4 fio is a tool that will spawn a number of threads or processes doing a
5 particular type of io action as specified by the user. fio takes a
6 number of global parameters, each inherited by the thread unless
7 otherwise parameters given to them overriding that setting is given.
8 The typical use of fio is to write a job file matching the io load
15 fio resides in a git repo, the canonical place is:
17 git://git.kernel.dk/fio.git
19 The http protocol also works, path is the same.
21 Snapshots are frequently generated and they include the git meta data as
22 well. You can download them here:
24 http://brick.kernel.dk/snaps/
26 Pascal Bleser <guru@unixtech.be> has fio RPMs in his repository for
27 SUSE variants, you can find them here:
29 http://linux01.gwdg.de/~pbleser/rpm-navigation.php?cat=System/fio
31 Dag Wieƫrs has RPMs for Red Hat related distros, find them here:
33 http://dag.wieers.com/rpm/packages/fio/
35 Mandriva has integrated fio into their package repository, so installing
36 on that distro should be as easy as typing 'urpmi fio'.
42 There's a mailing list associated with fio. It's meant for general
43 discussion, bug reporting, questions - basically anything that has to
44 do with fio. An automated mail detailing recent commits is automatically
45 sent to the list at most daily. The list address is fio-devel@kernel.dk,
46 subscribe by sending an empty email to fio-devel+subscribe@kernel.dk.
52 Just type 'make' and 'make install'. If on FreeBSD, for now you have to
53 specify the FreeBSD Makefile with -f, eg:
55 $ make -f Makefile.Freebsd && make -f Makefile.FreeBSD install
57 Likewise with OpenSolaris, use the Makefile.solaris to compile there.
58 This might change in the future if I opt for an autoconf type setup.
65 --debug Enable some debugging options (see below)
66 --output Write output to file
67 --runtime Runtime in seconds
68 --latency-log Generate per-job latency logs
69 --bandwidth-log Generate per-job bandwidth logs
70 --minimal Minimal (terse) output
71 --version Print version info and exit
72 --help Print this page
73 --cmdhelp=cmd Print command help, "all" for all of them
74 --showcmd Turn a job file into command line options
75 --readonly Turn on safety read-only checks
76 --eta=when When ETA estimate should be printed
77 May be "always", "never" or "auto"
80 Any parameters following the options will be assumed to be job files,
81 unless they match a job file parameter. You can add as many as you want,
82 each job file will be regarded as a separate group and fio will stonewall
85 The --readonly switch is an extra safety guard to prevent accidentically
86 turning on a write setting when that is not desired. Fio will only write
87 if rw=write/randwrite/rw/randrw is given, but this extra safety net can
88 be used as an extra precaution. It will also enable a write check in the
89 io engine core to prevent an accidental write due to a fio bug.
91 The debug switch allows adding options that trigger certain logging
92 options in fio. Currently the options are:
94 process Dump info related to processes
95 file Dump info related to file actions
96 io Dump info related to IO queuing
97 mem Dump info related to memory allocations
99 You can specify as many as you want, eg --debug=file,mem will enable
100 file and memory debugging. Specifying --debug=help or --debug=? will
101 dump the current modifier list.
107 See the HOWTO file for a more detailed description of parameters and what
108 they mean. This file contains the terse version. You can describe big and
109 complex setups with the command line, but generally it's a lot easier to
110 just write a simple job file to describe the workload. The job file format
111 is in the ini style format, as that is easy to read and write for the user.
113 The job file parameters are:
115 name=x Use 'x' as the identifier for this job.
116 description=x 'x' is a text description of the job.
117 directory=x Use 'x' as the top level directory for storing files
118 filename=x Force the use of 'x' as the filename for all files
119 in this thread. If not given, fio will make up
120 a suitable filename based on the thread and file
122 rw=x 'x' may be: read, randread, write, randwrite,
123 rw (read-write mix), randrw (read-write random mix)
124 rwmixcycle=x Base cycle for switching between read and write
126 rwmixread=x 'x' percentage of rw mix ios will be reads. If
127 rwmixwrite is also given, the last of the two will
128 be used if they don't add up to 100%.
129 rwmixwrite=x 'x' percentage of rw mix ios will be writes. See
131 rand_repeatable=x The sequence of random io blocks can be repeatable
132 across runs, if 'x' is 1.
133 size=x Set file size to x bytes (x string can include k/m/g)
134 ioengine=x 'x' may be: aio/libaio/linuxaio for Linux aio,
135 posixaio for POSIX aio, sync for regular read/write io,
136 psync for regular pread/pwrite io, mmap for mmap'ed io,
137 syslet-rw for syslet driven read/write, splice for using
138 splice/vmsplice, sgio for direct SG_IO io, net for
139 network io, or cpuio for a cycler burner load. sgio only
140 works on Linux on SCSI (or SCSI-like devices, such as
141 usb-storage or sata/libata driven) devices. Fio also has
142 a null io engine, which is mainly used for testing fio
144 iodepth=x For async io, allow 'x' ios in flight
145 overwrite=x If 'x', layout a write file first.
146 nrfiles=x Spread io load over 'x' number of files per job,
148 prio=x Run io at prio X, 0-7 is the kernel allowed range
149 prioclass=x Run io at prio class X
150 bs=x Use 'x' for thread blocksize. May include k/m postfix.
151 bsrange=x-y Mix thread block sizes randomly between x and y. May
152 also include k/m postfix.
153 direct=x 1 for direct IO, 0 for buffered IO
154 thinktime=x "Think" x usec after each io
155 rate=x Throttle rate to x KiB/sec
156 ratemin=x Quit if rate of x KiB/sec can't be met
157 ratecycle=x ratemin averaged over x msecs
158 cpumask=x Only allow job to run on CPUs defined by mask.
159 cpus_allowed=x Like 'cpumask', but allow text setting of CPU affinity.
160 fsync=x If writing with buffered IO, fsync after every
161 'x' blocks have been written.
162 end_fsync=x If 'x', run fsync() after end-of-job.
163 startdelay=x Start this thread x seconds after startup
164 runtime=x Terminate x seconds after startup. Can include a
165 normal time suffix if not given in seconds, such as
166 'm' for minutes, 'h' for hours, and 'd' for days.
167 offset=x Start io at offset x (x string can include k/m/g)
168 invalidate=x Invalidate page cache for file prior to doing io
169 sync=x Use sync writes if x and writing buffered IO.
170 mem=x If x == malloc, use malloc for buffers. If x == shm,
171 use shared memory for buffers. If x == mmap, use
173 exitall When one thread quits, terminate the others
174 bwavgtime=x Average bandwidth stats over an x msec window.
175 create_serialize=x If 'x', serialize file creation.
176 create_fsync=x If 'x', run fsync() after file creation.
177 unlink If set, unlink files when done.
178 loops=x Run the job 'x' number of times.
179 verify=x If 'x' == md5, use md5 for verifies. If 'x' == crc32,
180 use crc32 for verifies. md5 is 'safer', but crc32 is
181 a lot faster. Only makes sense for writing to a file.
182 stonewall Wait for preceeding jobs to end before running.
183 numjobs=x Create 'x' similar entries for this job
184 thread Use pthreads instead of forked jobs
186 zoneskip=y Zone options must be paired. If given, the job
187 will skip y bytes for every x read/written. This
188 can be used to gauge hard drive speed over the entire
189 platter, without reading everything. Both x/y can
190 include k/m/g suffix.
191 iolog=x Open and read io pattern from file 'x'. The file must
192 contain one io action per line in the following format:
194 where with rw=0/1 for read/write, and the offset
195 and length entries being in bytes.
196 write_iolog=x Write an iolog to file 'x' in the same format as iolog.
197 The iolog options are exclusive, if both given the
198 read iolog will be performed.
199 write_bw_log Write a bandwidth log.
200 write_lat_log Write a latency log.
201 lockmem=x Lock down x amount of memory on the machine, to
202 simulate a machine with less memory available. x can
203 include k/m/g suffix.
204 nice=x Run job at given nice value.
205 exec_prerun=x Run 'x' before job io is begun.
206 exec_postrun=x Run 'x' after job io has finished.
207 ioscheduler=x Use ioscheduler 'x' for this job.
208 cpuload=x For a CPU io thread, percentage of CPU time to attempt
210 cpuchunks=x Split burn cycles into pieces of x usecs.
216 Fio was written by Jens Axboe <axboe@kernel.dk> to enable flexible testing
217 of the Linux IO subsystem and schedulers. He got tired of writing
218 specific test applications to simulate a given workload, and found that
219 the existing io benchmark/test tools out there weren't flexible enough
220 to do what he wanted.
222 Jens Axboe <axboe@kernel.dk> 20060905