1 .TH fio 1 "September 2007" "User Manual"
3 fio \- flexible I/O tester
6 [\fIoptions\fR] [\fIjobfile\fR]...
9 is a tool that will spawn a number of threads or processes doing a
10 particular type of I/O action as specified by the user.
11 The typical use of fio is to write a job file matching the I/O load
12 one wants to simulate.
15 .BI \-\-output \fR=\fPfilename
16 Write output to \fIfilename\fR.
18 .BI \-\-timeout \fR=\fPtimeout
19 Limit run time to \fItimeout\fR seconds.
22 Generate per-job latency logs.
25 Generate per-job bandwidth logs.
28 Print statistics in a terse, semicolon-delimited format.
30 .BI \-\-showcmd \fR=\fPjobfile
31 Convert \fIjobfile\fR to a set of command-line options.
34 Enable read-only safety checks.
36 .BI \-\-eta \fR=\fPwhen
37 Specifies when real-time ETA estimate should be printed. \fIwhen\fR may
38 be one of `always', `never' or `auto'.
40 .BI \-\-section \fR=\fPsec
41 Only run section \fIsec\fR from job file.
43 .BI \-\-cmdhelp \fR=\fPcommand
44 Print help information for \fIcommand\fR. May be `all' for all commands.
46 .BI \-\-debug \fR=\fPtype
47 Enable verbose tracing of various fio actions. May be `all' for all types
48 or individual types seperated by a comma (eg --debug=io,file). `help' will
49 list all available tracing options.
52 Display usage information and exit.
55 Display version information and exit.
57 Job files are in `ini' format. They consist of one or more
58 job definitions, which begin with a job name in square brackets and
59 extend to the next job name. The job name can be any ASCII string
60 except `global', which has a special meaning. Following the job name is
61 a sequence of zero or more parameters, one per line, that define the
62 behavior of the job. Any line starting with a `;' or `#' character is
63 considered a comment and ignored.
65 If \fIjobfile\fR is specified as `-', the job file will be read from
68 The global section contains default parameters for jobs specified in the
69 job file. A job is only affected by global sections residing above it,
70 and there may be any number of global sections. Specific job definitions
71 may override any parameter set in global sections.
74 Some parameters may take arguments of a specific type. The types used are:
77 String: a sequence of alphanumeric characters.
80 SI integer: a whole number, possibly containing a suffix denoting the base unit
81 of the value. Accepted suffixes are `k', 'M' and 'G', denoting kilo (1024),
82 mega (1024*1024) and giga (1024*1024*1024) respectively. If prefixed with '0x',
83 the value is assumed to be base 16 (hexadecimal).
86 Boolean: a true or false value. `0' denotes false, `1' denotes true.
89 Integer range: a range of integers specified in the format
90 \fIlower\fR:\fIupper\fR or \fIlower\fR\-\fIupper\fR. \fIlower\fR and
91 \fIupper\fR may contain a suffix as described above. If an option allows two
92 sets of ranges, they are separated with a `,' or `/' character. For example:
97 May be used to override the job name. On the command line, this parameter
98 has the special purpose of signalling the start of a new job.
100 .BI description \fR=\fPstr
101 Human-readable description of the job. It is printed when the job is run, but
102 otherwise has no special purpose.
104 .BI directory \fR=\fPstr
105 Prefix filenames with this directory. Used to place files in a location other
108 .BI filename \fR=\fPstr
110 normally makes up a file name based on the job name, thread number, and file
111 number. If you want to share files between threads in a job or several jobs,
112 specify a \fIfilename\fR for each of them to override the default. If the I/O
113 engine used is `net', \fIfilename\fR is the host and port to connect to in the
114 format \fIhost\fR/\fIport\fR. If the I/O engine is file-based, you can specify
115 a number of files by separating the names with a `:' character. `\-' is a
116 reserved name, meaning stdin or stdout, depending on the read/write direction
119 .BI lockfile \fR=\fPstr
120 Fio defaults to not locking any files before it does IO to them. If a file or
121 file descriptor is shared, fio can serialize IO to that file to make the end
122 result consistent. This is usual for emulating real workloads that share files.
128 No locking. This is the default.
131 Only one thread or process may do IO at the time, excluding all others.
134 Read-write locking on the file. Many readers may access the file at the same
135 time, but writes get exclusive access.
138 The option may be post-fixed with a lock batch number. If set, then each
139 thread/process may do that amount of IOs to the file before giving up the lock.
140 Since lock acquisition is expensive, batching the lock/unlocks will speed up IO.
143 .BI opendir \fR=\fPstr
144 Recursively open any files below directory \fIstr\fR.
146 .BI readwrite \fR=\fPstr "\fR,\fP rw" \fR=\fPstr
147 Type of I/O pattern. Accepted values are:
164 Mixed sequential reads and writes.
167 Mixed random reads and writes.
170 For mixed I/O, the default split is 50/50. For random I/O, the number of I/Os
171 to perform before getting a new offset can be specified by appending
172 `:\fIint\fR' to the pattern type. The default is 1.
175 .BI randrepeat \fR=\fPbool
176 Seed the random number generator in a predictable way so results are repeatable
177 across runs. Default: true.
179 .BI fadvise_hint \fR=\fPbool
180 Disable use of \fIposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
181 are likely to be issued. Default: true.
184 Total size of I/O for this job. \fBfio\fR will run until this many bytes have
185 been transfered, unless limited by other options (\fBruntime\fR, for instance).
186 Unless \fBnr_files\fR and \fBfilesize\fR options are given, this amount will be
187 divided between the available files for the job.
189 .BI fill_device \fR=\fPbool
190 Sets size to something really large and waits for ENOSPC (no space left on
191 device) as the terminating condition. Only makes sense with sequential write.
192 For a read workload, the mount point will be filled first then IO started on
195 .BI filesize \fR=\fPirange
196 Individual file sizes. May be a range, in which case \fBfio\fR will select sizes
197 for files at random within the given range, limited to \fBsize\fR in total (if
198 that is given). If \fBfilesize\fR is not specified, each created file is the
201 .BI blocksize \fR=\fPint[,int] "\fR,\fB bs" \fR=\fPint[,int]
202 Block size for I/O units. Default: 4k. Values for reads and writes can be
203 specified seperately in the format \fIread\fR,\fIwrite\fR, either of
204 which may be empty to leave that value at its default.
206 .BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange]
207 Specify a range of I/O block sizes. The issued I/O unit will always be a
208 multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies
209 to both reads and writes if only one range is given, but can be specified
210 seperately with a comma seperating the values. Example: bsrange=1k-4k,2k-8k.
211 Also (see \fBblocksize\fR).
213 .BI bssplit \fR=\fPstr
214 This option allows even finer grained control of the block sizes issued,
215 not just even splits between them. With this option, you can weight various
216 block sizes for exact control of the issued IO for a job that has mixed
217 block sizes. The format of the option is bssplit=blocksize/percentage,
218 optionally adding as many definitions as needed seperated by a colon.
219 Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k
220 blocks and 40% 32k blocks. \fBbssplit\fR also supports giving separate
221 splits to reads and writes. The format is identical to what the
222 \fBbs\fR option accepts, the read and write parts are separated with a
225 .B blocksize_unaligned\fR,\fP bs_unaligned
226 If set, any size in \fBblocksize_range\fR may be used. This typically won't
227 work with direct I/O, as that normally requires sector alignment.
229 .BI blockalign \fR=\fPint[,int] "\fR,\fB ba" \fR=\fPint[,int]
230 At what boundary to align random IO offsets. Defaults to the same as 'blocksize'
231 the minimum blocksize given. Minimum alignment is typically 512b
232 for using direct IO, though it usually depends on the hardware block size.
233 This option is mutually exclusive with using a random map for files, so it
234 will turn off that option.
237 Initialise buffers with all zeros. Default: fill buffers with random data.
240 If this option is given, fio will refill the IO buffers on every submit. The
241 default is to only fill it at init time and reuse that data. Only makes sense
242 if zero_buffers isn't specified, naturally. If data verification is enabled,
243 refill_buffers is also automatically enabled.
245 .BI nrfiles \fR=\fPint
246 Number of files to use for this job. Default: 1.
248 .BI openfiles \fR=\fPint
249 Number of files to keep open at the same time. Default: \fBnrfiles\fR.
251 .BI file_service_type \fR=\fPstr
252 Defines how files to service are selected. The following types are defined:
257 Choose a file at random
260 Round robin over open files (default).
262 Do each file in the set sequentially.
265 The number of I/Os to issue before switching a new file can be specified by
266 appending `:\fIint\fR' to the service type.
269 .BI ioengine \fR=\fPstr
270 Defines how the job issues I/O. The following types are defined:
275 Basic \fIread\fR\|(2) or \fIwrite\fR\|(2) I/O. \fIfseek\fR\|(2) is used to
276 position the I/O location.
279 Basic \fIpread\fR\|(2) or \fIpwrite\fR\|(2) I/O.
282 Basic \fIreadv\fR\|(2) or \fIwritev\fR\|(2) I/O. Will emulate queuing by
283 coalescing adjacents IOs into a single submission.
286 Linux native asynchronous I/O.
289 glibc POSIX asynchronous I/O using \fIaio_read\fR\|(3) and \fIaio_write\fR\|(3).
292 File is memory mapped with \fImmap\fR\|(2) and data copied using
296 \fIsplice\fR\|(2) is used to transfer the data and \fIvmsplice\fR\|(2) to
297 transfer data from user-space to the kernel.
300 Use the syslet system calls to make regular read/write asynchronous.
303 SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if
304 the target is an sg character device, we use \fIread\fR\|(2) and
305 \fIwrite\fR\|(2) for asynchronous I/O.
308 Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR
309 itself and for debugging and testing purposes.
312 Transfer over the network. \fBfilename\fR must be set appropriately to
313 `\fIhost\fR/\fIport\fR' regardless of data direction. If receiving, only the
314 \fIport\fR argument is used.
317 Like \fBnet\fR, but uses \fIsplice\fR\|(2) and \fIvmsplice\fR\|(2) to map data
321 Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and
322 \fBcpucycles\fR parameters.
325 The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface
326 approach to asycnronous I/O.
328 See <http://www.xmailserver.org/guasi\-lib.html>.
331 Loads an external I/O engine object file. Append the engine filename as
336 .BI iodepth \fR=\fPint
337 Number of I/O units to keep in flight against the file. Default: 1.
339 .BI iodepth_batch \fR=\fPint
340 Number of I/Os to submit at once. Default: \fBiodepth\fR.
342 .BI iodepth_batch_complete \fR=\fPint
343 This defines how many pieces of IO to retrieve at once. It defaults to 1 which
344 means that we'll ask for a minimum of 1 IO in the retrieval process from the
345 kernel. The IO retrieval will go on until we hit the limit set by
346 \fBiodepth_low\fR. If this variable is set to 0, then fio will always check for
347 completed events before queuing more IO. This helps reduce IO latency, at the
348 cost of more retrieval system calls.
350 .BI iodepth_low \fR=\fPint
351 Low watermark indicating when to start filling the queue again. Default:
354 .BI direct \fR=\fPbool
355 If true, use non-buffered I/O (usually O_DIRECT). Default: false.
357 .BI buffered \fR=\fPbool
358 If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter.
361 .BI offset \fR=\fPint
362 Offset in the file to start I/O. Data before the offset will not be touched.
365 How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If
366 0, don't sync. Default: 0.
368 .BI fdatasync \fR=\fPint
369 Like \fBfsync\fR, but uses \fBfdatasync\fR\|(2) instead to only sync the
370 data parts of the file. Default: 0.
372 .BI overwrite \fR=\fPbool
373 If writing, setup the file first and do overwrites. Default: false.
375 .BI end_fsync \fR=\fPbool
376 Sync file contents when job exits. Default: false.
378 .BI fsync_on_close \fR=\fPbool
379 If true, sync file contents on close. This differs from \fBend_fsync\fR in that
380 it will happen on every close, not just at the end of the job. Default: false.
382 .BI rwmixcycle \fR=\fPint
383 How many milliseconds before switching between reads and writes for a mixed
384 workload. Default: 500ms.
386 .BI rwmixread \fR=\fPint
387 Percentage of a mixed workload that should be reads. Default: 50.
389 .BI rwmixwrite \fR=\fPint
390 Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and
391 \fBrwmixwrite\fR are given and do not sum to 100%, the latter of the two
392 overrides the first. This may interfere with a given rate setting, if fio is
393 asked to limit reads or writes to a certain rate. If that is the case, then
394 the distribution may be skewed. Default: 50.
397 Normally \fBfio\fR will cover every block of the file when doing random I/O. If
398 this parameter is given, a new offset will be chosen without looking at past
399 I/O history. This parameter is mutually exclusive with \fBverify\fR.
402 See \fBnorandommap\fR. If fio runs with the random block map enabled and it
403 fails to allocate the map, if this option is set it will continue without a
404 random block map. As coverage will not be as complete as with random maps, this
405 option is disabled by default.
408 Run job with given nice value. See \fInice\fR\|(2).
411 Set I/O priority value of this job between 0 (highest) and 7 (lowest). See
414 .BI prioclass \fR=\fPint
415 Set I/O priority class. See \fIionice\fR\|(1).
417 .BI thinktime \fR=\fPint
418 Stall job for given number of microseconds between issuing I/Os.
420 .BI thinktime_spin \fR=\fPint
421 Pretend to spend CPU time for given number of microseconds, sleeping the rest
422 of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set.
424 .BI thinktime_blocks \fR=\fPint
425 Number of blocks to issue before waiting \fBthinktime\fR microseconds.
429 Cap bandwidth used by this job. The number is in bytes/sec, the normal postfix
430 rules apply. You can use \fBrate\fR=500k to limit reads and writes to 500k each,
431 or you can specify read and writes separately. Using \fBrate\fR=1m,500k would
432 limit reads to 1MB/sec and writes to 500KB/sec. Capping only reads or writes
433 can be done with \fBrate\fR=,500k or \fBrate\fR=500k,. The former will only
434 limit writes (to 500KB/sec), the latter will only limit reads.
436 .BI ratemin \fR=\fPint
437 Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth.
438 Failing to meet this requirement will cause the job to exit. The same format
439 as \fBrate\fR is used for read vs write separation.
441 .BI rate_iops \fR=\fPint
442 Cap the bandwidth to this number of IOPS. Basically the same as rate, just
443 specified independently of bandwidth. The same format as \fBrate\fR is used for
444 read vs write seperation. If \fBblocksize\fR is a range, the smallest block
445 size is used as the metric.
447 .BI rate_iops_min \fR=\fPint
448 If this rate of I/O is not met, the job will exit. The same format as \fBrate\fR
449 is used for read vs write seperation.
451 .BI ratecycle \fR=\fPint
452 Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of
453 milliseconds. Default: 1000ms.
455 .BI cpumask \fR=\fPint
456 Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job
457 may run on. See \fBsched_setaffinity\fR\|(2).
459 .BI cpus_allowed \fR=\fPstr
460 Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers.
462 .BI startdelay \fR=\fPint
463 Delay start of job for the specified number of seconds.
465 .BI runtime \fR=\fPint
466 Terminate processing after the specified number of seconds.
469 If given, run for the specified \fBruntime\fR duration even if the files are
470 completely read or written. The same workload will be repeated as many times
471 as \fBruntime\fR allows.
473 .BI ramp_time \fR=\fPint
474 If set, fio will run the specified workload for this amount of time before
475 logging any performance numbers. Useful for letting performance settle before
476 logging results, thus minimizing the runtime required for stable results. Note
477 that the \fBramp_time\fR is considered lead in time for a job, thus it will
478 increase the total runtime if a special timeout or runtime is specified.
480 .BI invalidate \fR=\fPbool
481 Invalidate buffer-cache for the file prior to starting I/O. Default: true.
484 Use synchronous I/O for buffered writes. For the majority of I/O engines,
485 this means using O_SYNC. Default: false.
487 .BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr
488 Allocation method for I/O unit buffer. Allowed values are:
493 Allocate memory with \fImalloc\fR\|(3).
496 Use shared memory buffers allocated through \fIshmget\fR\|(2).
499 Same as \fBshm\fR, but use huge pages as backing.
502 Use \fImmap\fR\|(2) for allocation. Uses anonymous memory unless a filename
503 is given after the option in the format `:\fIfile\fR'.
506 Same as \fBmmap\fR, but use huge files as backing.
509 The amount of memory allocated is the maximum allowed \fBblocksize\fR for the
510 job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work,
511 the system must have free huge pages allocated. \fBmmaphuge\fR also needs to
512 have hugetlbfs mounted, and \fIfile\fR must point there.
515 .BI iomem_align \fR=\fPint
516 This indiciates the memory alignment of the IO memory buffers. Note that the
517 given alignment is applied to the first IO unit buffer, if using \fBiodepth\fR
518 the alignment of the following buffers are given by the \fBbs\fR used. In
519 other words, if using a \fBbs\fR that is a multiple of the page sized in the
520 system, all buffers will be aligned to this value. If using a \fBbs\fR that
521 is not page aligned, the alignment of subsequent IO memory buffers is the
522 sum of the \fBiomem_align\fR and \fBbs\fR used.
524 .BI hugepage\-size \fR=\fPint
525 Defines the size of a huge page. Must be at least equal to the system setting.
526 Should be a multiple of 1MiB. Default: 4MiB.
529 Terminate all jobs when one finishes. Default: wait for each job to finish.
531 .BI bwavgtime \fR=\fPint
532 Average bandwidth calculations over the given time in milliseconds. Default:
535 .BI create_serialize \fR=\fPbool
536 If true, serialize file creation for the jobs. Default: true.
538 .BI create_fsync \fR=\fPbool
539 \fIfsync\fR\|(2) data file after creation. Default: true.
541 .BI create_on_open \fR=\fPbool
542 If true, the files are not created until they are opened for IO by the job.
544 .BI pre_read \fR=\fPbool
545 If this is given, files will be pre-read into memory before starting the given
546 IO operation. This will also clear the \fR \fBinvalidate\fR flag, since it is
547 pointless to pre-read and then drop the cache. This will only work for IO
548 engines that are seekable, since they allow you to read the same data
549 multiple times. Thus it will not work on eg network or splice IO.
551 .BI unlink \fR=\fPbool
552 Unlink job files when done. Default: false.
555 Specifies the number of iterations (runs of the same workload) of this job.
558 .BI do_verify \fR=\fPbool
559 Run the verify phase after a write phase. Only valid if \fBverify\fR is set.
562 .BI verify \fR=\fPstr
563 Method of verifying file contents after each iteration of the job. Allowed
568 .B md5 crc16 crc32 crc64 crc7 sha256 sha512
569 Store appropriate checksum in the header of each block.
572 Write extra information about each I/O (timestamp, block number, etc.). The
573 block number is verified.
576 Fill I/O buffers with a specific pattern that is used to verify. The pattern is
577 specified by appending `:\fIint\fR' to the parameter. \fIint\fR cannot be larger
581 Pretend to verify. Used for testing internals.
585 .BI verify_sort \fR=\fPbool
586 If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to
587 read them back in a sorted manner. Default: true.
589 .BI verify_offset \fR=\fPint
590 Swap the verification header with data somewhere else in the block before
591 writing. It is swapped back before verifying.
593 .BI verify_interval \fR=\fPint
594 Write the verification header for this number of bytes, which should divide
595 \fBblocksize\fR. Default: \fBblocksize\fR.
597 .BI verify_fatal \fR=\fPbool
598 If true, exit the job on the first observed verification failure. Default:
601 .BI verify_async \fR=\fPint
602 Fio will normally verify IO inline from the submitting thread. This option
603 takes an integer describing how many async offload threads to create for IO
604 verification instead, causing fio to offload the duty of verifying IO contents
605 to one or more separate threads.
607 .BI verify_async_cpus \fR=\fPstr
608 Tell fio to set the given CPU affinity on the async IO verification threads.
609 See \fBcpus_allowed\fP for the format used.
612 Wait for preceeding jobs in the job file to exit before starting this one.
613 \fBstonewall\fR implies \fBnew_group\fR.
616 Start a new reporting group. If not given, all jobs in a file will be part
617 of the same reporting group, unless separated by a stonewall.
619 .BI numjobs \fR=\fPint
620 Number of clones (processes/threads performing the same workload) of this job.
624 If set, display per-group reports instead of per-job when \fBnumjobs\fR is
628 Use threads created with \fBpthread_create\fR\|(3) instead of processes created
629 with \fBfork\fR\|(2).
631 .BI zonesize \fR=\fPint
632 Divide file into zones of the specified size in bytes. See \fBzoneskip\fR.
634 .BI zoneskip \fR=\fPint
635 Skip the specified number of bytes when \fBzonesize\fR bytes of data have been
638 .BI write_iolog \fR=\fPstr
639 Write the issued I/O patterns to the specified file.
641 .BI read_iolog \fR=\fPstr
642 Replay the I/O patterns contained in the specified file generated by
643 \fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file.
645 .B write_bw_log \fR=\fPstr
646 If given, write a bandwidth log of the jobs in this job file. Can be used to
647 store data of the bandwidth of the jobs in their lifetime. The included
648 fio_generate_plots script uses gnuplot to turn these text files into nice
649 graphs. See \fBwrite_log_log\fR for behaviour of given filename. For this
650 option, the postfix is _bw.log.
653 Same as \fBwrite_bw_log\fR, but writes I/O completion latencies. If no
654 filename is given with this option, the default filename of "jobname_type.log"
655 is used. Even if the filename is given, fio will still append the type of log.
657 .B disable_clat \fR=\fPbool
658 Disable measurements of completion latency numbers. Useful only for cutting
659 back the number of calls to gettimeofday, as that does impact performance at
660 really high IOPS rates. Note that to really get rid of a large amount of these
661 calls, this option must be used with disable_slat and disable_bw as well.
663 .B disable_slat \fR=\fPbool
664 Disable measurements of submission latency numbers. See \fBdisable_clat\fR.
666 .B disable_bw_measurement \fR=\fPbool
667 Disable measurements of throughput/bandwidth numbers. See \fBdisable_clat\fR.
669 .BI lockmem \fR=\fPint
670 Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to
671 simulate a smaller amount of memory.
673 .BI exec_prerun \fR=\fPstr
674 Before running the job, execute the specified command with \fBsystem\fR\|(3).
676 .BI exec_postrun \fR=\fPstr
677 Same as \fBexec_prerun\fR, but the command is executed after the job completes.
679 .BI ioscheduler \fR=\fPstr
680 Attempt to switch the device hosting the file to the specified I/O scheduler.
682 .BI cpuload \fR=\fPint
683 If the job is a CPU cycle-eater, attempt to use the specified percentage of
686 .BI cpuchunks \fR=\fPint
687 If the job is a CPU cycle-eater, split the load into cycles of the
688 given time in milliseconds.
690 .BI disk_util \fR=\fPbool
691 Generate disk utilization statistics if the platform supports it. Default: true.
693 .BI gtod_reduce \fR=\fPbool
694 Enable all of the gettimeofday() reducing options (disable_clat, disable_slat,
695 disable_bw) plus reduce precision of the timeout somewhat to really shrink the
696 gettimeofday() call count. With this option enabled, we only do about 0.4% of
697 the gtod() calls we would have done if all time keeping was enabled.
699 .BI gtod_cpu \fR=\fPint
700 Sometimes it's cheaper to dedicate a single thread of execution to just getting
701 the current time. Fio (and databases, for instance) are very intensive on
702 gettimeofday() calls. With this option, you can set one CPU aside for doing
703 nothing but logging current time to a shared memory location. Then the other
704 threads/processes that run IO workloads need only copy that segment, instead of
705 entering the kernel with a gettimeofday() call. The CPU set aside for doing
706 these time calls will be excluded from other uses. Fio will manually clear it
707 from the CPU mask of other jobs.
709 .BI continue_on_error \fR=\fPbool
710 Normally fio will exit the job on the first observed failure. If this option is
711 set, fio will continue the job when there is a 'non-fatal error'
712 (\fBEIO\fR or \fBEILSEQ\fR) until the runtime is exceeded or the I/O size
713 specified is completed. If this option is used, there are two more stats that
714 are appended, the total error count and the first error. The error field given
715 in the stats is the first error that was hit during the run.
717 While running, \fBfio\fR will display the status of the created jobs. For
721 Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s]
724 The characters in the first set of brackets denote the current status of each
725 threads. The possible values are:
731 Setup but not started.
737 Initialized, waiting.
740 Running, doing sequential reads.
743 Running, doing random reads.
746 Running, doing sequential writes.
749 Running, doing random writes.
752 Running, doing mixed sequential reads/writes.
755 Running, doing mixed random reads/writes.
758 Running, currently waiting for \fBfsync\fR\|(2).
761 Running, verifying written data.
764 Exited, not reaped by main thread.
767 Exited, thread reaped.
771 The second set of brackets shows the estimated completion percentage of
772 the current group. The third set shows the read and write I/O rate,
773 respectively. Finally, the estimated run time of the job is displayed.
775 When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data
776 for each thread, each group of threads, and each disk, in that order.
778 Per-thread statistics first show the threads client number, group-id, and
779 error code. The remaining figures are as follows:
783 Number of megabytes of I/O performed.
786 Average data rate (bandwidth).
792 Submission latency minimum, maximum, average and standard deviation. This is
793 the time it took to submit the I/O.
796 Completion latency minimum, maximum, average and standard deviation. This
797 is the time between submission and completion.
800 Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average
801 and standard deviation.
804 CPU usage statistics. Includes user and system time, number of context switches
805 this thread went through and number of major and minor page faults.
808 Distribution of I/O depths. Each depth includes everything less than (or equal)
809 to it, but greater than the previous depth.
812 Number of read/write requests issued, and number of short read/write requests.
815 Distribution of I/O completion latencies. The numbers follow the same pattern
819 The group statistics show:
824 Number of megabytes I/O performed.
827 Aggregate bandwidth of threads in the group.
830 Minimum average bandwidth a thread saw.
833 Maximum average bandwidth a thread saw.
836 Shortest runtime of threads in the group.
839 Longest runtime of threads in the group.
843 Finally, disk statistics are printed with reads first:
848 Number of I/Os performed by all groups.
851 Number of merges in the I/O scheduler.
854 Number of ticks we kept the disk busy.
857 Total time spent in the disk queue.
864 If the \fB\-\-minimal\fR option is given, the results will be printed in a
865 semicolon-delimited format suitable for scripted use. The fields are:
868 .B jobname, groupid, error
872 .B KiB I/O, bandwidth \fR(KiB/s)\fP, runtime \fR(ms)\fP
876 .B min, max, mean, standard deviation
880 .B min, max, mean, standard deviation
884 .B min, max, aggregate percentage of total, mean, standard deviation
890 .B KiB I/O, bandwidth \fR(KiB/s)\fP, runtime \fR(ms)\fP
894 .B min, max, mean, standard deviation
898 .B min, max, mean, standard deviation
902 .B min, max, aggregate percentage of total, mean, standard deviation
908 .B user, system, context switches, major page faults, minor page faults
911 IO depth distribution:
913 .B <=1, 2, 4, 8, 16, 32, >=64
916 IO latency distribution (ms):
918 .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, >=2000
925 was written by Jens Axboe <jens.axboe@oracle.com>.
927 This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based
928 on documentation by Jens Axboe.
930 Report bugs to the \fBfio\fR mailing list <fio@vger.kernel.org>.
933 For further documentation see \fBHOWTO\fR and \fBREADME\fR.
935 Sample jobfiles are available in the \fBexamples\fR directory.
projects / fio.git / blob