| 1 | .TH fio 1 "September 2007" "User Manual" |
| 2 | .SH NAME |
| 3 | fio \- flexible I/O tester |
| 4 | .SH SYNOPSIS |
| 5 | .B fio |
| 6 | [\fIoptions\fR] [\fIjobfile\fR]... |
| 7 | .SH DESCRIPTION |
| 8 | .B fio |
| 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. |
| 13 | .SH OPTIONS |
| 14 | .TP |
| 15 | .BI \-\-debug \fR=\fPtype |
| 16 | Enable verbose tracing of various fio actions. May be `all' for all types |
| 17 | or individual types separated by a comma (eg \-\-debug=io,file). `help' will |
| 18 | list all available tracing options. |
| 19 | .TP |
| 20 | .BI \-\-output \fR=\fPfilename |
| 21 | Write output to \fIfilename\fR. |
| 22 | .TP |
| 23 | .BI \-\-timeout \fR=\fPtimeout |
| 24 | Limit run time to \fItimeout\fR seconds. |
| 25 | .TP |
| 26 | .B \-\-latency\-log |
| 27 | Generate per-job latency logs. |
| 28 | .TP |
| 29 | .B \-\-bandwidth\-log |
| 30 | Generate per-job bandwidth logs. |
| 31 | .TP |
| 32 | .B \-\-minimal |
| 33 | Print statistics in a terse, semicolon-delimited format. |
| 34 | .TP |
| 35 | .B \-\-version |
| 36 | Display version information and exit. |
| 37 | .TP |
| 38 | .BI \-\-terse\-version \fR=\fPversion |
| 39 | Set terse version output format (Current version 3, or older version 2). |
| 40 | .TP |
| 41 | .B \-\-help |
| 42 | Display usage information and exit. |
| 43 | .TP |
| 44 | .BI \-\-cmdhelp \fR=\fPcommand |
| 45 | Print help information for \fIcommand\fR. May be `all' for all commands. |
| 46 | .TP |
| 47 | .BI \-\-enghelp \fR=\fPioengine[,command] |
| 48 | List all commands defined by \fIioengine\fR, or print help for \fIcommand\fR defined by \fIioengine\fR. |
| 49 | .TP |
| 50 | .BI \-\-showcmd \fR=\fPjobfile |
| 51 | Convert \fIjobfile\fR to a set of command-line options. |
| 52 | .TP |
| 53 | .B \-\-readonly |
| 54 | Enable read-only safety checks. |
| 55 | .TP |
| 56 | .BI \-\-eta \fR=\fPwhen |
| 57 | Specifies when real-time ETA estimate should be printed. \fIwhen\fR may |
| 58 | be one of `always', `never' or `auto'. |
| 59 | .TP |
| 60 | .BI \-\-readonly |
| 61 | Turn on safety read-only checks, preventing any attempted write. |
| 62 | .TP |
| 63 | .BI \-\-section \fR=\fPsec |
| 64 | Only run section \fIsec\fR from job file. Multiple of these options can be given, adding more sections to run. |
| 65 | .TP |
| 66 | .BI \-\-alloc\-size \fR=\fPkb |
| 67 | Set the internal smalloc pool size to \fIkb\fP kilobytes. |
| 68 | .TP |
| 69 | .BI \-\-warnings\-fatal |
| 70 | All fio parser warnings are fatal, causing fio to exit with an error. |
| 71 | .TP |
| 72 | .BI \-\-max\-jobs \fR=\fPnr |
| 73 | Set the maximum allowed number of jobs (threads/processes) to suport. |
| 74 | .TP |
| 75 | .BI \-\-server \fR=\fPargs |
| 76 | Start a backend server, with \fIargs\fP specifying what to listen to. See client/server section. |
| 77 | .TP |
| 78 | .BI \-\-daemonize \fR=\fPpidfile |
| 79 | Background a fio server, writing the pid to the given pid file. |
| 80 | .TP |
| 81 | .BI \-\-client \fR=\fPhost |
| 82 | Instead of running the jobs locally, send and run them on the given host. |
| 83 | .SH "JOB FILE FORMAT" |
| 84 | Job files are in `ini' format. They consist of one or more |
| 85 | job definitions, which begin with a job name in square brackets and |
| 86 | extend to the next job name. The job name can be any ASCII string |
| 87 | except `global', which has a special meaning. Following the job name is |
| 88 | a sequence of zero or more parameters, one per line, that define the |
| 89 | behavior of the job. Any line starting with a `;' or `#' character is |
| 90 | considered a comment and ignored. |
| 91 | .P |
| 92 | If \fIjobfile\fR is specified as `-', the job file will be read from |
| 93 | standard input. |
| 94 | .SS "Global Section" |
| 95 | The global section contains default parameters for jobs specified in the |
| 96 | job file. A job is only affected by global sections residing above it, |
| 97 | and there may be any number of global sections. Specific job definitions |
| 98 | may override any parameter set in global sections. |
| 99 | .SH "JOB PARAMETERS" |
| 100 | .SS Types |
| 101 | Some parameters may take arguments of a specific type. The types used are: |
| 102 | .TP |
| 103 | .I str |
| 104 | String: a sequence of alphanumeric characters. |
| 105 | .TP |
| 106 | .I int |
| 107 | SI integer: a whole number, possibly containing a suffix denoting the base unit |
| 108 | of the value. Accepted suffixes are `k', 'M', 'G', 'T', and 'P', denoting |
| 109 | kilo (1024), mega (1024^2), giga (1024^3), tera (1024^4), and peta (1024^5) |
| 110 | respectively. The suffix is not case sensitive. If prefixed with '0x', the |
| 111 | value is assumed to be base 16 (hexadecimal). A suffix may include a trailing 'b', |
| 112 | for instance 'kb' is identical to 'k'. You can specify a base 10 value |
| 113 | by using 'KiB', 'MiB', 'GiB', etc. This is useful for disk drives where |
| 114 | values are often given in base 10 values. Specifying '30GiB' will get you |
| 115 | 30*1000^3 bytes. |
| 116 | .TP |
| 117 | .I bool |
| 118 | Boolean: a true or false value. `0' denotes false, `1' denotes true. |
| 119 | .TP |
| 120 | .I irange |
| 121 | Integer range: a range of integers specified in the format |
| 122 | \fIlower\fR:\fIupper\fR or \fIlower\fR\-\fIupper\fR. \fIlower\fR and |
| 123 | \fIupper\fR may contain a suffix as described above. If an option allows two |
| 124 | sets of ranges, they are separated with a `,' or `/' character. For example: |
| 125 | `8\-8k/8M\-4G'. |
| 126 | .TP |
| 127 | .I float_list |
| 128 | List of floating numbers: A list of floating numbers, separated by |
| 129 | a ':' charcater. |
| 130 | .SS "Parameter List" |
| 131 | .TP |
| 132 | .BI name \fR=\fPstr |
| 133 | May be used to override the job name. On the command line, this parameter |
| 134 | has the special purpose of signalling the start of a new job. |
| 135 | .TP |
| 136 | .BI description \fR=\fPstr |
| 137 | Human-readable description of the job. It is printed when the job is run, but |
| 138 | otherwise has no special purpose. |
| 139 | .TP |
| 140 | .BI directory \fR=\fPstr |
| 141 | Prefix filenames with this directory. Used to place files in a location other |
| 142 | than `./'. |
| 143 | .TP |
| 144 | .BI filename \fR=\fPstr |
| 145 | .B fio |
| 146 | normally makes up a file name based on the job name, thread number, and file |
| 147 | number. If you want to share files between threads in a job or several jobs, |
| 148 | specify a \fIfilename\fR for each of them to override the default. |
| 149 | If the I/O engine is file-based, you can specify |
| 150 | a number of files by separating the names with a `:' character. `\-' is a |
| 151 | reserved name, meaning stdin or stdout, depending on the read/write direction |
| 152 | set. |
| 153 | .TP |
| 154 | .BI lockfile \fR=\fPstr |
| 155 | Fio defaults to not locking any files before it does IO to them. If a file or |
| 156 | file descriptor is shared, fio can serialize IO to that file to make the end |
| 157 | result consistent. This is usual for emulating real workloads that share files. |
| 158 | The lock modes are: |
| 159 | .RS |
| 160 | .RS |
| 161 | .TP |
| 162 | .B none |
| 163 | No locking. This is the default. |
| 164 | .TP |
| 165 | .B exclusive |
| 166 | Only one thread or process may do IO at the time, excluding all others. |
| 167 | .TP |
| 168 | .B readwrite |
| 169 | Read-write locking on the file. Many readers may access the file at the same |
| 170 | time, but writes get exclusive access. |
| 171 | .RE |
| 172 | .P |
| 173 | The option may be post-fixed with a lock batch number. If set, then each |
| 174 | thread/process may do that amount of IOs to the file before giving up the lock. |
| 175 | Since lock acquisition is expensive, batching the lock/unlocks will speed up IO. |
| 176 | .RE |
| 177 | .P |
| 178 | .BI opendir \fR=\fPstr |
| 179 | Recursively open any files below directory \fIstr\fR. |
| 180 | .TP |
| 181 | .BI readwrite \fR=\fPstr "\fR,\fP rw" \fR=\fPstr |
| 182 | Type of I/O pattern. Accepted values are: |
| 183 | .RS |
| 184 | .RS |
| 185 | .TP |
| 186 | .B read |
| 187 | Sequential reads. |
| 188 | .TP |
| 189 | .B write |
| 190 | Sequential writes. |
| 191 | .TP |
| 192 | .B randread |
| 193 | Random reads. |
| 194 | .TP |
| 195 | .B randwrite |
| 196 | Random writes. |
| 197 | .TP |
| 198 | .B rw |
| 199 | Mixed sequential reads and writes. |
| 200 | .TP |
| 201 | .B randrw |
| 202 | Mixed random reads and writes. |
| 203 | .RE |
| 204 | .P |
| 205 | For mixed I/O, the default split is 50/50. For certain types of io the result |
| 206 | may still be skewed a bit, since the speed may be different. It is possible to |
| 207 | specify a number of IO's to do before getting a new offset, this is one by |
| 208 | appending a `:\fI<nr>\fR to the end of the string given. For a random read, it |
| 209 | would look like \fBrw=randread:8\fR for passing in an offset modifier with a |
| 210 | value of 8. If the postfix is used with a sequential IO pattern, then the value |
| 211 | specified will be added to the generated offset for each IO. For instance, |
| 212 | using \fBrw=write:4k\fR will skip 4k for every write. It turns sequential IO |
| 213 | into sequential IO with holes. See the \fBrw_sequencer\fR option. |
| 214 | .RE |
| 215 | .TP |
| 216 | .BI rw_sequencer \fR=\fPstr |
| 217 | If an offset modifier is given by appending a number to the \fBrw=<str>\fR line, |
| 218 | then this option controls how that number modifies the IO offset being |
| 219 | generated. Accepted values are: |
| 220 | .RS |
| 221 | .RS |
| 222 | .TP |
| 223 | .B sequential |
| 224 | Generate sequential offset |
| 225 | .TP |
| 226 | .B identical |
| 227 | Generate the same offset |
| 228 | .RE |
| 229 | .P |
| 230 | \fBsequential\fR is only useful for random IO, where fio would normally |
| 231 | generate a new random offset for every IO. If you append eg 8 to randread, you |
| 232 | would get a new random offset for every 8 IO's. The result would be a seek for |
| 233 | only every 8 IO's, instead of for every IO. Use \fBrw=randread:8\fR to specify |
| 234 | that. As sequential IO is already sequential, setting \fBsequential\fR for that |
| 235 | would not result in any differences. \fBidentical\fR behaves in a similar |
| 236 | fashion, except it sends the same offset 8 number of times before generating a |
| 237 | new offset. |
| 238 | .RE |
| 239 | .P |
| 240 | .TP |
| 241 | .BI kb_base \fR=\fPint |
| 242 | The base unit for a kilobyte. The defacto base is 2^10, 1024. Storage |
| 243 | manufacturers like to use 10^3 or 1000 as a base ten unit instead, for obvious |
| 244 | reasons. Allow values are 1024 or 1000, with 1024 being the default. |
| 245 | .TP |
| 246 | .BI randrepeat \fR=\fPbool |
| 247 | Seed the random number generator in a predictable way so results are repeatable |
| 248 | across runs. Default: true. |
| 249 | .TP |
| 250 | .BI use_os_rand \fR=\fPbool |
| 251 | Fio can either use the random generator supplied by the OS to generator random |
| 252 | offsets, or it can use it's own internal generator (based on Tausworthe). |
| 253 | Default is to use the internal generator, which is often of better quality and |
| 254 | faster. Default: false. |
| 255 | .TP |
| 256 | .BI fallocate \fR=\fPstr |
| 257 | Whether pre-allocation is performed when laying down files. Accepted values |
| 258 | are: |
| 259 | .RS |
| 260 | .RS |
| 261 | .TP |
| 262 | .B none |
| 263 | Do not pre-allocate space. |
| 264 | .TP |
| 265 | .B posix |
| 266 | Pre-allocate via posix_fallocate(). |
| 267 | .TP |
| 268 | .B keep |
| 269 | Pre-allocate via fallocate() with FALLOC_FL_KEEP_SIZE set. |
| 270 | .TP |
| 271 | .B 0 |
| 272 | Backward-compatible alias for 'none'. |
| 273 | .TP |
| 274 | .B 1 |
| 275 | Backward-compatible alias for 'posix'. |
| 276 | .RE |
| 277 | .P |
| 278 | May not be available on all supported platforms. 'keep' is only |
| 279 | available on Linux. If using ZFS on Solaris this must be set to 'none' |
| 280 | because ZFS doesn't support it. Default: 'posix'. |
| 281 | .RE |
| 282 | .TP |
| 283 | .BI fadvise_hint \fR=\fPbool |
| 284 | Use of \fIposix_fadvise\fR\|(2) to advise the kernel what I/O patterns |
| 285 | are likely to be issued. Default: true. |
| 286 | .TP |
| 287 | .BI size \fR=\fPint |
| 288 | Total size of I/O for this job. \fBfio\fR will run until this many bytes have |
| 289 | been transfered, unless limited by other options (\fBruntime\fR, for instance). |
| 290 | Unless \fBnrfiles\fR and \fBfilesize\fR options are given, this amount will be |
| 291 | divided between the available files for the job. If not set, fio will use the |
| 292 | full size of the given files or devices. If the the files do not exist, size |
| 293 | must be given. It is also possible to give size as a percentage between 1 and |
| 294 | 100. If size=20% is given, fio will use 20% of the full size of the given files |
| 295 | or devices. |
| 296 | .TP |
| 297 | .BI fill_device \fR=\fPbool "\fR,\fB fill_fs" \fR=\fPbool |
| 298 | Sets size to something really large and waits for ENOSPC (no space left on |
| 299 | device) as the terminating condition. Only makes sense with sequential write. |
| 300 | For a read workload, the mount point will be filled first then IO started on |
| 301 | the result. This option doesn't make sense if operating on a raw device node, |
| 302 | since the size of that is already known by the file system. Additionally, |
| 303 | writing beyond end-of-device will not return ENOSPC there. |
| 304 | .TP |
| 305 | .BI filesize \fR=\fPirange |
| 306 | Individual file sizes. May be a range, in which case \fBfio\fR will select sizes |
| 307 | for files at random within the given range, limited to \fBsize\fR in total (if |
| 308 | that is given). If \fBfilesize\fR is not specified, each created file is the |
| 309 | same size. |
| 310 | .TP |
| 311 | .BI blocksize \fR=\fPint[,int] "\fR,\fB bs" \fR=\fPint[,int] |
| 312 | Block size for I/O units. Default: 4k. Values for reads and writes can be |
| 313 | specified separately in the format \fIread\fR,\fIwrite\fR, either of |
| 314 | which may be empty to leave that value at its default. |
| 315 | .TP |
| 316 | .BI blocksize_range \fR=\fPirange[,irange] "\fR,\fB bsrange" \fR=\fPirange[,irange] |
| 317 | Specify a range of I/O block sizes. The issued I/O unit will always be a |
| 318 | multiple of the minimum size, unless \fBblocksize_unaligned\fR is set. Applies |
| 319 | to both reads and writes if only one range is given, but can be specified |
| 320 | separately with a comma seperating the values. Example: bsrange=1k-4k,2k-8k. |
| 321 | Also (see \fBblocksize\fR). |
| 322 | .TP |
| 323 | .BI bssplit \fR=\fPstr |
| 324 | This option allows even finer grained control of the block sizes issued, |
| 325 | not just even splits between them. With this option, you can weight various |
| 326 | block sizes for exact control of the issued IO for a job that has mixed |
| 327 | block sizes. The format of the option is bssplit=blocksize/percentage, |
| 328 | optionally adding as many definitions as needed separated by a colon. |
| 329 | Example: bssplit=4k/10:64k/50:32k/40 would issue 50% 64k blocks, 10% 4k |
| 330 | blocks and 40% 32k blocks. \fBbssplit\fR also supports giving separate |
| 331 | splits to reads and writes. The format is identical to what the |
| 332 | \fBbs\fR option accepts, the read and write parts are separated with a |
| 333 | comma. |
| 334 | .TP |
| 335 | .B blocksize_unaligned\fR,\fP bs_unaligned |
| 336 | If set, any size in \fBblocksize_range\fR may be used. This typically won't |
| 337 | work with direct I/O, as that normally requires sector alignment. |
| 338 | .TP |
| 339 | .BI blockalign \fR=\fPint[,int] "\fR,\fB ba" \fR=\fPint[,int] |
| 340 | At what boundary to align random IO offsets. Defaults to the same as 'blocksize' |
| 341 | the minimum blocksize given. Minimum alignment is typically 512b |
| 342 | for using direct IO, though it usually depends on the hardware block size. |
| 343 | This option is mutually exclusive with using a random map for files, so it |
| 344 | will turn off that option. |
| 345 | .TP |
| 346 | .B zero_buffers |
| 347 | Initialise buffers with all zeros. Default: fill buffers with random data. |
| 348 | .TP |
| 349 | .B refill_buffers |
| 350 | If this option is given, fio will refill the IO buffers on every submit. The |
| 351 | default is to only fill it at init time and reuse that data. Only makes sense |
| 352 | if zero_buffers isn't specified, naturally. If data verification is enabled, |
| 353 | refill_buffers is also automatically enabled. |
| 354 | .TP |
| 355 | .BI scramble_buffers \fR=\fPbool |
| 356 | If \fBrefill_buffers\fR is too costly and the target is using data |
| 357 | deduplication, then setting this option will slightly modify the IO buffer |
| 358 | contents to defeat normal de-dupe attempts. This is not enough to defeat |
| 359 | more clever block compression attempts, but it will stop naive dedupe |
| 360 | of blocks. Default: true. |
| 361 | .TP |
| 362 | .BI buffer_compress_percentage \fR=\fPint |
| 363 | If this is set, then fio will attempt to provide IO buffer content (on WRITEs) |
| 364 | that compress to the specified level. Fio does this by providing a mix of |
| 365 | random data and zeroes. Note that this is per block size unit, for file/disk |
| 366 | wide compression level that matches this setting, you'll also want to set |
| 367 | \fBrefill_buffers\fR. |
| 368 | .TP |
| 369 | .BI buffer_compress_chunk \fR=\fPint |
| 370 | See \fBbuffer_compress_percentage\fR. This setting allows fio to manage how |
| 371 | big the ranges of random data and zeroed data is. Without this set, fio will |
| 372 | provide \fBbuffer_compress_percentage\fR of blocksize random data, followed by |
| 373 | the remaining zeroed. With this set to some chunk size smaller than the block |
| 374 | size, fio can alternate random and zeroed data throughout the IO buffer. |
| 375 | .TP |
| 376 | .BI nrfiles \fR=\fPint |
| 377 | Number of files to use for this job. Default: 1. |
| 378 | .TP |
| 379 | .BI openfiles \fR=\fPint |
| 380 | Number of files to keep open at the same time. Default: \fBnrfiles\fR. |
| 381 | .TP |
| 382 | .BI file_service_type \fR=\fPstr |
| 383 | Defines how files to service are selected. The following types are defined: |
| 384 | .RS |
| 385 | .RS |
| 386 | .TP |
| 387 | .B random |
| 388 | Choose a file at random |
| 389 | .TP |
| 390 | .B roundrobin |
| 391 | Round robin over open files (default). |
| 392 | .B sequential |
| 393 | Do each file in the set sequentially. |
| 394 | .RE |
| 395 | .P |
| 396 | The number of I/Os to issue before switching a new file can be specified by |
| 397 | appending `:\fIint\fR' to the service type. |
| 398 | .RE |
| 399 | .TP |
| 400 | .BI ioengine \fR=\fPstr |
| 401 | Defines how the job issues I/O. The following types are defined: |
| 402 | .RS |
| 403 | .RS |
| 404 | .TP |
| 405 | .B sync |
| 406 | Basic \fIread\fR\|(2) or \fIwrite\fR\|(2) I/O. \fIfseek\fR\|(2) is used to |
| 407 | position the I/O location. |
| 408 | .TP |
| 409 | .B psync |
| 410 | Basic \fIpread\fR\|(2) or \fIpwrite\fR\|(2) I/O. |
| 411 | .TP |
| 412 | .B vsync |
| 413 | Basic \fIreadv\fR\|(2) or \fIwritev\fR\|(2) I/O. Will emulate queuing by |
| 414 | coalescing adjacents IOs into a single submission. |
| 415 | .TP |
| 416 | .B libaio |
| 417 | Linux native asynchronous I/O. This ioengine defines engine specific options. |
| 418 | .TP |
| 419 | .B posixaio |
| 420 | POSIX asynchronous I/O using \fIaio_read\fR\|(3) and \fIaio_write\fR\|(3). |
| 421 | .TP |
| 422 | .B solarisaio |
| 423 | Solaris native asynchronous I/O. |
| 424 | .TP |
| 425 | .B windowsaio |
| 426 | Windows native asynchronous I/O. |
| 427 | .TP |
| 428 | .B mmap |
| 429 | File is memory mapped with \fImmap\fR\|(2) and data copied using |
| 430 | \fImemcpy\fR\|(3). |
| 431 | .TP |
| 432 | .B splice |
| 433 | \fIsplice\fR\|(2) is used to transfer the data and \fIvmsplice\fR\|(2) to |
| 434 | transfer data from user-space to the kernel. |
| 435 | .TP |
| 436 | .B syslet-rw |
| 437 | Use the syslet system calls to make regular read/write asynchronous. |
| 438 | .TP |
| 439 | .B sg |
| 440 | SCSI generic sg v3 I/O. May be either synchronous using the SG_IO ioctl, or if |
| 441 | the target is an sg character device, we use \fIread\fR\|(2) and |
| 442 | \fIwrite\fR\|(2) for asynchronous I/O. |
| 443 | .TP |
| 444 | .B null |
| 445 | Doesn't transfer any data, just pretends to. Mainly used to exercise \fBfio\fR |
| 446 | itself and for debugging and testing purposes. |
| 447 | .TP |
| 448 | .B net |
| 449 | Transfer over the network. The protocol to be used can be defined with the |
| 450 | \fBprotocol\fR parameter. Depending on the protocol, \fBfilename\fR, |
| 451 | \fBhostname\fR, \fBport\fR, or \fBlisten\fR must be specified. |
| 452 | This ioengine defines engine specific options. |
| 453 | .TP |
| 454 | .B netsplice |
| 455 | Like \fBnet\fR, but uses \fIsplice\fR\|(2) and \fIvmsplice\fR\|(2) to map data |
| 456 | and send/receive. This ioengine defines engine specific options. |
| 457 | .TP |
| 458 | .B cpuio |
| 459 | Doesn't transfer any data, but burns CPU cycles according to \fBcpuload\fR and |
| 460 | \fBcpucycles\fR parameters. |
| 461 | .TP |
| 462 | .B guasi |
| 463 | The GUASI I/O engine is the Generic Userspace Asynchronous Syscall Interface |
| 464 | approach to asycnronous I/O. |
| 465 | .br |
| 466 | See <http://www.xmailserver.org/guasi\-lib.html>. |
| 467 | .TP |
| 468 | .B rdma |
| 469 | The RDMA I/O engine supports both RDMA memory semantics (RDMA_WRITE/RDMA_READ) |
| 470 | and channel semantics (Send/Recv) for the InfiniBand, RoCE and iWARP protocols. |
| 471 | .TP |
| 472 | .B external |
| 473 | Loads an external I/O engine object file. Append the engine filename as |
| 474 | `:\fIenginepath\fR'. |
| 475 | .RE |
| 476 | .RE |
| 477 | .TP |
| 478 | .BI iodepth \fR=\fPint |
| 479 | Number of I/O units to keep in flight against the file. Note that increasing |
| 480 | iodepth beyond 1 will not affect synchronous ioengines (except for small |
| 481 | degress when verify_async is in use). Even async engines my impose OS |
| 482 | restrictions causing the desired depth not to be achieved. This may happen on |
| 483 | Linux when using libaio and not setting \fBdirect\fR=1, since buffered IO is |
| 484 | not async on that OS. Keep an eye on the IO depth distribution in the |
| 485 | fio output to verify that the achieved depth is as expected. Default: 1. |
| 486 | .TP |
| 487 | .BI iodepth_batch \fR=\fPint |
| 488 | Number of I/Os to submit at once. Default: \fBiodepth\fR. |
| 489 | .TP |
| 490 | .BI iodepth_batch_complete \fR=\fPint |
| 491 | This defines how many pieces of IO to retrieve at once. It defaults to 1 which |
| 492 | means that we'll ask for a minimum of 1 IO in the retrieval process from the |
| 493 | kernel. The IO retrieval will go on until we hit the limit set by |
| 494 | \fBiodepth_low\fR. If this variable is set to 0, then fio will always check for |
| 495 | completed events before queuing more IO. This helps reduce IO latency, at the |
| 496 | cost of more retrieval system calls. |
| 497 | .TP |
| 498 | .BI iodepth_low \fR=\fPint |
| 499 | Low watermark indicating when to start filling the queue again. Default: |
| 500 | \fBiodepth\fR. |
| 501 | .TP |
| 502 | .BI direct \fR=\fPbool |
| 503 | If true, use non-buffered I/O (usually O_DIRECT). Default: false. |
| 504 | .TP |
| 505 | .BI buffered \fR=\fPbool |
| 506 | If true, use buffered I/O. This is the opposite of the \fBdirect\fR parameter. |
| 507 | Default: true. |
| 508 | .TP |
| 509 | .BI offset \fR=\fPint |
| 510 | Offset in the file to start I/O. Data before the offset will not be touched. |
| 511 | .TP |
| 512 | .BI offset_increment \fR=\fPint |
| 513 | If this is provided, then the real offset becomes the |
| 514 | offset + offset_increment * thread_number, where the thread number is a counter |
| 515 | that starts at 0 and is incremented for each job. This option is useful if |
| 516 | there are several jobs which are intended to operate on a file in parallel in |
| 517 | disjoint segments, with even spacing between the starting points. |
| 518 | .TP |
| 519 | .BI fsync \fR=\fPint |
| 520 | How many I/Os to perform before issuing an \fBfsync\fR\|(2) of dirty data. If |
| 521 | 0, don't sync. Default: 0. |
| 522 | .TP |
| 523 | .BI fdatasync \fR=\fPint |
| 524 | Like \fBfsync\fR, but uses \fBfdatasync\fR\|(2) instead to only sync the |
| 525 | data parts of the file. Default: 0. |
| 526 | .TP |
| 527 | .BI sync_file_range \fR=\fPstr:int |
| 528 | Use sync_file_range() for every \fRval\fP number of write operations. Fio will |
| 529 | track range of writes that have happened since the last sync_file_range() call. |
| 530 | \fRstr\fP can currently be one or more of: |
| 531 | .RS |
| 532 | .TP |
| 533 | .B wait_before |
| 534 | SYNC_FILE_RANGE_WAIT_BEFORE |
| 535 | .TP |
| 536 | .B write |
| 537 | SYNC_FILE_RANGE_WRITE |
| 538 | .TP |
| 539 | .B wait_after |
| 540 | SYNC_FILE_RANGE_WRITE |
| 541 | .TP |
| 542 | .RE |
| 543 | .P |
| 544 | So if you do sync_file_range=wait_before,write:8, fio would use |
| 545 | \fBSYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE\fP for every 8 writes. |
| 546 | Also see the sync_file_range(2) man page. This option is Linux specific. |
| 547 | .TP |
| 548 | .BI overwrite \fR=\fPbool |
| 549 | If writing, setup the file first and do overwrites. Default: false. |
| 550 | .TP |
| 551 | .BI end_fsync \fR=\fPbool |
| 552 | Sync file contents when job exits. Default: false. |
| 553 | .TP |
| 554 | .BI fsync_on_close \fR=\fPbool |
| 555 | If true, sync file contents on close. This differs from \fBend_fsync\fR in that |
| 556 | it will happen on every close, not just at the end of the job. Default: false. |
| 557 | .TP |
| 558 | .BI rwmixcycle \fR=\fPint |
| 559 | How many milliseconds before switching between reads and writes for a mixed |
| 560 | workload. Default: 500ms. |
| 561 | .TP |
| 562 | .BI rwmixread \fR=\fPint |
| 563 | Percentage of a mixed workload that should be reads. Default: 50. |
| 564 | .TP |
| 565 | .BI rwmixwrite \fR=\fPint |
| 566 | Percentage of a mixed workload that should be writes. If \fBrwmixread\fR and |
| 567 | \fBrwmixwrite\fR are given and do not sum to 100%, the latter of the two |
| 568 | overrides the first. This may interfere with a given rate setting, if fio is |
| 569 | asked to limit reads or writes to a certain rate. If that is the case, then |
| 570 | the distribution may be skewed. Default: 50. |
| 571 | .TP |
| 572 | .B norandommap |
| 573 | Normally \fBfio\fR will cover every block of the file when doing random I/O. If |
| 574 | this parameter is given, a new offset will be chosen without looking at past |
| 575 | I/O history. This parameter is mutually exclusive with \fBverify\fR. |
| 576 | .TP |
| 577 | .BI softrandommap \fR=\fPbool |
| 578 | See \fBnorandommap\fR. If fio runs with the random block map enabled and it |
| 579 | fails to allocate the map, if this option is set it will continue without a |
| 580 | random block map. As coverage will not be as complete as with random maps, this |
| 581 | option is disabled by default. |
| 582 | .TP |
| 583 | .BI nice \fR=\fPint |
| 584 | Run job with given nice value. See \fInice\fR\|(2). |
| 585 | .TP |
| 586 | .BI prio \fR=\fPint |
| 587 | Set I/O priority value of this job between 0 (highest) and 7 (lowest). See |
| 588 | \fIionice\fR\|(1). |
| 589 | .TP |
| 590 | .BI prioclass \fR=\fPint |
| 591 | Set I/O priority class. See \fIionice\fR\|(1). |
| 592 | .TP |
| 593 | .BI thinktime \fR=\fPint |
| 594 | Stall job for given number of microseconds between issuing I/Os. |
| 595 | .TP |
| 596 | .BI thinktime_spin \fR=\fPint |
| 597 | Pretend to spend CPU time for given number of microseconds, sleeping the rest |
| 598 | of the time specified by \fBthinktime\fR. Only valid if \fBthinktime\fR is set. |
| 599 | .TP |
| 600 | .BI thinktime_blocks \fR=\fPint |
| 601 | Number of blocks to issue before waiting \fBthinktime\fR microseconds. |
| 602 | Default: 1. |
| 603 | .TP |
| 604 | .BI rate \fR=\fPint |
| 605 | Cap bandwidth used by this job. The number is in bytes/sec, the normal postfix |
| 606 | rules apply. You can use \fBrate\fR=500k to limit reads and writes to 500k each, |
| 607 | or you can specify read and writes separately. Using \fBrate\fR=1m,500k would |
| 608 | limit reads to 1MB/sec and writes to 500KB/sec. Capping only reads or writes |
| 609 | can be done with \fBrate\fR=,500k or \fBrate\fR=500k,. The former will only |
| 610 | limit writes (to 500KB/sec), the latter will only limit reads. |
| 611 | .TP |
| 612 | .BI ratemin \fR=\fPint |
| 613 | Tell \fBfio\fR to do whatever it can to maintain at least the given bandwidth. |
| 614 | Failing to meet this requirement will cause the job to exit. The same format |
| 615 | as \fBrate\fR is used for read vs write separation. |
| 616 | .TP |
| 617 | .BI rate_iops \fR=\fPint |
| 618 | Cap the bandwidth to this number of IOPS. Basically the same as rate, just |
| 619 | specified independently of bandwidth. The same format as \fBrate\fR is used for |
| 620 | read vs write seperation. If \fBblocksize\fR is a range, the smallest block |
| 621 | size is used as the metric. |
| 622 | .TP |
| 623 | .BI rate_iops_min \fR=\fPint |
| 624 | If this rate of I/O is not met, the job will exit. The same format as \fBrate\fR |
| 625 | is used for read vs write seperation. |
| 626 | .TP |
| 627 | .BI ratecycle \fR=\fPint |
| 628 | Average bandwidth for \fBrate\fR and \fBratemin\fR over this number of |
| 629 | milliseconds. Default: 1000ms. |
| 630 | .TP |
| 631 | .BI cpumask \fR=\fPint |
| 632 | Set CPU affinity for this job. \fIint\fR is a bitmask of allowed CPUs the job |
| 633 | may run on. See \fBsched_setaffinity\fR\|(2). |
| 634 | .TP |
| 635 | .BI cpus_allowed \fR=\fPstr |
| 636 | Same as \fBcpumask\fR, but allows a comma-delimited list of CPU numbers. |
| 637 | .TP |
| 638 | .BI startdelay \fR=\fPint |
| 639 | Delay start of job for the specified number of seconds. |
| 640 | .TP |
| 641 | .BI runtime \fR=\fPint |
| 642 | Terminate processing after the specified number of seconds. |
| 643 | .TP |
| 644 | .B time_based |
| 645 | If given, run for the specified \fBruntime\fR duration even if the files are |
| 646 | completely read or written. The same workload will be repeated as many times |
| 647 | as \fBruntime\fR allows. |
| 648 | .TP |
| 649 | .BI ramp_time \fR=\fPint |
| 650 | If set, fio will run the specified workload for this amount of time before |
| 651 | logging any performance numbers. Useful for letting performance settle before |
| 652 | logging results, thus minimizing the runtime required for stable results. Note |
| 653 | that the \fBramp_time\fR is considered lead in time for a job, thus it will |
| 654 | increase the total runtime if a special timeout or runtime is specified. |
| 655 | .TP |
| 656 | .BI invalidate \fR=\fPbool |
| 657 | Invalidate buffer-cache for the file prior to starting I/O. Default: true. |
| 658 | .TP |
| 659 | .BI sync \fR=\fPbool |
| 660 | Use synchronous I/O for buffered writes. For the majority of I/O engines, |
| 661 | this means using O_SYNC. Default: false. |
| 662 | .TP |
| 663 | .BI iomem \fR=\fPstr "\fR,\fP mem" \fR=\fPstr |
| 664 | Allocation method for I/O unit buffer. Allowed values are: |
| 665 | .RS |
| 666 | .RS |
| 667 | .TP |
| 668 | .B malloc |
| 669 | Allocate memory with \fImalloc\fR\|(3). |
| 670 | .TP |
| 671 | .B shm |
| 672 | Use shared memory buffers allocated through \fIshmget\fR\|(2). |
| 673 | .TP |
| 674 | .B shmhuge |
| 675 | Same as \fBshm\fR, but use huge pages as backing. |
| 676 | .TP |
| 677 | .B mmap |
| 678 | Use \fImmap\fR\|(2) for allocation. Uses anonymous memory unless a filename |
| 679 | is given after the option in the format `:\fIfile\fR'. |
| 680 | .TP |
| 681 | .B mmaphuge |
| 682 | Same as \fBmmap\fR, but use huge files as backing. |
| 683 | .RE |
| 684 | .P |
| 685 | The amount of memory allocated is the maximum allowed \fBblocksize\fR for the |
| 686 | job multiplied by \fBiodepth\fR. For \fBshmhuge\fR or \fBmmaphuge\fR to work, |
| 687 | the system must have free huge pages allocated. \fBmmaphuge\fR also needs to |
| 688 | have hugetlbfs mounted, and \fIfile\fR must point there. At least on Linux, |
| 689 | huge pages must be manually allocated. See \fB/proc/sys/vm/nr_hugehages\fR |
| 690 | and the documentation for that. Normally you just need to echo an appropriate |
| 691 | number, eg echoing 8 will ensure that the OS has 8 huge pages ready for |
| 692 | use. |
| 693 | .RE |
| 694 | .TP |
| 695 | .BI iomem_align \fR=\fPint "\fR,\fP mem_align" \fR=\fPint |
| 696 | This indiciates the memory alignment of the IO memory buffers. Note that the |
| 697 | given alignment is applied to the first IO unit buffer, if using \fBiodepth\fR |
| 698 | the alignment of the following buffers are given by the \fBbs\fR used. In |
| 699 | other words, if using a \fBbs\fR that is a multiple of the page sized in the |
| 700 | system, all buffers will be aligned to this value. If using a \fBbs\fR that |
| 701 | is not page aligned, the alignment of subsequent IO memory buffers is the |
| 702 | sum of the \fBiomem_align\fR and \fBbs\fR used. |
| 703 | .TP |
| 704 | .BI hugepage\-size \fR=\fPint |
| 705 | Defines the size of a huge page. Must be at least equal to the system setting. |
| 706 | Should be a multiple of 1MB. Default: 4MB. |
| 707 | .TP |
| 708 | .B exitall |
| 709 | Terminate all jobs when one finishes. Default: wait for each job to finish. |
| 710 | .TP |
| 711 | .BI bwavgtime \fR=\fPint |
| 712 | Average bandwidth calculations over the given time in milliseconds. Default: |
| 713 | 500ms. |
| 714 | .TP |
| 715 | .BI iopsavgtime \fR=\fPint |
| 716 | Average IOPS calculations over the given time in milliseconds. Default: |
| 717 | 500ms. |
| 718 | .TP |
| 719 | .BI create_serialize \fR=\fPbool |
| 720 | If true, serialize file creation for the jobs. Default: true. |
| 721 | .TP |
| 722 | .BI create_fsync \fR=\fPbool |
| 723 | \fIfsync\fR\|(2) data file after creation. Default: true. |
| 724 | .TP |
| 725 | .BI create_on_open \fR=\fPbool |
| 726 | If true, the files are not created until they are opened for IO by the job. |
| 727 | .TP |
| 728 | .BI pre_read \fR=\fPbool |
| 729 | If this is given, files will be pre-read into memory before starting the given |
| 730 | IO operation. This will also clear the \fR \fBinvalidate\fR flag, since it is |
| 731 | pointless to pre-read and then drop the cache. This will only work for IO |
| 732 | engines that are seekable, since they allow you to read the same data |
| 733 | multiple times. Thus it will not work on eg network or splice IO. |
| 734 | .TP |
| 735 | .BI unlink \fR=\fPbool |
| 736 | Unlink job files when done. Default: false. |
| 737 | .TP |
| 738 | .BI loops \fR=\fPint |
| 739 | Specifies the number of iterations (runs of the same workload) of this job. |
| 740 | Default: 1. |
| 741 | .TP |
| 742 | .BI do_verify \fR=\fPbool |
| 743 | Run the verify phase after a write phase. Only valid if \fBverify\fR is set. |
| 744 | Default: true. |
| 745 | .TP |
| 746 | .BI verify \fR=\fPstr |
| 747 | Method of verifying file contents after each iteration of the job. Allowed |
| 748 | values are: |
| 749 | .RS |
| 750 | .RS |
| 751 | .TP |
| 752 | .B md5 crc16 crc32 crc32c crc32c-intel crc64 crc7 sha256 sha512 sha1 |
| 753 | Store appropriate checksum in the header of each block. crc32c-intel is |
| 754 | hardware accelerated SSE4.2 driven, falls back to regular crc32c if |
| 755 | not supported by the system. |
| 756 | .TP |
| 757 | .B meta |
| 758 | Write extra information about each I/O (timestamp, block number, etc.). The |
| 759 | block number is verified. See \fBverify_pattern\fR as well. |
| 760 | .TP |
| 761 | .B null |
| 762 | Pretend to verify. Used for testing internals. |
| 763 | .RE |
| 764 | |
| 765 | This option can be used for repeated burn-in tests of a system to make sure |
| 766 | that the written data is also correctly read back. If the data direction given |
| 767 | is a read or random read, fio will assume that it should verify a previously |
| 768 | written file. If the data direction includes any form of write, the verify will |
| 769 | be of the newly written data. |
| 770 | .RE |
| 771 | .TP |
| 772 | .BI verify_sort \fR=\fPbool |
| 773 | If true, written verify blocks are sorted if \fBfio\fR deems it to be faster to |
| 774 | read them back in a sorted manner. Default: true. |
| 775 | .TP |
| 776 | .BI verify_offset \fR=\fPint |
| 777 | Swap the verification header with data somewhere else in the block before |
| 778 | writing. It is swapped back before verifying. |
| 779 | .TP |
| 780 | .BI verify_interval \fR=\fPint |
| 781 | Write the verification header for this number of bytes, which should divide |
| 782 | \fBblocksize\fR. Default: \fBblocksize\fR. |
| 783 | .TP |
| 784 | .BI verify_pattern \fR=\fPstr |
| 785 | If set, fio will fill the io buffers with this pattern. Fio defaults to filling |
| 786 | with totally random bytes, but sometimes it's interesting to fill with a known |
| 787 | pattern for io verification purposes. Depending on the width of the pattern, |
| 788 | fio will fill 1/2/3/4 bytes of the buffer at the time(it can be either a |
| 789 | decimal or a hex number). The verify_pattern if larger than a 32-bit quantity |
| 790 | has to be a hex number that starts with either "0x" or "0X". Use with |
| 791 | \fBverify\fP=meta. |
| 792 | .TP |
| 793 | .BI verify_fatal \fR=\fPbool |
| 794 | If true, exit the job on the first observed verification failure. Default: |
| 795 | false. |
| 796 | .TP |
| 797 | .BI verify_dump \fR=\fPbool |
| 798 | If set, dump the contents of both the original data block and the data block we |
| 799 | read off disk to files. This allows later analysis to inspect just what kind of |
| 800 | data corruption occurred. Off by default. |
| 801 | .TP |
| 802 | .BI verify_async \fR=\fPint |
| 803 | Fio will normally verify IO inline from the submitting thread. This option |
| 804 | takes an integer describing how many async offload threads to create for IO |
| 805 | verification instead, causing fio to offload the duty of verifying IO contents |
| 806 | to one or more separate threads. If using this offload option, even sync IO |
| 807 | engines can benefit from using an \fBiodepth\fR setting higher than 1, as it |
| 808 | allows them to have IO in flight while verifies are running. |
| 809 | .TP |
| 810 | .BI verify_async_cpus \fR=\fPstr |
| 811 | Tell fio to set the given CPU affinity on the async IO verification threads. |
| 812 | See \fBcpus_allowed\fP for the format used. |
| 813 | .TP |
| 814 | .BI verify_backlog \fR=\fPint |
| 815 | Fio will normally verify the written contents of a job that utilizes verify |
| 816 | once that job has completed. In other words, everything is written then |
| 817 | everything is read back and verified. You may want to verify continually |
| 818 | instead for a variety of reasons. Fio stores the meta data associated with an |
| 819 | IO block in memory, so for large verify workloads, quite a bit of memory would |
| 820 | be used up holding this meta data. If this option is enabled, fio will write |
| 821 | only N blocks before verifying these blocks. |
| 822 | .TP |
| 823 | .BI verify_backlog_batch \fR=\fPint |
| 824 | Control how many blocks fio will verify if verify_backlog is set. If not set, |
| 825 | will default to the value of \fBverify_backlog\fR (meaning the entire queue is |
| 826 | read back and verified). If \fBverify_backlog_batch\fR is less than |
| 827 | \fBverify_backlog\fR then not all blocks will be verified, if |
| 828 | \fBverify_backlog_batch\fR is larger than \fBverify_backlog\fR, some blocks |
| 829 | will be verified more than once. |
| 830 | .TP |
| 831 | .B stonewall "\fR,\fP wait_for_previous" |
| 832 | Wait for preceding jobs in the job file to exit before starting this one. |
| 833 | \fBstonewall\fR implies \fBnew_group\fR. |
| 834 | .TP |
| 835 | .B new_group |
| 836 | Start a new reporting group. If not given, all jobs in a file will be part |
| 837 | of the same reporting group, unless separated by a stonewall. |
| 838 | .TP |
| 839 | .BI numjobs \fR=\fPint |
| 840 | Number of clones (processes/threads performing the same workload) of this job. |
| 841 | Default: 1. |
| 842 | .TP |
| 843 | .B group_reporting |
| 844 | If set, display per-group reports instead of per-job when \fBnumjobs\fR is |
| 845 | specified. |
| 846 | .TP |
| 847 | .B thread |
| 848 | Use threads created with \fBpthread_create\fR\|(3) instead of processes created |
| 849 | with \fBfork\fR\|(2). |
| 850 | .TP |
| 851 | .BI zonesize \fR=\fPint |
| 852 | Divide file into zones of the specified size in bytes. See \fBzoneskip\fR. |
| 853 | .TP |
| 854 | .BI zoneskip \fR=\fPint |
| 855 | Skip the specified number of bytes when \fBzonesize\fR bytes of data have been |
| 856 | read. |
| 857 | .TP |
| 858 | .BI write_iolog \fR=\fPstr |
| 859 | Write the issued I/O patterns to the specified file. Specify a separate file |
| 860 | for each job, otherwise the iologs will be interspersed and the file may be |
| 861 | corrupt. |
| 862 | .TP |
| 863 | .BI read_iolog \fR=\fPstr |
| 864 | Replay the I/O patterns contained in the specified file generated by |
| 865 | \fBwrite_iolog\fR, or may be a \fBblktrace\fR binary file. |
| 866 | .TP |
| 867 | .BI replay_no_stall \fR=\fPint |
| 868 | While replaying I/O patterns using \fBread_iolog\fR the default behavior |
| 869 | attempts to respect timing information between I/Os. Enabling |
| 870 | \fBreplay_no_stall\fR causes I/Os to be replayed as fast as possible while |
| 871 | still respecting ordering. |
| 872 | .TP |
| 873 | .BI replay_redirect \fR=\fPstr |
| 874 | While replaying I/O patterns using \fBread_iolog\fR the default behavior |
| 875 | is to replay the IOPS onto the major/minor device that each IOP was recorded |
| 876 | from. Setting \fBreplay_redirect\fR causes all IOPS to be replayed onto the |
| 877 | single specified device regardless of the device it was recorded from. |
| 878 | .TP |
| 879 | .BI write_bw_log \fR=\fPstr |
| 880 | If given, write a bandwidth log of the jobs in this job file. Can be used to |
| 881 | store data of the bandwidth of the jobs in their lifetime. The included |
| 882 | fio_generate_plots script uses gnuplot to turn these text files into nice |
| 883 | graphs. See \fBwrite_log_log\fR for behaviour of given filename. For this |
| 884 | option, the postfix is _bw.log. |
| 885 | .TP |
| 886 | .BI write_lat_log \fR=\fPstr |
| 887 | Same as \fBwrite_bw_log\fR, but writes I/O completion latencies. If no |
| 888 | filename is given with this option, the default filename of "jobname_type.log" |
| 889 | is used. Even if the filename is given, fio will still append the type of log. |
| 890 | .TP |
| 891 | .BI write_iops_log \fR=\fPstr |
| 892 | Same as \fBwrite_bw_log\fR, but writes IOPS. If no filename is given with this |
| 893 | option, the default filename of "jobname_type.log" is used. Even if the |
| 894 | filename is given, fio will still append the type of log. |
| 895 | .TP |
| 896 | .BI log_avg_msec \fR=\fPint |
| 897 | By default, fio will log an entry in the iops, latency, or bw log for every |
| 898 | IO that completes. When writing to the disk log, that can quickly grow to a |
| 899 | very large size. Setting this option makes fio average the each log entry |
| 900 | over the specified period of time, reducing the resolution of the log. |
| 901 | Defaults to 0. |
| 902 | .TP |
| 903 | .BI disable_lat \fR=\fPbool |
| 904 | Disable measurements of total latency numbers. Useful only for cutting |
| 905 | back the number of calls to gettimeofday, as that does impact performance at |
| 906 | really high IOPS rates. Note that to really get rid of a large amount of these |
| 907 | calls, this option must be used with disable_slat and disable_bw as well. |
| 908 | .TP |
| 909 | .BI disable_clat \fR=\fPbool |
| 910 | Disable measurements of completion latency numbers. See \fBdisable_lat\fR. |
| 911 | .TP |
| 912 | .BI disable_slat \fR=\fPbool |
| 913 | Disable measurements of submission latency numbers. See \fBdisable_lat\fR. |
| 914 | .TP |
| 915 | .BI disable_bw_measurement \fR=\fPbool |
| 916 | Disable measurements of throughput/bandwidth numbers. See \fBdisable_lat\fR. |
| 917 | .TP |
| 918 | .BI lockmem \fR=\fPint |
| 919 | Pin the specified amount of memory with \fBmlock\fR\|(2). Can be used to |
| 920 | simulate a smaller amount of memory. |
| 921 | .TP |
| 922 | .BI exec_prerun \fR=\fPstr |
| 923 | Before running the job, execute the specified command with \fBsystem\fR\|(3). |
| 924 | .TP |
| 925 | .BI exec_postrun \fR=\fPstr |
| 926 | Same as \fBexec_prerun\fR, but the command is executed after the job completes. |
| 927 | .TP |
| 928 | .BI ioscheduler \fR=\fPstr |
| 929 | Attempt to switch the device hosting the file to the specified I/O scheduler. |
| 930 | .TP |
| 931 | .BI cpuload \fR=\fPint |
| 932 | If the job is a CPU cycle-eater, attempt to use the specified percentage of |
| 933 | CPU cycles. |
| 934 | .TP |
| 935 | .BI cpuchunks \fR=\fPint |
| 936 | If the job is a CPU cycle-eater, split the load into cycles of the |
| 937 | given time in milliseconds. |
| 938 | .TP |
| 939 | .BI disk_util \fR=\fPbool |
| 940 | Generate disk utilization statistics if the platform supports it. Default: true. |
| 941 | .TP |
| 942 | .BI gtod_reduce \fR=\fPbool |
| 943 | Enable all of the gettimeofday() reducing options (disable_clat, disable_slat, |
| 944 | disable_bw) plus reduce precision of the timeout somewhat to really shrink the |
| 945 | gettimeofday() call count. With this option enabled, we only do about 0.4% of |
| 946 | the gtod() calls we would have done if all time keeping was enabled. |
| 947 | .TP |
| 948 | .BI gtod_cpu \fR=\fPint |
| 949 | Sometimes it's cheaper to dedicate a single thread of execution to just getting |
| 950 | the current time. Fio (and databases, for instance) are very intensive on |
| 951 | gettimeofday() calls. With this option, you can set one CPU aside for doing |
| 952 | nothing but logging current time to a shared memory location. Then the other |
| 953 | threads/processes that run IO workloads need only copy that segment, instead of |
| 954 | entering the kernel with a gettimeofday() call. The CPU set aside for doing |
| 955 | these time calls will be excluded from other uses. Fio will manually clear it |
| 956 | from the CPU mask of other jobs. |
| 957 | .TP |
| 958 | .BI cgroup \fR=\fPstr |
| 959 | Add job to this control group. If it doesn't exist, it will be created. |
| 960 | The system must have a mounted cgroup blkio mount point for this to work. If |
| 961 | your system doesn't have it mounted, you can do so with: |
| 962 | |
| 963 | # mount \-t cgroup \-o blkio none /cgroup |
| 964 | .TP |
| 965 | .BI cgroup_weight \fR=\fPint |
| 966 | Set the weight of the cgroup to this value. See the documentation that comes |
| 967 | with the kernel, allowed values are in the range of 100..1000. |
| 968 | .TP |
| 969 | .BI cgroup_nodelete \fR=\fPbool |
| 970 | Normally fio will delete the cgroups it has created after the job completion. |
| 971 | To override this behavior and to leave cgroups around after the job completion, |
| 972 | set cgroup_nodelete=1. This can be useful if one wants to inspect various |
| 973 | cgroup files after job completion. Default: false |
| 974 | .TP |
| 975 | .BI uid \fR=\fPint |
| 976 | Instead of running as the invoking user, set the user ID to this value before |
| 977 | the thread/process does any work. |
| 978 | .TP |
| 979 | .BI gid \fR=\fPint |
| 980 | Set group ID, see \fBuid\fR. |
| 981 | .TP |
| 982 | .BI flow_id \fR=\fPint |
| 983 | The ID of the flow. If not specified, it defaults to being a global flow. See |
| 984 | \fBflow\fR. |
| 985 | .TP |
| 986 | .BI flow \fR=\fPint |
| 987 | Weight in token-based flow control. If this value is used, then there is a |
| 988 | \fBflow counter\fR which is used to regulate the proportion of activity between |
| 989 | two or more jobs. fio attempts to keep this flow counter near zero. The |
| 990 | \fBflow\fR parameter stands for how much should be added or subtracted to the |
| 991 | flow counter on each iteration of the main I/O loop. That is, if one job has |
| 992 | \fBflow=8\fR and another job has \fBflow=-1\fR, then there will be a roughly |
| 993 | 1:8 ratio in how much one runs vs the other. |
| 994 | .TP |
| 995 | .BI flow_watermark \fR=\fPint |
| 996 | The maximum value that the absolute value of the flow counter is allowed to |
| 997 | reach before the job must wait for a lower value of the counter. |
| 998 | .TP |
| 999 | .BI flow_sleep \fR=\fPint |
| 1000 | The period of time, in microseconds, to wait after the flow watermark has been |
| 1001 | exceeded before retrying operations |
| 1002 | .TP |
| 1003 | .BI clat_percentiles \fR=\fPbool |
| 1004 | Enable the reporting of percentiles of completion latencies. |
| 1005 | .TP |
| 1006 | .BI percentile_list \fR=\fPfloat_list |
| 1007 | Overwrite the default list of percentiles for completion |
| 1008 | latencies. Each number is a floating number in the range (0,100], and |
| 1009 | the maximum length of the list is 20. Use ':' to separate the |
| 1010 | numbers. For example, \-\-percentile_list=99.5:99.9 will cause fio to |
| 1011 | report the values of completion latency below which 99.5% and 99.9% of |
| 1012 | the observed latencies fell, respectively. |
| 1013 | .SS "Ioengine Parameters List" |
| 1014 | Some parameters are only valid when a specific ioengine is in use. These are |
| 1015 | used identically to normal parameters, with the caveat that when used on the |
| 1016 | command line, the must come after the ioengine that defines them is selected. |
| 1017 | .TP |
| 1018 | .BI (cpu)cpuload \fR=\fPint |
| 1019 | Attempt to use the specified percentage of CPU cycles. |
| 1020 | .TP |
| 1021 | .BI (cpu)cpuchunks \fR=\fPint |
| 1022 | Split the load into cycles of the given time. In microseconds. |
| 1023 | .TP |
| 1024 | .BI (libaio)userspace_reap |
| 1025 | Normally, with the libaio engine in use, fio will use |
| 1026 | the io_getevents system call to reap newly returned events. |
| 1027 | With this flag turned on, the AIO ring will be read directly |
| 1028 | from user-space to reap events. The reaping mode is only |
| 1029 | enabled when polling for a minimum of 0 events (eg when |
| 1030 | iodepth_batch_complete=0). |
| 1031 | .TP |
| 1032 | .BI (net,netsplice)hostname \fR=\fPstr |
| 1033 | The host name or IP address to use for TCP or UDP based IO. |
| 1034 | If the job is a TCP listener or UDP reader, the hostname is not |
| 1035 | used and must be omitted. |
| 1036 | .TP |
| 1037 | .BI (net,netsplice)port \fR=\fPint |
| 1038 | The TCP or UDP port to bind to or connect to. |
| 1039 | .TP |
| 1040 | .BI (net,netsplice)protocol \fR=\fPstr "\fR,\fP proto" \fR=\fPstr |
| 1041 | The network protocol to use. Accepted values are: |
| 1042 | .RS |
| 1043 | .RS |
| 1044 | .TP |
| 1045 | .B tcp |
| 1046 | Transmission control protocol |
| 1047 | .TP |
| 1048 | .B udp |
| 1049 | Unreliable datagram protocol |
| 1050 | .TP |
| 1051 | .B unix |
| 1052 | UNIX domain socket |
| 1053 | .RE |
| 1054 | .P |
| 1055 | When the protocol is TCP or UDP, the port must also be given, |
| 1056 | as well as the hostname if the job is a TCP listener or UDP |
| 1057 | reader. For unix sockets, the normal filename option should be |
| 1058 | used and the port is invalid. |
| 1059 | .RE |
| 1060 | .TP |
| 1061 | .BI (net,netsplice)listen |
| 1062 | For TCP network connections, tell fio to listen for incoming |
| 1063 | connections rather than initiating an outgoing connection. The |
| 1064 | hostname must be omitted if this option is used. |
| 1065 | .SH OUTPUT |
| 1066 | While running, \fBfio\fR will display the status of the created jobs. For |
| 1067 | example: |
| 1068 | .RS |
| 1069 | .P |
| 1070 | Threads: 1: [_r] [24.8% done] [ 13509/ 8334 kb/s] [eta 00h:01m:31s] |
| 1071 | .RE |
| 1072 | .P |
| 1073 | The characters in the first set of brackets denote the current status of each |
| 1074 | threads. The possible values are: |
| 1075 | .P |
| 1076 | .PD 0 |
| 1077 | .RS |
| 1078 | .TP |
| 1079 | .B P |
| 1080 | Setup but not started. |
| 1081 | .TP |
| 1082 | .B C |
| 1083 | Thread created. |
| 1084 | .TP |
| 1085 | .B I |
| 1086 | Initialized, waiting. |
| 1087 | .TP |
| 1088 | .B R |
| 1089 | Running, doing sequential reads. |
| 1090 | .TP |
| 1091 | .B r |
| 1092 | Running, doing random reads. |
| 1093 | .TP |
| 1094 | .B W |
| 1095 | Running, doing sequential writes. |
| 1096 | .TP |
| 1097 | .B w |
| 1098 | Running, doing random writes. |
| 1099 | .TP |
| 1100 | .B M |
| 1101 | Running, doing mixed sequential reads/writes. |
| 1102 | .TP |
| 1103 | .B m |
| 1104 | Running, doing mixed random reads/writes. |
| 1105 | .TP |
| 1106 | .B F |
| 1107 | Running, currently waiting for \fBfsync\fR\|(2). |
| 1108 | .TP |
| 1109 | .B V |
| 1110 | Running, verifying written data. |
| 1111 | .TP |
| 1112 | .B E |
| 1113 | Exited, not reaped by main thread. |
| 1114 | .TP |
| 1115 | .B \- |
| 1116 | Exited, thread reaped. |
| 1117 | .RE |
| 1118 | .PD |
| 1119 | .P |
| 1120 | The second set of brackets shows the estimated completion percentage of |
| 1121 | the current group. The third set shows the read and write I/O rate, |
| 1122 | respectively. Finally, the estimated run time of the job is displayed. |
| 1123 | .P |
| 1124 | When \fBfio\fR completes (or is interrupted by Ctrl-C), it will show data |
| 1125 | for each thread, each group of threads, and each disk, in that order. |
| 1126 | .P |
| 1127 | Per-thread statistics first show the threads client number, group-id, and |
| 1128 | error code. The remaining figures are as follows: |
| 1129 | .RS |
| 1130 | .TP |
| 1131 | .B io |
| 1132 | Number of megabytes of I/O performed. |
| 1133 | .TP |
| 1134 | .B bw |
| 1135 | Average data rate (bandwidth). |
| 1136 | .TP |
| 1137 | .B runt |
| 1138 | Threads run time. |
| 1139 | .TP |
| 1140 | .B slat |
| 1141 | Submission latency minimum, maximum, average and standard deviation. This is |
| 1142 | the time it took to submit the I/O. |
| 1143 | .TP |
| 1144 | .B clat |
| 1145 | Completion latency minimum, maximum, average and standard deviation. This |
| 1146 | is the time between submission and completion. |
| 1147 | .TP |
| 1148 | .B bw |
| 1149 | Bandwidth minimum, maximum, percentage of aggregate bandwidth received, average |
| 1150 | and standard deviation. |
| 1151 | .TP |
| 1152 | .B cpu |
| 1153 | CPU usage statistics. Includes user and system time, number of context switches |
| 1154 | this thread went through and number of major and minor page faults. |
| 1155 | .TP |
| 1156 | .B IO depths |
| 1157 | Distribution of I/O depths. Each depth includes everything less than (or equal) |
| 1158 | to it, but greater than the previous depth. |
| 1159 | .TP |
| 1160 | .B IO issued |
| 1161 | Number of read/write requests issued, and number of short read/write requests. |
| 1162 | .TP |
| 1163 | .B IO latencies |
| 1164 | Distribution of I/O completion latencies. The numbers follow the same pattern |
| 1165 | as \fBIO depths\fR. |
| 1166 | .RE |
| 1167 | .P |
| 1168 | The group statistics show: |
| 1169 | .PD 0 |
| 1170 | .RS |
| 1171 | .TP |
| 1172 | .B io |
| 1173 | Number of megabytes I/O performed. |
| 1174 | .TP |
| 1175 | .B aggrb |
| 1176 | Aggregate bandwidth of threads in the group. |
| 1177 | .TP |
| 1178 | .B minb |
| 1179 | Minimum average bandwidth a thread saw. |
| 1180 | .TP |
| 1181 | .B maxb |
| 1182 | Maximum average bandwidth a thread saw. |
| 1183 | .TP |
| 1184 | .B mint |
| 1185 | Shortest runtime of threads in the group. |
| 1186 | .TP |
| 1187 | .B maxt |
| 1188 | Longest runtime of threads in the group. |
| 1189 | .RE |
| 1190 | .PD |
| 1191 | .P |
| 1192 | Finally, disk statistics are printed with reads first: |
| 1193 | .PD 0 |
| 1194 | .RS |
| 1195 | .TP |
| 1196 | .B ios |
| 1197 | Number of I/Os performed by all groups. |
| 1198 | .TP |
| 1199 | .B merge |
| 1200 | Number of merges in the I/O scheduler. |
| 1201 | .TP |
| 1202 | .B ticks |
| 1203 | Number of ticks we kept the disk busy. |
| 1204 | .TP |
| 1205 | .B io_queue |
| 1206 | Total time spent in the disk queue. |
| 1207 | .TP |
| 1208 | .B util |
| 1209 | Disk utilization. |
| 1210 | .RE |
| 1211 | .PD |
| 1212 | .SH TERSE OUTPUT |
| 1213 | If the \fB\-\-minimal\fR option is given, the results will be printed in a |
| 1214 | semicolon-delimited format suitable for scripted use - a job description |
| 1215 | (if provided) follows on a new line. Note that the first |
| 1216 | number in the line is the version number. If the output has to be changed |
| 1217 | for some reason, this number will be incremented by 1 to signify that |
| 1218 | change. The fields are: |
| 1219 | .P |
| 1220 | .RS |
| 1221 | .B terse version, fio version, jobname, groupid, error |
| 1222 | .P |
| 1223 | Read status: |
| 1224 | .RS |
| 1225 | .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP |
| 1226 | .P |
| 1227 | Submission latency: |
| 1228 | .RS |
| 1229 | .B min, max, mean, standard deviation |
| 1230 | .RE |
| 1231 | Completion latency: |
| 1232 | .RS |
| 1233 | .B min, max, mean, standard deviation |
| 1234 | .RE |
| 1235 | Completion latency percentiles (20 fields): |
| 1236 | .RS |
| 1237 | .B Xth percentile=usec |
| 1238 | .RE |
| 1239 | Total latency: |
| 1240 | .RS |
| 1241 | .B min, max, mean, standard deviation |
| 1242 | .RE |
| 1243 | Bandwidth: |
| 1244 | .RS |
| 1245 | .B min, max, aggregate percentage of total, mean, standard deviation |
| 1246 | .RE |
| 1247 | .RE |
| 1248 | .P |
| 1249 | Write status: |
| 1250 | .RS |
| 1251 | .B Total I/O \fR(KB)\fP, bandwidth \fR(KB/s)\fP, IOPS, runtime \fR(ms)\fP |
| 1252 | .P |
| 1253 | Submission latency: |
| 1254 | .RS |
| 1255 | .B min, max, mean, standard deviation |
| 1256 | .RE |
| 1257 | Completion latency: |
| 1258 | .RS |
| 1259 | .B min, max, mean, standard deviation |
| 1260 | .RE |
| 1261 | Completion latency percentiles (20 fields): |
| 1262 | .RS |
| 1263 | .B Xth percentile=usec |
| 1264 | .RE |
| 1265 | Total latency: |
| 1266 | .RS |
| 1267 | .B min, max, mean, standard deviation |
| 1268 | .RE |
| 1269 | Bandwidth: |
| 1270 | .RS |
| 1271 | .B min, max, aggregate percentage of total, mean, standard deviation |
| 1272 | .RE |
| 1273 | .RE |
| 1274 | .P |
| 1275 | CPU usage: |
| 1276 | .RS |
| 1277 | .B user, system, context switches, major page faults, minor page faults |
| 1278 | .RE |
| 1279 | .P |
| 1280 | IO depth distribution: |
| 1281 | .RS |
| 1282 | .B <=1, 2, 4, 8, 16, 32, >=64 |
| 1283 | .RE |
| 1284 | .P |
| 1285 | IO latency distribution: |
| 1286 | .RS |
| 1287 | Microseconds: |
| 1288 | .RS |
| 1289 | .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000 |
| 1290 | .RE |
| 1291 | Milliseconds: |
| 1292 | .RS |
| 1293 | .B <=2, 4, 10, 20, 50, 100, 250, 500, 750, 1000, 2000, >=2000 |
| 1294 | .RE |
| 1295 | .RE |
| 1296 | .P |
| 1297 | Disk utilization (1 for each disk used): |
| 1298 | .RS |
| 1299 | .B name, read ios, write ios, read merges, write merges, read ticks, write ticks, read in-queue time, write in-queue time, disk utilization percentage |
| 1300 | .RE |
| 1301 | .P |
| 1302 | Error Info (dependent on continue_on_error, default off): |
| 1303 | .RS |
| 1304 | .B total # errors, first error code |
| 1305 | .RE |
| 1306 | .P |
| 1307 | .B text description (if provided in config - appears on newline) |
| 1308 | .RE |
| 1309 | .SH CLIENT / SERVER |
| 1310 | Normally you would run fio as a stand-alone application on the machine |
| 1311 | where the IO workload should be generated. However, it is also possible to |
| 1312 | run the frontend and backend of fio separately. This makes it possible to |
| 1313 | have a fio server running on the machine(s) where the IO workload should |
| 1314 | be running, while controlling it from another machine. |
| 1315 | |
| 1316 | To start the server, you would do: |
| 1317 | |
| 1318 | \fBfio \-\-server=args\fR |
| 1319 | |
| 1320 | on that machine, where args defines what fio listens to. The arguments |
| 1321 | are of the form 'type:hostname or IP:port'. 'type' is either 'ip' (or ip4) |
| 1322 | for TCP/IP v4, 'ip6' for TCP/IP v6, or 'sock' for a local unix domain socket. |
| 1323 | 'hostname' is either a hostname or IP address, and 'port' is the port to |
| 1324 | listen to (only valid for TCP/IP, not a local socket). Some examples: |
| 1325 | |
| 1326 | 1) fio --server |
| 1327 | |
| 1328 | Start a fio server, listening on all interfaces on the default port (8765). |
| 1329 | |
| 1330 | 2) fio --server=ip:hostname,4444 |
| 1331 | |
| 1332 | Start a fio server, listening on IP belonging to hostname and on port 4444. |
| 1333 | |
| 1334 | 3) fio --server=ip6:::1,4444 |
| 1335 | |
| 1336 | Start a fio server, listening on IPv6 localhost ::1 and on port 4444. |
| 1337 | |
| 1338 | 4) fio --server=,4444 |
| 1339 | |
| 1340 | Start a fio server, listening on all interfaces on port 4444. |
| 1341 | |
| 1342 | 5) fio --server=1.2.3.4 |
| 1343 | |
| 1344 | Start a fio server, listening on IP 1.2.3.4 on the default port. |
| 1345 | |
| 1346 | 6) fio --server=sock:/tmp/fio.sock |
| 1347 | |
| 1348 | Start a fio server, listening on the local socket /tmp/fio.sock. |
| 1349 | |
| 1350 | When a server is running, you can connect to it from a client. The client |
| 1351 | is run with: |
| 1352 | |
| 1353 | fio --local-args --client=server --remote-args <job file(s)> |
| 1354 | |
| 1355 | where --local-args are arguments that are local to the client where it is |
| 1356 | running, 'server' is the connect string, and --remote-args and <job file(s)> |
| 1357 | are sent to the server. The 'server' string follows the same format as it |
| 1358 | does on the server side, to allow IP/hostname/socket and port strings. |
| 1359 | You can connect to multiple clients as well, to do that you could run: |
| 1360 | |
| 1361 | fio --client=server2 --client=server2 <job file(s)> |
| 1362 | .SH AUTHORS |
| 1363 | |
| 1364 | .B fio |
| 1365 | was written by Jens Axboe <jens.axboe@oracle.com>, |
| 1366 | now Jens Axboe <jaxboe@fusionio.com>. |
| 1367 | .br |
| 1368 | This man page was written by Aaron Carroll <aaronc@cse.unsw.edu.au> based |
| 1369 | on documentation by Jens Axboe. |
| 1370 | .SH "REPORTING BUGS" |
| 1371 | Report bugs to the \fBfio\fR mailing list <fio@vger.kernel.org>. |
| 1372 | See \fBREADME\fR. |
| 1373 | .SH "SEE ALSO" |
| 1374 | For further documentation see \fBHOWTO\fR and \fBREADME\fR. |
| 1375 | .br |
| 1376 | Sample jobfiles are available in the \fBexamples\fR directory. |
| 1377 | |