[PATCH] IO completion time histograms
[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
17git://brick.kernel.dk/data/git/fio.git
18
79809113
JA
19Snapshots are frequently generated and they include the git meta data as
20well. You can download them here:
2b02b546
JA
21
22http://brick.kernel.dk/snaps/
23
1053a106
JA
24Pascal Bleser <guru@unixtech.be> has fio RPMs in his repository, you
25can find them here:
26
27http://linux01.gwdg.de/~pbleser/rpm-navigation.php?cat=System/fio
28
2b02b546 29
726f6ff0
JA
30Mailing list
31------------
32
33There's a mailing list associated with fio. It's meant for general
34discussion, bug reporting, questions - basically anything that has to
35do with fio. An automated mail detailing recent commits is automatically
36sent to the list at most daily. The list address is fio-devel@kernel.dk,
37subscribe by sending an empty email to fio-devel+subscribe@kernel.dk.
38
39
bbfd6b00
JA
40Building
41--------
42
43Just type 'make' and 'make install'. If on FreeBSD, for now you have to
44specify the FreeBSD Makefile with -f, eg:
45
46$ make -f Makefile.Freebsd && make -f Makefile.FreeBSD install
47
edffcb96 48Likewise with OpenSolaris, use the Makefile.solaris to compile there.
bbfd6b00
JA
49This might change in the future if I opt for an autoconf type setup.
50
51
972cfd25
JA
52Command line
53------------
ebac4655
JA
54
55$ fio
b4692828 56 --output Write output to file
03b74b3e 57 --runtime Runtime in seconds
b4692828
JA
58 --latency-log Generate per-job latency logs
59 --bandwidth-log Generate per-job bandwidth logs
60 --minimal Minimal (terse) output
61 --version Print version info and exit
fd28ca49
JA
62 --help Print this page
63 --cmdhelp=cmd Print command help, "all" for all of them
b4692828
JA
64
65Any parameters following the options will be assumed to be job files,
66unless they match a job file parameter. You can add as many as you want,
67each job file will be regarded as a separate group and fio will stonewall
68its execution.
972cfd25 69
79809113
JA
70
71Job file
72--------
73
71bfa161
JA
74See the HOWTO file for a more detailed description of parameters and what
75they mean. This file contains the terse version. Only a few options can
76be controlled with command line parameters, generally it's a lot easier to
77just write a simple job file to describe the workload. The job file format
78is in the ini style format, as it's easy to read and write for the user.
79809113
JA
79
80The job file parameters are:
ebac4655 81
01452055 82 name=x Use 'x' as the identifier for this job.
61697c37 83 description=x 'x' is a text description of the job.
ebac4655 84 directory=x Use 'x' as the top level directory for storing files
b50b8755
JA
85 filename=x Force the use of 'x' as the filename for all files
86 in this thread. If not given, fio will make up
87 a suitable filename based on the thread and file
88 number.
3d60d1ed
JA
89 rw=x 'x' may be: read, randread, write, randwrite,
90 rw (read-write mix), randrw (read-write random mix)
a6ccc7be
JA
91 rwmixcycle=x Base cycle for switching between read and write
92 in msecs.
93 rwmixread=x 'x' percentage of rw mix ios will be reads. If
94 rwmixwrite is also given, the last of the two will
95 be used if they don't add up to 100%.
96 rwmixwrite=x 'x' percentage of rw mix ios will be writes. See
97 rwmixread.
9ebc27e1
JA
98 rand_repeatable=x The sequence of random io blocks can be repeatable
99 across runs, if 'x' is 1.
ebac4655
JA
100 size=x Set file size to x bytes (x string can include k/m/g)
101 ioengine=x 'x' may be: aio/libaio/linuxaio for Linux aio,
102 posixaio for POSIX aio, sync for regular read/write io,
d0ff85df
JA
103 mmap for mmap'ed io, syslet-rw for syslet driven
104 read/write, splice for using splice/vmsplice,
ed92ac0c
JA
105 sgio for direct SG_IO io, or net for network io. sgio
106 only works on Linux on SCSI (or SCSI-like devices,
107 such as usb-storage or sata/libata driven) devices.
108 Fio also has a null io engine, which is mainly used
109 for testing fio itself.
ebac4655
JA
110 iodepth=x For async io, allow 'x' ios in flight
111 overwrite=x If 'x', layout a write file first.
53cdc686
JA
112 nrfiles=x Spread io load over 'x' number of files per job,
113 if possible.
ebac4655
JA
114 prio=x Run io at prio X, 0-7 is the kernel allowed range
115 prioclass=x Run io at prio class X
116 bs=x Use 'x' for thread blocksize. May include k/m postfix.
117 bsrange=x-y Mix thread block sizes randomly between x and y. May
118 also include k/m postfix.
119 direct=x 1 for direct IO, 0 for buffered IO
120 thinktime=x "Think" x usec after each io
121 rate=x Throttle rate to x KiB/sec
122 ratemin=x Quit if rate of x KiB/sec can't be met
123 ratecycle=x ratemin averaged over x msecs
124 cpumask=x Only allow job to run on CPUs defined by mask.
795407ca
JA
125 fsync=x If writing with buffered IO, fsync after every
126 'x' blocks have been written.
127 end_fsync=x If 'x', run fsync() after end-of-job.
ebac4655 128 startdelay=x Start this thread x seconds after startup
03b74b3e 129 runtime=x Terminate x seconds after startup. Can include a
906c8d75
JA
130 normal time suffix if not given in seconds, such as
131 'm' for minutes, 'h' for hours, and 'd' for days.
ebac4655
JA
132 offset=x Start io at offset x (x string can include k/m/g)
133 invalidate=x Invalidate page cache for file prior to doing io
795407ca 134 sync=x Use sync writes if x and writing buffered IO.
ebac4655 135 mem=x If x == malloc, use malloc for buffers. If x == shm,
795407ca
JA
136 use shared memory for buffers. If x == mmap, use
137 anonymous mmap.
ebac4655
JA
138 exitall When one thread quits, terminate the others
139 bwavgtime=x Average bandwidth stats over an x msec window.
140 create_serialize=x If 'x', serialize file creation.
141 create_fsync=x If 'x', run fsync() after file creation.
f6cbb269 142 unlink If set, unlink files when done.
ebac4655
JA
143 loops=x Run the job 'x' number of times.
144 verify=x If 'x' == md5, use md5 for verifies. If 'x' == crc32,
145 use crc32 for verifies. md5 is 'safer', but crc32 is
146 a lot faster. Only makes sense for writing to a file.
147 stonewall Wait for preceeding jobs to end before running.
148 numjobs=x Create 'x' similar entries for this job
149 thread Use pthreads instead of forked jobs
20dc95c4
JA
150 zonesize=x
151 zoneskip=y Zone options must be paired. If given, the job
152 will skip y bytes for every x read/written. This
153 can be used to gauge hard drive speed over the entire
154 platter, without reading everything. Both x/y can
155 include k/m/g suffix.
aea47d44
JA
156 iolog=x Open and read io pattern from file 'x'. The file must
157 contain one io action per line in the following format:
158 rw, offset, length
159 where with rw=0/1 for read/write, and the offset
160 and length entries being in bytes.
843a7413
JA
161 write_iolog=x Write an iolog to file 'x' in the same format as iolog.
162 The iolog options are exclusive, if both given the
163 read iolog will be performed.
ec94ec56
JA
164 write_bw_log Write a bandwidth log.
165 write_lat_log Write a latency log.
c04f7ec3
JA
166 lockmem=x Lock down x amount of memory on the machine, to
167 simulate a machine with less memory available. x can
168 include k/m/g suffix.
b6f4d880 169 nice=x Run job at given nice value.
4e0ba8af
JA
170 exec_prerun=x Run 'x' before job io is begun.
171 exec_postrun=x Run 'x' after job io has finished.
da86774e 172 ioscheduler=x Use ioscheduler 'x' for this job.
b990b5c0
JA
173 cpuload=x For a CPU io thread, percentage of CPU time to attempt
174 to burn.
175 cpuchunks=x Split burn cycles into pieces of x.
ebac4655 176
79809113 177
79809113
JA
178Author
179------
180
aae22ca7 181Fio was written by Jens Axboe <axboe@kernel.dk> to enable flexible testing
79809113
JA
182of the Linux IO subsystem and schedulers. He got tired of writing
183specific test applications to simulate a given workload, and found that
184the existing io benchmark/test tools out there weren't flexible enough
185to do what he wanted.
186
aae22ca7 187Jens Axboe <axboe@kernel.dk> 20060905
79809113 188