Add vsync io engine
[fio.git] / README
CommitLineData
ebac4655
JA
1fio
2---
3
79809113
JA
4fio is a tool that will spawn a number of threads or processes doing a
5particular type of io action as specified by the user. fio takes a
6number of global parameters, each inherited by the thread unless
7otherwise parameters given to them overriding that setting is given.
8The typical use of fio is to write a job file matching the io load
9one wants to simulate.
ebac4655 10
2b02b546
JA
11
12Source
13------
14
15fio resides in a git repo, the canonical place is:
16
6b3eccb1 17git://git.kernel.dk/fio.git
97f049c9
JA
18
19The http protocol also works, path is the same.
2b02b546 20
79809113
JA
21Snapshots are frequently generated and they include the git meta data as
22well. You can download them here:
2b02b546
JA
23
24http://brick.kernel.dk/snaps/
25
a68594cb
JA
26Pascal Bleser <guru@unixtech.be> has fio RPMs in his repository for
27SUSE variants, you can find them here:
1053a106
JA
28
29http://linux01.gwdg.de/~pbleser/rpm-navigation.php?cat=System/fio
30
a68594cb
JA
31Dag Wieƫrs has RPMs for Red Hat related distros, find them here:
32
33http://dag.wieers.com/rpm/packages/fio/
34
244e170e
JA
35Mandriva has integrated fio into their package repository, so installing
36on that distro should be as easy as typing 'urpmi fio'.
37
2b02b546 38
726f6ff0
JA
39Mailing list
40------------
41
42There's a mailing list associated with fio. It's meant for general
43discussion, bug reporting, questions - basically anything that has to
44do with fio. An automated mail detailing recent commits is automatically
45sent to the list at most daily. The list address is fio-devel@kernel.dk,
46subscribe by sending an empty email to fio-devel+subscribe@kernel.dk.
47
48
bbfd6b00
JA
49Building
50--------
51
52Just type 'make' and 'make install'. If on FreeBSD, for now you have to
53specify the FreeBSD Makefile with -f, eg:
54
55$ make -f Makefile.Freebsd && make -f Makefile.FreeBSD install
56
edffcb96 57Likewise with OpenSolaris, use the Makefile.solaris to compile there.
bbfd6b00
JA
58This might change in the future if I opt for an autoconf type setup.
59
60
972cfd25
JA
61Command line
62------------
ebac4655
JA
63
64$ fio
ee56ad50 65 --debug Enable some debugging options (see below)
b4692828 66 --output Write output to file
03b74b3e 67 --runtime Runtime in seconds
b4692828
JA
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
fd28ca49
JA
72 --help Print this page
73 --cmdhelp=cmd Print command help, "all" for all of them
cca73aa7 74 --showcmd Turn a job file into command line options
724e4435 75 --readonly Turn on safety read-only checks
e592a06b
AC
76 --eta=when When ETA estimate should be printed
77 May be "always", "never" or "auto"
78
b4692828
JA
79
80Any parameters following the options will be assumed to be job files,
81unless they match a job file parameter. You can add as many as you want,
82each job file will be regarded as a separate group and fio will stonewall
83its execution.
972cfd25 84
724e4435
JA
85The --readonly switch is an extra safety guard to prevent accidentically
86turning on a write setting when that is not desired. Fio will only write
87if rw=write/randwrite/rw/randrw is given, but this extra safety net can
88be used as an extra precaution. It will also enable a write check in the
89io engine core to prevent an accidental write due to a fio bug.
90
ee56ad50
JA
91The debug switch allows adding options that trigger certain logging
92options in fio. Currently the options are:
93
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
bd6f78b2
JA
98 blktrace Dump info related to blktrace setup
99 verify Dump info related to IO verification
100 all Enable all debug options
101 ? or help Show available debug options.
ee56ad50
JA
102
103You can specify as many as you want, eg --debug=file,mem will enable
bd6f78b2 104file and memory debugging.
ee56ad50 105
79809113
JA
106
107Job file
108--------
109
71bfa161 110See the HOWTO file for a more detailed description of parameters and what
4661f3d0
JA
111they mean. This file contains the terse version. You can describe big and
112complex setups with the command line, but generally it's a lot easier to
71bfa161 113just write a simple job file to describe the workload. The job file format
4661f3d0 114is in the ini style format, as that is easy to read and write for the user.
79809113
JA
115
116The job file parameters are:
ebac4655 117
01452055 118 name=x Use 'x' as the identifier for this job.
61697c37 119 description=x 'x' is a text description of the job.
ebac4655 120 directory=x Use 'x' as the top level directory for storing files
b50b8755
JA
121 filename=x Force the use of 'x' as the filename for all files
122 in this thread. If not given, fio will make up
123 a suitable filename based on the thread and file
124 number.
3d60d1ed
JA
125 rw=x 'x' may be: read, randread, write, randwrite,
126 rw (read-write mix), randrw (read-write random mix)
a6ccc7be
JA
127 rwmixcycle=x Base cycle for switching between read and write
128 in msecs.
129 rwmixread=x 'x' percentage of rw mix ios will be reads. If
130 rwmixwrite is also given, the last of the two will
131 be used if they don't add up to 100%.
132 rwmixwrite=x 'x' percentage of rw mix ios will be writes. See
133 rwmixread.
9ebc27e1
JA
134 rand_repeatable=x The sequence of random io blocks can be repeatable
135 across runs, if 'x' is 1.
ebac4655
JA
136 size=x Set file size to x bytes (x string can include k/m/g)
137 ioengine=x 'x' may be: aio/libaio/linuxaio for Linux aio,
138 posixaio for POSIX aio, sync for regular read/write io,
1d2af02a
JA
139 psync for regular pread/pwrite io, vsync for regular
140 readv/writev (with queuing emulation) mmap for mmap'ed
141 io, syslet-rw for syslet driven read/write, splice for
142 using splice/vmsplice, sgio for direct SG_IO io, net
143 for network io, or cpuio for a cycler burner load. sgio
144 only works on Linux on SCSI (or SCSI-like devices, such
145 as usb-storage or sata/libata driven) devices. Fio also
146 has a null io engine, which is mainly used for testing
147 fio itself.
148
ebac4655
JA
149 iodepth=x For async io, allow 'x' ios in flight
150 overwrite=x If 'x', layout a write file first.
53cdc686
JA
151 nrfiles=x Spread io load over 'x' number of files per job,
152 if possible.
ebac4655
JA
153 prio=x Run io at prio X, 0-7 is the kernel allowed range
154 prioclass=x Run io at prio class X
155 bs=x Use 'x' for thread blocksize. May include k/m postfix.
156 bsrange=x-y Mix thread block sizes randomly between x and y. May
157 also include k/m postfix.
158 direct=x 1 for direct IO, 0 for buffered IO
159 thinktime=x "Think" x usec after each io
160 rate=x Throttle rate to x KiB/sec
161 ratemin=x Quit if rate of x KiB/sec can't be met
162 ratecycle=x ratemin averaged over x msecs
163 cpumask=x Only allow job to run on CPUs defined by mask.
d2e268b0 164 cpus_allowed=x Like 'cpumask', but allow text setting of CPU affinity.
795407ca
JA
165 fsync=x If writing with buffered IO, fsync after every
166 'x' blocks have been written.
167 end_fsync=x If 'x', run fsync() after end-of-job.
ebac4655 168 startdelay=x Start this thread x seconds after startup
03b74b3e 169 runtime=x Terminate x seconds after startup. Can include a
906c8d75
JA
170 normal time suffix if not given in seconds, such as
171 'm' for minutes, 'h' for hours, and 'd' for days.
ebac4655
JA
172 offset=x Start io at offset x (x string can include k/m/g)
173 invalidate=x Invalidate page cache for file prior to doing io
795407ca 174 sync=x Use sync writes if x and writing buffered IO.
ebac4655 175 mem=x If x == malloc, use malloc for buffers. If x == shm,
795407ca
JA
176 use shared memory for buffers. If x == mmap, use
177 anonymous mmap.
ebac4655
JA
178 exitall When one thread quits, terminate the others
179 bwavgtime=x Average bandwidth stats over an x msec window.
180 create_serialize=x If 'x', serialize file creation.
181 create_fsync=x If 'x', run fsync() after file creation.
f6cbb269 182 unlink If set, unlink files when done.
ebac4655
JA
183 loops=x Run the job 'x' number of times.
184 verify=x If 'x' == md5, use md5 for verifies. If 'x' == crc32,
185 use crc32 for verifies. md5 is 'safer', but crc32 is
186 a lot faster. Only makes sense for writing to a file.
187 stonewall Wait for preceeding jobs to end before running.
188 numjobs=x Create 'x' similar entries for this job
189 thread Use pthreads instead of forked jobs
20dc95c4
JA
190 zonesize=x
191 zoneskip=y Zone options must be paired. If given, the job
192 will skip y bytes for every x read/written. This
193 can be used to gauge hard drive speed over the entire
194 platter, without reading everything. Both x/y can
195 include k/m/g suffix.
aea47d44
JA
196 iolog=x Open and read io pattern from file 'x'. The file must
197 contain one io action per line in the following format:
198 rw, offset, length
199 where with rw=0/1 for read/write, and the offset
200 and length entries being in bytes.
843a7413
JA
201 write_iolog=x Write an iolog to file 'x' in the same format as iolog.
202 The iolog options are exclusive, if both given the
203 read iolog will be performed.
ec94ec56
JA
204 write_bw_log Write a bandwidth log.
205 write_lat_log Write a latency log.
c04f7ec3
JA
206 lockmem=x Lock down x amount of memory on the machine, to
207 simulate a machine with less memory available. x can
208 include k/m/g suffix.
b6f4d880 209 nice=x Run job at given nice value.
4e0ba8af
JA
210 exec_prerun=x Run 'x' before job io is begun.
211 exec_postrun=x Run 'x' after job io has finished.
da86774e 212 ioscheduler=x Use ioscheduler 'x' for this job.
b990b5c0
JA
213 cpuload=x For a CPU io thread, percentage of CPU time to attempt
214 to burn.
ba0fbe10 215 cpuchunks=x Split burn cycles into pieces of x usecs.
ebac4655 216
79809113 217
79809113
JA
218Author
219------
220
aae22ca7 221Fio was written by Jens Axboe <axboe@kernel.dk> to enable flexible testing
79809113
JA
222of the Linux IO subsystem and schedulers. He got tired of writing
223specific test applications to simulate a given workload, and found that
224the existing io benchmark/test tools out there weren't flexible enough
225to do what he wanted.
226
aae22ca7 227Jens Axboe <axboe@kernel.dk> 20060905
79809113 228