
2045 to single kilobytes. It contains information on the number of completions
2046 done, the last X completions, etc.
2048 A trigger is invoked either through creation ('touch') of a specified
2049 file in the system, or through a timeout setting. If fio is run with
2050 --trigger-file=/tmp/trigger-file, then it will continually check for
2051 the existence of /tmp/trigger-file. When it sees this file, it will
2052 fire off the trigger (thus saving state, and executing the trigger
2055 For client/server runs, there's both a local and remote trigger. If
2056 fio is running as a server backend, it will send the job states back
2057 to the client for safe storage, then execute the remote trigger, if
2058 specified. If a local trigger is specified, the server will still send
2059 back the write state, but the client will then execute the trigger.
2061 10.1 Verification trigger example
2062 ---------------------------------
2063 Lets say we want to run a powercut test on the remote machine 'server'.
2064 Our write workload is in write-test.fio. We want to cut power to 'server'
2065 at some point during the run, and we'll run this test from the safety
2066 or our local machine, 'localbox'. On the server, we'll start the fio
2069 server# fio --server
2071 and on the client, we'll fire off the workload:
2073 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger-remote="bash -c \"echo b > /proc/sysrq-triger\""
2075 We set /tmp/my-trigger as the trigger file, and we tell fio to execute
2077 echo b > /proc/sysrq-trigger
2079 on the server once it has received the trigger and sent us the write
2080 state. This will work, but it's not _really_ cutting power to the server,
2081 it's merely abruptly rebooting it. If we have a remote way of cutting
2082 power to the server through IPMI or similar, we could do that through
2083 a local trigger command instead. Lets assume we have a script that does
2084 IPMI reboot of a given hostname, ipmi-reboot. On localbox, we could
2085 then have run fio with a local trigger instead:
2087 localbox$ fio --client=server --trigger-file=/tmp/my-trigger --trigger="ipmi-reboot server"
2089 For this case, fio would wait for the server to send us the write state,
2090 then execute 'ipmi-reboot server' when that happened.
2092 10.1 Loading verify state
2093 -------------------------
2094 To load store write state, read verification job file must contain
2095 the verify_state_load option. If that is set, fio will load the previously
2096 stored state. For a local fio run this is done by loading the files directly,
2097 and on a client/server run, the server backend will ask the client to send
2098 the files over and load them from there.