| 1 |
.\" Man page generated from reStructuredText |
| 2 |
.\" by the Docutils 0.22.4 manpage writer. |
| 3 |
. |
| 4 |
. |
| 5 |
.nr rst2man-indent-level 0 |
| 6 |
. |
| 7 |
.de1 rstReportMargin |
| 8 |
\\$1 \\n[an-margin] |
| 9 |
level \\n[rst2man-indent-level] |
| 10 |
level margin: \\n[rst2man-indent\\n[rst2man-indent-level]] |
| 11 |
- |
| 12 |
\\n[rst2man-indent0] |
| 13 |
\\n[rst2man-indent1] |
| 14 |
\\n[rst2man-indent2] |
| 15 |
.. |
| 16 |
.de1 INDENT |
| 17 |
.\" .rstReportMargin pre: |
| 18 |
. RS \\$1 |
| 19 |
. nr rst2man-indent\\n[rst2man-indent-level] \\n[an-margin] |
| 20 |
. nr rst2man-indent-level +1 |
| 21 |
.\" .rstReportMargin post: |
| 22 |
.. |
| 23 |
.de UNINDENT |
| 24 |
. RE |
| 25 |
.\" indent \\n[an-margin] |
| 26 |
.\" old: \\n[rst2man-indent\\n[rst2man-indent-level]] |
| 27 |
.nr rst2man-indent-level -1 |
| 28 |
.\" new: \\n[rst2man-indent\\n[rst2man-indent-level]] |
| 29 |
.in \\n[rst2man-indent\\n[rst2man-indent-level]]u |
| 30 |
.. |
| 31 |
.TH "FLENT" "1" "June 16, 2026" "2.3.0" "Flent: The FLExible Network Tester" |
| 32 |
.SH NAME |
| 33 |
flent \- Flent: The FLExible Network Tester |
| 34 |
.SH INTRODUCTION |
| 35 |
.sp |
| 36 |
Flent is a wrapper around \fInetperf\fP and similar tools to run predefined |
| 37 |
tests and aggregate and plot the results. It defines several tests that |
| 38 |
can be run against one or more hosts, primarily targeted at testing for |
| 39 |
the presence of bufferbloat under various conditions. |
| 40 |
.sp |
| 41 |
The aggregated data is saved in (gzipped) JSON format for later |
| 42 |
processing and/or import into other tools. The JSON format is documented |
| 43 |
below. |
| 44 |
.sp |
| 45 |
Apart from the JSON format, the data can be output as csv values, emacs |
| 46 |
org mode tables or plots. Each test can specify several different plots, |
| 47 |
including time\-series plots of the values against each other, as well as |
| 48 |
CDF plots of (e.g.) ping times. |
| 49 |
.sp |
| 50 |
Plots can be output to the formats supported by matplotlib by specifying |
| 51 |
the output filename with \fB\-o\fP \fIoutput.{png,ps,pdf,svg}\fP\&. If no output |
| 52 |
file is specified, the plot is displayed using matplotlib’s interactive |
| 53 |
plot browser, which also allows saving of the output (in .png format). |
| 54 |
.SS Requirements |
| 55 |
.sp |
| 56 |
Flent runs on Python, versions 2.7+ and 3.3+. Plotting requires a functional |
| 57 |
\fBmatplotlib\fP installation (but everything else can run without |
| 58 |
\fBmatplotlib\fP). For the interactive plot viewer, a graphical display (and |
| 59 |
suitably configured \fBmatplotlib\fP) is required. |
| 60 |
.sp |
| 61 |
Most tests employ the \fBnetperf\fP benchmarking tool to run the tests. |
| 62 |
Version 2.6 or higher is required, and netperf must be compiled with the |
| 63 |
\fB\-\-enable\-demo\fP option passed to \fB\&./configure\fP\&. Some tests use |
| 64 |
iperf in addition to, or instead of netperf. Both tools must be available in the |
| 65 |
\fBPATH\fP\&. |
| 66 |
.sp |
| 67 |
For ICMP ping measurements, the version of ping employed must support output |
| 68 |
timestamping (the \-D parameter to GNU ping). This is not supported by the BSD |
| 69 |
and OSX versions of ping. As an alternative to the regular ping command, the |
| 70 |
\fBfping\fP utility (see \%<http://\:fping\:.org>) can be employed. In that case |
| 71 |
fping must be version 3.5 or greater. Flent will attempt to detect the presence |
| 72 |
of fping in the \fBPATH\fP and check for support for the \-D parameter. If |
| 73 |
this check is successful, \fBfping\fP will be employed for ping data, |
| 74 |
otherwise the system ping will be used. |
| 75 |
.sp |
| 76 |
The \fBirtt\fP tool is highly recommended for UDP measurements. See |
| 77 |
\%<https://\:github\:.com/\:peteheist/\:irtt>\&. Flent will automatically detect if irtt is |
| 78 |
available in the \fBPATH\fP and use it if it is detected. Note that the |
| 79 |
server component of irtt needs to be running on the target host(s) as well. |
| 80 |
.SH RUNNING FLENT |
| 81 |
.sp |
| 82 |
When run, flent must be supplied either (a) a test name and one or more host |
| 83 |
names to connect to, or (b) one or more input files containing data from |
| 84 |
previous runs to post\-process. |
| 85 |
.sp |
| 86 |
Test names, hostnames and input file names can all be specified as unqualified |
| 87 |
arguments, and flent will do its best to guess which is which. For each |
| 88 |
argument, if it is an existing file, it is assumed to be an input file, if it is |
| 89 |
the name of an existing test configuration it’s assumed to be a test name, and |
| 90 |
if neither of those are true, it is assumed to be a host name. The \fB\-i\fP |
| 91 |
and \fB\-H\fP switches can be used to explicitly specify the interpretation |
| 92 |
of an argument. |
| 93 |
.SS Invocation |
| 94 |
.sp |
| 95 |
\fBflent\fP [\fIoptions\fP] \fI<host|test|input file\fP ...\fI>\fP |
| 96 |
.SS General options |
| 97 |
.INDENT 0.0 |
| 98 |
.TP |
| 99 |
.B \-o OUTPUT, \-\-output=OUTPUT |
| 100 |
File to write processed output to (default standard out). |
| 101 |
.UNINDENT |
| 102 |
.INDENT 0.0 |
| 103 |
.TP |
| 104 |
.B \-D DATA_DIR, \-\-data\-dir=DATA_DIR |
| 105 |
Directory to store data files in. Defaults to the current directory. |
| 106 |
.UNINDENT |
| 107 |
.INDENT 0.0 |
| 108 |
.TP |
| 109 |
.B \-i INPUT, \-\-input=INPUT |
| 110 |
File to read input from (instead of running tests). Input files can also be |
| 111 |
specified as unqualified arguments without using the \fB\-i\fP switch. |
| 112 |
.UNINDENT |
| 113 |
.INDENT 0.0 |
| 114 |
.TP |
| 115 |
.B \-f FORMAT, \-\-format=FORMAT |
| 116 |
Select output format (plot, csv, org_table, stats). Default is no processed |
| 117 |
output (just writes the JSON data file). |
| 118 |
.UNINDENT |
| 119 |
.INDENT 0.0 |
| 120 |
.TP |
| 121 |
.B \-p PLOT, \-\-plot=PLOT |
| 122 |
Select which plot to output for the given test (implies \fB\-f\fP plot). Use the |
| 123 |
\fB\-\-list\-plots\fP option to see available plots. |
| 124 |
.UNINDENT |
| 125 |
.INDENT 0.0 |
| 126 |
.TP |
| 127 |
.B \-t TITLE, \-\-title\-extra=TITLE |
| 128 |
Text to add to plot title and data file name. |
| 129 |
.UNINDENT |
| 130 |
.INDENT 0.0 |
| 131 |
.TP |
| 132 |
.B \-n NOTE, \-\-note=NOTE |
| 133 |
Add arbitrary text as a note to be stored in the JSON data file (under the |
| 134 |
\fINOTE\fP key in the metadata object). |
| 135 |
.UNINDENT |
| 136 |
.INDENT 0.0 |
| 137 |
.TP |
| 138 |
.B \-r RCFILE, \-\-rcfile=RCFILE |
| 139 |
Load configuration data from \fIRCFILE\fP (default \fI~/.flentrc\fP). See section |
| 140 |
below for information on the rc file format. |
| 141 |
.UNINDENT |
| 142 |
.INDENT 0.0 |
| 143 |
.TP |
| 144 |
.B \-x, \-\-extended\-metadata |
| 145 |
Collect extended metadata and store it with the data file. May include |
| 146 |
details of your machine you don’t want to distribute; see the section on the |
| 147 |
data format below. |
| 148 |
.UNINDENT |
| 149 |
.INDENT 0.0 |
| 150 |
.TP |
| 151 |
.B \-\-remote\-metadata=HOSTNAME |
| 152 |
Collect extended metadata from a remote host. \fIHOSTNAME\fP is passed verbatim |
| 153 |
to ssh, so can include hosts specified in ~/.ssh/config. Note that gathering |
| 154 |
the data can take some time, since it involves executing several remote |
| 155 |
commands. This option can be specified multiple times and implies |
| 156 |
\fB\-\-extended\-metadata\fP\&. |
| 157 |
.UNINDENT |
| 158 |
.INDENT 0.0 |
| 159 |
.TP |
| 160 |
.B \-\-gui |
| 161 |
Run the flent GUI. All other options are used as defaults in the GUI, but can |
| 162 |
be changed once it is running. The GUI can also be started by running the |
| 163 |
\fBflent\-gui\fP binary. For more information on the GUI, see the |
| 164 |
The Flent GUI section. |
| 165 |
.UNINDENT |
| 166 |
.INDENT 0.0 |
| 167 |
.TP |
| 168 |
.B \-\-new\-gui\-instance |
| 169 |
Start a new GUI instance. Otherwise, flent will try to connect to an already |
| 170 |
running GUI instance and have that load any new data files specified as |
| 171 |
arguments. Implies \fB\-\-gui\fP when passed on the command line, but not when |
| 172 |
set in the rc file. Note that when multiple GUI instances are running, there |
| 173 |
is no guarantee as to which instance will get a subsequent open request (if |
| 174 |
run again without \fB\-\-new\-gui\-instance\fP). |
| 175 |
.UNINDENT |
| 176 |
.INDENT 0.0 |
| 177 |
.TP |
| 178 |
.B \-\-gui\-no\-defer |
| 179 |
Normally, the GUI defers redrawing plots until they are needed to avoid |
| 180 |
redrawing all open plots every time an option changes. This switch turns off |
| 181 |
that optimisation in favour of always redrawing everything straight away. |
| 182 |
This is useful when loading a bunch of plots from the command line and then |
| 183 |
wanting to flip through them without drawing delay. |
| 184 |
.UNINDENT |
| 185 |
.INDENT 0.0 |
| 186 |
.TP |
| 187 |
.B \-b BATCH_NAME, \-\-batch\-name=BATCH_NAME |
| 188 |
Run test batch BATCH_NAME (must be specified in a batch file loaded by the |
| 189 |
\fB\-\-batch\-file\fP option). Can be supplied multiple times. |
| 190 |
.UNINDENT |
| 191 |
.INDENT 0.0 |
| 192 |
.TP |
| 193 |
.B \-B BATCH_FILE, \-\-batch\-file=BATCH_FILE |
| 194 |
Load batch file BATCH_FILE. Can be specified multiple times, in which case |
| 195 |
the files will be combined (with identically\-named sections being overridden |
| 196 |
by later files). See appropriate section below for an explanation of the |
| 197 |
batch file format. |
| 198 |
.UNINDENT |
| 199 |
.INDENT 0.0 |
| 200 |
.TP |
| 201 |
.B \-\-batch\-override=key=value |
| 202 |
Override parameter ’key’ in the batch config and set it to ’value’. The key |
| 203 |
name will be case folded to lower case. Can be specified multiple times. |
| 204 |
.UNINDENT |
| 205 |
.INDENT 0.0 |
| 206 |
.TP |
| 207 |
.B \-\-batch\-dry\-run |
| 208 |
Dry batch run. Prints what would be done, but doesn’t actually run any tests. |
| 209 |
.UNINDENT |
| 210 |
.INDENT 0.0 |
| 211 |
.TP |
| 212 |
.B \-\-batch\-verbose |
| 213 |
Be verbose during batch run: Print all commands executed. |
| 214 |
.UNINDENT |
| 215 |
.INDENT 0.0 |
| 216 |
.TP |
| 217 |
.B \-\-batch\-no\-shuffle |
| 218 |
Do not randomise the order of test runs within each batch. |
| 219 |
.UNINDENT |
| 220 |
.INDENT 0.0 |
| 221 |
.TP |
| 222 |
.B \-\-batch\-repetitions=REPETITIONS |
| 223 |
Shorthand for \fB\-\-batch\-override\fP \fB\(aqrepetitions=REPETITIONS’\fP\&. |
| 224 |
.UNINDENT |
| 225 |
.INDENT 0.0 |
| 226 |
.TP |
| 227 |
.B \-\-batch\-title=TITLE |
| 228 |
Shorthand for \fB\-\-batch\-override\fP \fB\(aqbatch_title=TITLE’\fP\&. |
| 229 |
.UNINDENT |
| 230 |
.INDENT 0.0 |
| 231 |
.TP |
| 232 |
.B \-\-batch\-resume=DIR |
| 233 |
Try to resume a previously interrupted batch run. The argument is the |
| 234 |
top\-level output directory from the previous run. |
| 235 |
.sp |
| 236 |
This will attempt to find a data file in the resume directory and load the |
| 237 |
BATCH_TIME from the previous run from that and continue. The assumption is |
| 238 |
that the output directory and filenames are generated from the batch time, so |
| 239 |
that they will match with the previous run when the same time is used. Then, |
| 240 |
tests for which data files already exist will be skipped on this run. If the |
| 241 |
rest of the batch invocation is different from the one being resumed, results |
| 242 |
may not be what you want. |
| 243 |
.sp |
| 244 |
There\(aqs a check to ensure that the generated output path is a subdirectory of |
| 245 |
the resume directory, and the whole run will be aborted if it isn\(aqt. |
| 246 |
.UNINDENT |
| 247 |
.SS Test configuration options |
| 248 |
.sp |
| 249 |
These options affect the behaviour of the test being run and have no effect when |
| 250 |
parsing input files. |
| 251 |
.INDENT 0.0 |
| 252 |
.TP |
| 253 |
.B \-H HOST, \-\-host=HOST |
| 254 |
Host to connect to for tests. For tests that support it, multiple hosts can |
| 255 |
be specified by supplying this option multiple times. Hosts can also be |
| 256 |
specified as unqualified arguments; this parameter guarantees that the |
| 257 |
argument be interpreted as a host name (rather than being subject to |
| 258 |
auto\-detection between input files, hostnames and test names). |
| 259 |
.UNINDENT |
| 260 |
.INDENT 0.0 |
| 261 |
.TP |
| 262 |
.B \-\-local\-bind=IP |
| 263 |
Local hostname or IP address to bind to (for test tools that support this). |
| 264 |
Can be specified multiple times for tests that connect to more than one host; |
| 265 |
if it is, it must be specified as many times as there are hosts. |
| 266 |
.UNINDENT |
| 267 |
.INDENT 0.0 |
| 268 |
.TP |
| 269 |
.B \-\-remote\-host=idx=HOSTNAME |
| 270 |
A remote hostname to connect to when starting a test. The idx is the runner |
| 271 |
index, which is assigned sequentially to each \fIrunner\fP (and so it is \fInot\fP |
| 272 |
the same as the sequence of hostnames). Look for the \(aqIDX\(aq key in SERIES_META |
| 273 |
for a test get the idx used here, but note that the idx assignment depends on |
| 274 |
the exact arguments to the test. |
| 275 |
.sp |
| 276 |
This works by simply prepending \(aqssh HOSTNAME\(aq to the runner command, so it |
| 277 |
relies on the same binaries being in the same places on both machines, and |
| 278 |
won\(aqt work for all runners. |
| 279 |
.sp |
| 280 |
This option can be specified multiple times to have multiple runners run on |
| 281 |
remote hosts. |
| 282 |
.UNINDENT |
| 283 |
.INDENT 0.0 |
| 284 |
.TP |
| 285 |
.B \-l LENGTH, \-\-length=LENGTH |
| 286 |
Base test length (some tests may add some time to this). |
| 287 |
.UNINDENT |
| 288 |
.INDENT 0.0 |
| 289 |
.TP |
| 290 |
.B \-s STEP_SIZE, \-\-step\-size=STEP_SIZE |
| 291 |
Measurement data point step size. |
| 292 |
.UNINDENT |
| 293 |
.INDENT 0.0 |
| 294 |
.TP |
| 295 |
.B \-d DELAY, \-\-delay=DELAY |
| 296 |
Number of seconds to delay parts of test (such as bandwidth loaders). |
| 297 |
.UNINDENT |
| 298 |
.INDENT 0.0 |
| 299 |
.TP |
| 300 |
.B \-4, \-\-ipv4 |
| 301 |
Use IPv4 for tests (some tests may ignore this). |
| 302 |
.UNINDENT |
| 303 |
.INDENT 0.0 |
| 304 |
.TP |
| 305 |
.B \-6, \-\-ipv6 |
| 306 |
Use IPv6 for tests (some tests may ignore this). |
| 307 |
.UNINDENT |
| 308 |
.INDENT 0.0 |
| 309 |
.TP |
| 310 |
.B \-\-socket\-timeout=SOCKET_TIMEOUT |
| 311 |
Socket timeout (in seconds) used for UDP delay measurement, to prevent stalls |
| 312 |
on packet loss. Only enabled if the installed netperf version is detected to |
| 313 |
support this (requires SVN version of netperf). |
| 314 |
.sp |
| 315 |
For the default value, see the output of flent \fB\-h\fP\&. The value of |
| 316 |
this parameter is an implicit upper bound on how long a round\-trip time that |
| 317 |
can be measured. As such you may need to adjust it if you are experiencing |
| 318 |
latency above the default value. Set to 0 to disable. |
| 319 |
.UNINDENT |
| 320 |
.INDENT 0.0 |
| 321 |
.TP |
| 322 |
.B \-\-send\-size=SEND_SIZE |
| 323 |
Send size (in bytes) used for TCP tests. Netperf uses the socket buffer size |
| 324 |
by default, which if too large can cause spikes in the throughput results. |
| 325 |
Lowering this value will increase CPU usage but also improves the fidelity |
| 326 |
of the throughput results without having to decrease the socket buffer size. |
| 327 |
.sp |
| 328 |
Can be specified multiple times, with each value corresponding to a stream |
| 329 |
of a test. If only specified once, the same value will be applied to all |
| 330 |
flows. |
| 331 |
.UNINDENT |
| 332 |
.INDENT 0.0 |
| 333 |
.TP |
| 334 |
.B \-\-test\-parameter=key=value |
| 335 |
Arbitrary test parameter in key=value format. Key will be case folded to |
| 336 |
lower case. The values are stored with the results metadata, and so can be |
| 337 |
used for storing arbitrary information relevant for a particular test run. |
| 338 |
.sp |
| 339 |
In addition to serving as simple metadata, the test parameters can also |
| 340 |
affect the behaviour of some test configurations. See the Supplied Tests |
| 341 |
section for information on these. |
| 342 |
.sp |
| 343 |
This option can be specified multiple times to set multiple test parameters. |
| 344 |
.UNINDENT |
| 345 |
.INDENT 0.0 |
| 346 |
.TP |
| 347 |
.B \-\-swap\-up\-down |
| 348 |
Switch upstream and downstream directions for data transfer. This means that |
| 349 |
’upload’ will become ’download’ and vice versa. Works by exchanging netperf |
| 350 |
\fBTCP_MAERTS\fP and \fBTCP_STREAM\fP parameters, so only works for tests that employ |
| 351 |
these as their data transfer, and only for the TCP streams. |
| 352 |
.UNINDENT |
| 353 |
.INDENT 0.0 |
| 354 |
.TP |
| 355 |
.B \-\-socket\-stats |
| 356 |
Parse socket stats during test. This will capture and parse socket |
| 357 |
statistics for all TCP upload flows during a test, adding TCP cwnd and RTT |
| 358 |
values to the test data. Requires the \(aqss\(aq utility to be present on the |
| 359 |
system, and can fail if there are too many simultaneous upload flows; which |
| 360 |
is why this option is not enabled by default. |
| 361 |
.UNINDENT |
| 362 |
.INDENT 0.0 |
| 363 |
.TP |
| 364 |
.B \-\-marking\-name |
| 365 |
Define a new symbolic name that can be used when specifying flow markings |
| 366 |
using the \(aqmarkings\(aq test parameter. This can be used to make it easier to |
| 367 |
specify custom diffserv markings on flows by using symbolic names for each |
| 368 |
marking value instead of the hex codes. Values specified here will be used |
| 369 |
in addition to the common values (listed below), and cannot override the |
| 370 |
built\-in names. Names will be case\-folded when matching. |
| 371 |
.sp |
| 372 |
The list of symbolic markings natively supported, along with their hex |
| 373 |
expansions, are: |
| 374 |
.INDENT 7.0 |
| 375 |
.INDENT 3.5 |
| 376 |
.sp |
| 377 |
.EX |
| 378 |
AF11: 0x28 CS0: 0x00 |
| 379 |
AF12: 0x30 CS1: 0x20 |
| 380 |
AF13: 0x38 CS2: 0x40 |
| 381 |
AF21: 0x48 CS3: 0x60 |
| 382 |
AF22: 0x50 CS4: 0x80 |
| 383 |
AF23: 0x58 CS5: 0xa0 |
| 384 |
AF31: 0x68 CS6: 0xc0 |
| 385 |
AF32: 0x70 CS7: 0xe0 |
| 386 |
AF33: 0x78 EF: 0xb8 |
| 387 |
AF41: 0x88 |
| 388 |
AF42: 0x90 |
| 389 |
AF43: 0x98 |
| 390 |
.EE |
| 391 |
.UNINDENT |
| 392 |
.UNINDENT |
| 393 |
.sp |
| 394 |
Note that the hexadecimal values denote the value of the full ToS byte |
| 395 |
(including the two ECN bits), so they need to be right\-shifted by two bits |
| 396 |
to get the corresponding diffserv code points. |
| 397 |
.UNINDENT |
| 398 |
.SS Plot configuration options |
| 399 |
.sp |
| 400 |
These options are used to configure the appearance of plot output and only make |
| 401 |
sense combined with \fB\-f\fP \fIplot\fP\&. |
| 402 |
.INDENT 0.0 |
| 403 |
.TP |
| 404 |
.B \-\-label\-x=LABEL |
| 405 |
.UNINDENT |
| 406 |
.INDENT 0.0 |
| 407 |
.TP |
| 408 |
.B \-\-label\-y=LABEL |
| 409 |
Override the figure axis labels. Can be specified twice, corresponding to |
| 410 |
figures with multiple axes. |
| 411 |
.UNINDENT |
| 412 |
.INDENT 0.0 |
| 413 |
.TP |
| 414 |
.B \-I, \-\-invert\-latency\-y |
| 415 |
Invert latency data series axis (typically the Y\-axis), making plots show |
| 416 |
’better\(aq values upwards. |
| 417 |
.UNINDENT |
| 418 |
.INDENT 0.0 |
| 419 |
.TP |
| 420 |
.B \-z, \-\-zero\-y |
| 421 |
Always start Y axis of plot at zero, instead of autoscaling the axis. |
| 422 |
Autoscaling is still enabled for the upper bound. This also disables log |
| 423 |
scale if enabled. |
| 424 |
.UNINDENT |
| 425 |
.INDENT 0.0 |
| 426 |
.TP |
| 427 |
.B \-\-log\-scale={log2,log10} |
| 428 |
Use the specified logarithmic scale on plots. |
| 429 |
.UNINDENT |
| 430 |
.INDENT 0.0 |
| 431 |
.TP |
| 432 |
.B \-\-norm\-factor=FACTOR |
| 433 |
Data normalisation factor. Divide all data points by this value. Can be |
| 434 |
specified multiple times, in which case each value corresponds to a data |
| 435 |
series. |
| 436 |
.UNINDENT |
| 437 |
.INDENT 0.0 |
| 438 |
.TP |
| 439 |
.B \-\-bounds\-x=BOUNDS |
| 440 |
.UNINDENT |
| 441 |
.INDENT 0.0 |
| 442 |
.TP |
| 443 |
.B \-\-bounds\-y=BOUNDS |
| 444 |
Specify bounds of the plot axes. If specifying one number, that will become |
| 445 |
the upper bound. Specify two numbers separated by a comma to specify both |
| 446 |
upper and lower bounds. To specify just the lower bound, add a comma |
| 447 |
afterwards. Can be specified twice, corresponding to figures with multiple |
| 448 |
axes. |
| 449 |
.UNINDENT |
| 450 |
.INDENT 0.0 |
| 451 |
.TP |
| 452 |
.B \-S, \-\-scale\-mode |
| 453 |
Treat file names (except for the first one) passed as unqualified arguments |
| 454 |
as if passed as \fB\-\-scale\-data\fP (default as if passed as |
| 455 |
\fB\-\-input\fP). |
| 456 |
.UNINDENT |
| 457 |
.INDENT 0.0 |
| 458 |
.TP |
| 459 |
.B \-\-concatenate |
| 460 |
Concatenate multiple result sets into one data series. This means that each |
| 461 |
data file will have its time axis shifted by the preceding series duration |
| 462 |
and appended to the first data set specified. Only works for data sets from |
| 463 |
the same test, obviously. |
| 464 |
.UNINDENT |
| 465 |
.INDENT 0.0 |
| 466 |
.TP |
| 467 |
.B \-\-absolute\-time |
| 468 |
Plot data points with absolute UNIX time on the x\-axis. This requires the |
| 469 |
absolute starting time for the test run to be stored in the data file, and so |
| 470 |
it won’t work with data files that predates this feature. |
| 471 |
.UNINDENT |
| 472 |
.INDENT 0.0 |
| 473 |
.TP |
| 474 |
.B \-\-subplot\-combine |
| 475 |
When plotting multiple data series, plot each one on a separate subplot |
| 476 |
instead of combining them into one plot. This mode is not supported for all |
| 477 |
plot types, and only works when \fB\-\-scale\-mode\fP is disabled. |
| 478 |
.UNINDENT |
| 479 |
.INDENT 0.0 |
| 480 |
.TP |
| 481 |
.B \-\-skip\-missing\-series |
| 482 |
Skip missing series entirely from bar plots, instead of leaving an empty |
| 483 |
space for it. |
| 484 |
.UNINDENT |
| 485 |
.INDENT 0.0 |
| 486 |
.TP |
| 487 |
.B \-\-no\-print\-n |
| 488 |
Do not print the number of data points on combined plots. When using plot |
| 489 |
types that combines results from several test runs, the number of data series |
| 490 |
in each combined data point is normally added after the series name, (n=X) |
| 491 |
for X data series. This option turns that off. |
| 492 |
.UNINDENT |
| 493 |
.INDENT 0.0 |
| 494 |
.TP |
| 495 |
.B \-\-no\-annotation |
| 496 |
Exclude annotation with hostnames, time and test length from plots. |
| 497 |
.UNINDENT |
| 498 |
.INDENT 0.0 |
| 499 |
.TP |
| 500 |
.B \-\-figure\-note=NOTE, \-\-fig\-note=NOTE |
| 501 |
Add a note (arbitrary text) to the bottom\-left of the figure. |
| 502 |
.UNINDENT |
| 503 |
.INDENT 0.0 |
| 504 |
.TP |
| 505 |
.B \-\-no\-title |
| 506 |
Exclude title from plots. |
| 507 |
.UNINDENT |
| 508 |
.INDENT 0.0 |
| 509 |
.TP |
| 510 |
.B \-\-override\-title=TITLE |
| 511 |
Override plot title with this string. Completely discards the configured |
| 512 |
title (from the test configuration), as well as the title stored in the data |
| 513 |
set, and replaces it with the value supplied here. This is useful to override |
| 514 |
the plot title \fIat the time of plotting\fP, for instance to add a title to an |
| 515 |
aggregate plot from several data series. When this parameter is specified, |
| 516 |
\fB\-\-no\-title\fP has no effect. |
| 517 |
.UNINDENT |
| 518 |
.INDENT 0.0 |
| 519 |
.TP |
| 520 |
.B \-\-no\-labels |
| 521 |
Hides tick labels from box and bar plots. |
| 522 |
.UNINDENT |
| 523 |
.INDENT 0.0 |
| 524 |
.TP |
| 525 |
.B \-\-no\-markers |
| 526 |
Don’t use line markers to differentiate data series on plots. |
| 527 |
.UNINDENT |
| 528 |
.INDENT 0.0 |
| 529 |
.TP |
| 530 |
.B \-\-no\-legend |
| 531 |
Exclude legend from plots. |
| 532 |
.UNINDENT |
| 533 |
.INDENT 0.0 |
| 534 |
.TP |
| 535 |
.B \-\-horizontal\-legend |
| 536 |
Place a horizontal legend below the plot instead of a vertical one next to |
| 537 |
it. Doesn\(aqt always work well if there are too many items in the legend. |
| 538 |
.UNINDENT |
| 539 |
.INDENT 0.0 |
| 540 |
.TP |
| 541 |
.B \-\-legend\-title=LEGEND_TITLE |
| 542 |
Override legend title on plot. |
| 543 |
.UNINDENT |
| 544 |
.INDENT 0.0 |
| 545 |
.TP |
| 546 |
.B \-\-legend\-placement=LEGEND_PLACEMENT |
| 547 |
Control legend placement. Enabling this option will place the legend inside |
| 548 |
the plot at the specified location. Can be one of \(aqbest\(aq, \(aqupper right\(aq, |
| 549 |
\(aqupper left\(aq, \(aqlower left\(aq, \(aqlower right\(aq, \(aqright\(aq, \(aqcenter left\(aq, \(aqcenter |
| 550 |
right\(aq, \(aqlower center\(aq, \(aqupper center\(aq or \(aqcenter\(aq. |
| 551 |
.UNINDENT |
| 552 |
.INDENT 0.0 |
| 553 |
.TP |
| 554 |
.B \-\-legend\-columns=LEGEND_COLUMNS |
| 555 |
.TP |
| 556 |
.B Set the number of columns in the legend. |
| 557 |
.UNINDENT |
| 558 |
.INDENT 0.0 |
| 559 |
.TP |
| 560 |
.B \-\-reverse\-legend |
| 561 |
Reverse the order of items in the legend. This can be useful to make the |
| 562 |
legend order match the data series in some cases. |
| 563 |
.UNINDENT |
| 564 |
.INDENT 0.0 |
| 565 |
.TP |
| 566 |
.B \-\-filter\-legend |
| 567 |
Filter legend labels by removing the longest common substring from all |
| 568 |
entries. This is not particularly smart, so use with care. |
| 569 |
.UNINDENT |
| 570 |
.INDENT 0.0 |
| 571 |
.TP |
| 572 |
.B \-\-replace\-legend=src=dest |
| 573 |
Replace \(aqsrc\(aq with \(aqdst\(aq in legends. Can be specified multiple times. |
| 574 |
.UNINDENT |
| 575 |
.INDENT 0.0 |
| 576 |
.TP |
| 577 |
.B \-\-filter\-regexp=REGEXP |
| 578 |
Filter the plot legend by the supplied regular expression (removing any text |
| 579 |
from the legend that matches the expression). Can be specified multiple |
| 580 |
times, in which case the filters will be applied in the order specified. Note |
| 581 |
that for combining several plot results, the regular expression is also |
| 582 |
applied before the grouping logic, meaning that a too wide filter can mess up |
| 583 |
the grouping. |
| 584 |
.UNINDENT |
| 585 |
.INDENT 0.0 |
| 586 |
.TP |
| 587 |
.B \-\-override\-label=LABEL |
| 588 |
Override dataset label. Can be specified multiple times when multiple |
| 589 |
datasets are being plotted, in which case the order of labels corresponds to |
| 590 |
the order of datasets. |
| 591 |
.sp |
| 592 |
Like \fB\-\-override\-title\fP, this is applied \fIat the time of plotting\fP\&. |
| 593 |
.UNINDENT |
| 594 |
.INDENT 0.0 |
| 595 |
.TP |
| 596 |
.B \-\-filter\-series=SERIES |
| 597 |
Filter out specified series from plot. Can be specified multiple times. |
| 598 |
.UNINDENT |
| 599 |
.INDENT 0.0 |
| 600 |
.TP |
| 601 |
.B \-\-split\-group=LABEL |
| 602 |
Split data sets into groups when creating box plots. Specify this option |
| 603 |
multiple times to define the new groups; the value of each option is the |
| 604 |
group name. |
| 605 |
.sp |
| 606 |
Say you\(aqre plotting nine datasets which are really testing two variables with |
| 607 |
three values each. In this case, it can be useful to have the box plot of the |
| 608 |
results be split into three parts (corresponding to the values of one |
| 609 |
variable) with each three boxes in each of them (corresponding to the values |
| 610 |
of the second variable). This option makes this possible; simply specify it |
| 611 |
three times with the labels to be used for the three groups. |
| 612 |
.sp |
| 613 |
A constraint on this option is that the number of datasets being plotted must |
| 614 |
be divisible by the number of groups. |
| 615 |
.UNINDENT |
| 616 |
.INDENT 0.0 |
| 617 |
.TP |
| 618 |
.B \-\-colours=COLOURS |
| 619 |
Comma\-separated list of colours to be used for the plot colour cycle. Can be |
| 620 |
specified in any format understood by matplotlib (including HTML hex values |
| 621 |
prefixed with a #). |
| 622 |
.sp |
| 623 |
Yes, this option uses British spelling. No, American spelling is not |
| 624 |
supported. |
| 625 |
.UNINDENT |
| 626 |
.INDENT 0.0 |
| 627 |
.TP |
| 628 |
.B \-\-override\-colour\-mode=MODE |
| 629 |
Override colour_mode attribute. This changes the way colours are |
| 630 |
assigned to bar plots. The default is \(aqgroups\(aq which assigns a separate |
| 631 |
colour to each group of data series. The alternative is \(aqseries\(aq which |
| 632 |
assigns a separate colour to each series, repeating them for each data |
| 633 |
group. |
| 634 |
.UNINDENT |
| 635 |
.INDENT 0.0 |
| 636 |
.TP |
| 637 |
.B \-\-override\-group\-by=GROUP |
| 638 |
Override the \fBgroup_by\fP setting for combination plots. This is useful to, |
| 639 |
for instance, switch to splitting up combined data sets by batch run instead |
| 640 |
of by file name. |
| 641 |
.UNINDENT |
| 642 |
.INDENT 0.0 |
| 643 |
.TP |
| 644 |
.B \-\-combine\-save\-dir=DIRNAME |
| 645 |
When doing a combination plot save the intermediate data to \fBDIRNAME\fP\&. This |
| 646 |
can then be used for subsequent plotting to avoid having to load all the |
| 647 |
source data files again on each plot. |
| 648 |
.UNINDENT |
| 649 |
.INDENT 0.0 |
| 650 |
.TP |
| 651 |
.B \-\-figure\-width=FIG_WIDTH |
| 652 |
Figure width in inches. Used when saving plots to file and for default size |
| 653 |
of the interactive plot window. |
| 654 |
.UNINDENT |
| 655 |
.INDENT 0.0 |
| 656 |
.TP |
| 657 |
.B \-\-figure\-height=FIG_HEIGHT |
| 658 |
Figure height in inches. Used when saving plots to file and for default size |
| 659 |
of the interactive plot window. |
| 660 |
.UNINDENT |
| 661 |
.INDENT 0.0 |
| 662 |
.TP |
| 663 |
.B \-\-figure\-dpi=FIG_DPI |
| 664 |
Figure DPI. Used when saving plots to raster format files. |
| 665 |
.UNINDENT |
| 666 |
.INDENT 0.0 |
| 667 |
.TP |
| 668 |
.B \-\-fallback\-layout |
| 669 |
Use the fallback layout engine (tight_layout built in to matplotlib). Use |
| 670 |
this if text is cut off on saved figures. The downside to the fallback engine |
| 671 |
is that the size of the figure (as specified by \fB\-\-figure\-width\fP and |
| 672 |
\fB\-\-figure\-height\fP) is no longer kept constant.) |
| 673 |
.UNINDENT |
| 674 |
.INDENT 0.0 |
| 675 |
.TP |
| 676 |
.B \-\-no\-matplotlibrc |
| 677 |
Don’t load included matplotlibrc values. Use this if autodetection of custom |
| 678 |
matplotlibrc fails and flent is inadvertently overriding rc values. |
| 679 |
.UNINDENT |
| 680 |
.INDENT 0.0 |
| 681 |
.TP |
| 682 |
.B \-\-no\-hover\-highlight |
| 683 |
Don\(aqt highlight data series on hover in interactive plot views. Use this if |
| 684 |
redrawing is too slow, or the highlighting is undesired for other reasons. |
| 685 |
.UNINDENT |
| 686 |
.INDENT 0.0 |
| 687 |
.TP |
| 688 |
.B \-\-scale\-data=SCALE_DATA |
| 689 |
Additional data files to consider when scaling the plot axes (for plotting |
| 690 |
several plots with identical axes). Note, this displays only the first data |
| 691 |
set, but with axis scaling taking into account the additional data sets. Can |
| 692 |
be supplied multiple times; see also \fB\-\-scale\-mode\fP\&. |
| 693 |
.UNINDENT |
| 694 |
.SS Test tool\-related options |
| 695 |
.INDENT 0.0 |
| 696 |
.TP |
| 697 |
.B \-\-control\-host=HOST |
| 698 |
Hostname for the test control connection (for test tools that support this). |
| 699 |
Default: First hostname of test target. |
| 700 |
.sp |
| 701 |
When running tests that uses D\-ITG as a test tool (such as the voip\-* |
| 702 |
tests), this switch controls where flent will look for the D\-ITG control |
| 703 |
server (see section below on running tests with D\-ITG). For Netperf\-based |
| 704 |
tests, this option is passed to Netperf to control where to point the control |
| 705 |
connection. This is useful to, for instance, to run the control server |
| 706 |
communication over a separate control network so as to not interfere with |
| 707 |
test traffic. |
| 708 |
.sp |
| 709 |
There is also a per\-flow setting for this for tests that connect to multiple |
| 710 |
hosts; see the \fIcontrol_hosts\fP test parameter in Supplied Tests\&. If both are |
| 711 |
set, the per\-flow setting takes precedence for those tests that use it. |
| 712 |
.UNINDENT |
| 713 |
.INDENT 0.0 |
| 714 |
.TP |
| 715 |
.B \-\-control\-local\-bind=IP |
| 716 |
Local hostname or IP to bind control connection to (for test tools that |
| 717 |
support it; currently netperf). If not supplied, the value for |
| 718 |
\fB\-\-local\-bind\fP will be used. Note that if this value is passed but |
| 719 |
\fB\-\-local\-bind\fP is \fInot\fP, netperf will use the value specified here to |
| 720 |
bind the data connections to as well. |
| 721 |
.UNINDENT |
| 722 |
.INDENT 0.0 |
| 723 |
.TP |
| 724 |
.B \-\-netperf\-control\-port=PORT |
| 725 |
Port for Netperf control server. Default: 12865. |
| 726 |
.UNINDENT |
| 727 |
.INDENT 0.0 |
| 728 |
.TP |
| 729 |
.B \-\-ditg\-control\-port=PORT |
| 730 |
Port for D\-ITG control server. Default: 8000. |
| 731 |
.UNINDENT |
| 732 |
.INDENT 0.0 |
| 733 |
.TP |
| 734 |
.B \-\-ditg\-control\-secret=SECRET |
| 735 |
Secret for D\-ITG control server authentication. Default: ’’. |
| 736 |
.UNINDENT |
| 737 |
.INDENT 0.0 |
| 738 |
.TP |
| 739 |
.B \-\-http\-getter\-urllist=FILENAME |
| 740 |
When running HTTP tests, the http\-getter tool is used to fetch URLs (see |
| 741 |
\%<https://\:github\:.com/\:tohojo/\:http-getter>). This option specifies the filename |
| 742 |
containing the list of HTTP URLs to get. Can also be a URL, which will then |
| 743 |
be downloaded as part of each test iteration. If not specified, this is set |
| 744 |
to \%<http:/>/<hostname>/filelist.txt where <hostname> is the first test |
| 745 |
hostname. |
| 746 |
.UNINDENT |
| 747 |
.INDENT 0.0 |
| 748 |
.TP |
| 749 |
.B \-\-http\-getter\-dns\-servers=DNS_SERVERS |
| 750 |
DNS servers to use for http\-getter lookups. Format is |
| 751 |
host[:port][,host[:port]]... This option will only work if libcurl supports |
| 752 |
it (needs to be built with the ares resolver). Default is none (use the |
| 753 |
system resolver). |
| 754 |
.UNINDENT |
| 755 |
.INDENT 0.0 |
| 756 |
.TP |
| 757 |
.B \-\-http\-getter\-timeout=MILLISECONDS |
| 758 |
Timeout for HTTP connections. Default is to use the test length. |
| 759 |
.UNINDENT |
| 760 |
.INDENT 0.0 |
| 761 |
.TP |
| 762 |
.B \-\-http\-getter\-workers=NUMBER |
| 763 |
Number of workers to use for getting HTTP urls. Default is 4. |
| 764 |
.UNINDENT |
| 765 |
.INDENT 0.0 |
| 766 |
.TP |
| 767 |
.B \-\-irtt\-sampling\-interval=MILLISECONDS |
| 768 |
Override the sampling interval passed to irtt, in milliseconds. Can be used |
| 769 |
to run irtt with a higher sampling frequency than the rest of the test. If |
| 770 |
set, this will override the sampling interval for all instances of irtt used |
| 771 |
in the test. |
| 772 |
.UNINDENT |
| 773 |
.SS Misc and debugging options: |
| 774 |
.INDENT 0.0 |
| 775 |
.TP |
| 776 |
.B \-L LOG_FILE, \-\-log\-file=LOG_FILE |
| 777 |
Write debug log (test program output) to log file. |
| 778 |
.UNINDENT |
| 779 |
.INDENT 0.0 |
| 780 |
.TP |
| 781 |
.B \-\-list\-tests |
| 782 |
List available tests and exit. |
| 783 |
.UNINDENT |
| 784 |
.INDENT 0.0 |
| 785 |
.TP |
| 786 |
.B \-\-list\-plots |
| 787 |
List available plots for selected test and exit. |
| 788 |
.UNINDENT |
| 789 |
.INDENT 0.0 |
| 790 |
.TP |
| 791 |
.B \-V, \-\-version |
| 792 |
Show Flent version information and exit. |
| 793 |
.UNINDENT |
| 794 |
.INDENT 0.0 |
| 795 |
.TP |
| 796 |
.B \-v, \-\-verbose |
| 797 |
Enable verbose logging to console. |
| 798 |
.UNINDENT |
| 799 |
.INDENT 0.0 |
| 800 |
.TP |
| 801 |
.B \-q, \-\-quiet |
| 802 |
Disable normal logging to console (and only log warnings and errors). |
| 803 |
.UNINDENT |
| 804 |
.INDENT 0.0 |
| 805 |
.TP |
| 806 |
.B \-\-debug\-error |
| 807 |
Print full exception backtraces to console. |
| 808 |
.UNINDENT |
| 809 |
.INDENT 0.0 |
| 810 |
.TP |
| 811 |
.B \-h, \-\-help |
| 812 |
Show usage help message and exit. |
| 813 |
.UNINDENT |
| 814 |
.SS Signals |
| 815 |
.sp |
| 816 |
Flent will abort what it is currently doing on receiving a \fBSIGINT\fP \-\- this |
| 817 |
includes killing all runners, cleaning up temporary files and shutting down as |
| 818 |
gracefully as possible. Runners are killed with \fBSIGTERM\fP in this mode, and |
| 819 |
their output is discarded. If a batch run is in progress, the current test will |
| 820 |
be interrupted in this way, and the rest of the batch run is aborted. Previously |
| 821 |
completed tests and their results are not aborted. Post\-commands marked as |
| 822 |
’essential’ will be run after the test is interrupted. Additionally, flent |
| 823 |
converts \fBSIGTERM\fP into \fBSIGINT\fP internally and reacts accordingly. |
| 824 |
.sp |
| 825 |
Upon receiving a \fBSIGUSR1\fP, flent will try to gracefully abort the test it is |
| 826 |
currently running, and parse the output of the runners to the extent that any |
| 827 |
such output exists. That is, each runner will be killed by a \fBSIGINT\fP, which |
| 828 |
will cause a graceful shutdown for at least ping and netperf (although netperf |
| 829 |
running in \fITCP_MAERTS\fP mode will bug out when interrupted like this, so |
| 830 |
end\-of\-tests statistics will be missing). Flent will only react once to a |
| 831 |
\fBSIGUSR1\fP, sending exactly one \fBSIGINT\fP to the active runners, then wait for |
| 832 |
them to exit. This may take several seconds in the case of netperf. If the |
| 833 |
runners for some reason fail to exit, flent will be stuck and will need to be |
| 834 |
killed with \fBSIGINT\fP\&. If running in batch mode, \fBSIGUSR1\fP will only affect |
| 835 |
the currently running test; subsequent tests will still be run. |
| 836 |
.SH SUPPLIED TESTS |
| 837 |
.sp |
| 838 |
Test are supplied as Python files and can specify commands to run etc. |
| 839 |
For a full list of the tests supported by flent, see the |
| 840 |
\fB\-\-list\-tests\fP option. |
| 841 |
.SS The Realtime Response Under Load (RRUL) test |
| 842 |
.sp |
| 843 |
This test exists in a couple of variants and is a partial implementation of the |
| 844 |
RRUL specification as written by Dave Taht (see |
| 845 |
\%<https://\:www\:.bufferbloat\:.net/\:projects/\:bloat/\:wiki/\:RRUL_Spec/>). It works |
| 846 |
by running RTT measurement using ICMP ping and UDP roundtrip time measurement, |
| 847 |
while loading up the link with eight TCP streams (four downloads, four uploads). |
| 848 |
This quite reliably saturates the measured link (wherever the bottleneck might |
| 849 |
be), and thus exposes bufferbloat when it is present. |
| 850 |
.SS Simple TCP flow tests |
| 851 |
.sp |
| 852 |
These tests combine a TCP flow (either in one direction, or both) with an ICMP |
| 853 |
ping measurement. It’s a simpler test than RRUL, but in some cases the single |
| 854 |
TCP flow can be sufficient to saturate the link. |
| 855 |
.SS UDP flood test |
| 856 |
.sp |
| 857 |
This test runs \fIiperf\fP configured to emit 100Mbps of UDP packets targeted at the |
| 858 |
test host, while measuring RTT using ICMP ping. It is useful for observing |
| 859 |
latency in the face of a completely unresponsive packet stream. |
| 860 |
.SS Test parameters |
| 861 |
.sp |
| 862 |
Some test parameters (set with \fB\-\-test\-parameter\fP ) affect the way tests |
| 863 |
behave. These are: |
| 864 |
.INDENT 0.0 |
| 865 |
.TP |
| 866 |
.B upload_streams |
| 867 |
.UNINDENT |
| 868 |
.INDENT 0.0 |
| 869 |
.TP |
| 870 |
.B download_streams |
| 871 |
.UNINDENT |
| 872 |
.INDENT 0.0 |
| 873 |
.TP |
| 874 |
.B bidir_streams |
| 875 |
These set the number of streams for the tests that are configurable. The |
| 876 |
\fBtcp_nup\fP, \fBtcp_ndown\fP and \fBrrul_be_nflows\fP tests understand |
| 877 |
\fBupload_streams\fP and \fBdownload_streams\fP, while the \fBrrul_var\fP test |
| 878 |
understands \fBbidir_streams\fP\&. If any of these parameter is set to the |
| 879 |
special value \fBnum_cpus\fP the number of streams will be set to the number of |
| 880 |
CPUs on the system (if this information is available). |
| 881 |
.UNINDENT |
| 882 |
.INDENT 0.0 |
| 883 |
.TP |
| 884 |
.B tcp_cong_control |
| 885 |
Set the congestion control used for TCP flows, for platforms that supports |
| 886 |
setting it. This can be specified as a simple string to set the same value |
| 887 |
for upstream and downstream, or two comma\-separated values to set it |
| 888 |
separately for the upstream and downstream directions. On Linux, any value in |
| 889 |
the sysctl \fBnet.ipv4.tcp_allowed_congestion_control\fP can be used. |
| 890 |
.sp |
| 891 |
If a congestion control is specified that is not available on the system |
| 892 |
running the test, setting it will simply fail. In addition, some tests |
| 893 |
override the congestion control for one or more flows. The actual congestion |
| 894 |
control used is stored in the \fBCONG_CONTROL\fP per\-test metadata field. |
| 895 |
.UNINDENT |
| 896 |
.INDENT 0.0 |
| 897 |
.TP |
| 898 |
.B udp_bandwidth |
| 899 |
.UNINDENT |
| 900 |
.INDENT 0.0 |
| 901 |
.TP |
| 902 |
.B udp_bandwidths |
| 903 |
.UNINDENT |
| 904 |
.INDENT 0.0 |
| 905 |
.TP |
| 906 |
.B udp_pktsize |
| 907 |
.UNINDENT |
| 908 |
.INDENT 0.0 |
| 909 |
.TP |
| 910 |
.B udp_pktsizes |
| 911 |
This sets the bandwidth and packet size of each UDP stream in the \fBudp_*\fP |
| 912 |
tests. The option is passed to \fBiperf\fP so can be in any syntax the iperf |
| 913 |
understands (e.g. \fB20M\fP for 20 Mbps). |
| 914 |
.sp |
| 915 |
When running multiple UDP streams use the plural versions of the options |
| 916 |
(\fBudp_bandwidths\fP and \fBudp_pktsizes\fP) to specify individual per\-stream |
| 917 |
values (comma\-separated per stream), or the singular versions to specify the |
| 918 |
same value for all streams. |
| 919 |
.UNINDENT |
| 920 |
.INDENT 0.0 |
| 921 |
.TP |
| 922 |
.B burst_length |
| 923 |
.UNINDENT |
| 924 |
.INDENT 0.0 |
| 925 |
.TP |
| 926 |
.B burst_ports |
| 927 |
.UNINDENT |
| 928 |
.INDENT 0.0 |
| 929 |
.TP |
| 930 |
.B burst_psize |
| 931 |
.UNINDENT |
| 932 |
.INDENT 0.0 |
| 933 |
.TP |
| 934 |
.B burst_tos |
| 935 |
These set the length, number of ports to use, packet size and TOS value for |
| 936 |
the packet bursts generated in the \fBburst*\fP tests. |
| 937 |
.UNINDENT |
| 938 |
.INDENT 0.0 |
| 939 |
.TP |
| 940 |
.B cpu_stats_hosts |
| 941 |
.UNINDENT |
| 942 |
.INDENT 0.0 |
| 943 |
.TP |
| 944 |
.B netstat_hosts |
| 945 |
.UNINDENT |
| 946 |
.INDENT 0.0 |
| 947 |
.TP |
| 948 |
.B qdisc_stats_hosts |
| 949 |
.UNINDENT |
| 950 |
.INDENT 0.0 |
| 951 |
.TP |
| 952 |
.B wifi_stats_hosts |
| 953 |
These set hostnames to gather statistics from from during the test. The |
| 954 |
hostnames are passed to SSH, so can be anything understood by SSH (including |
| 955 |
using \fBusername@host\fP syntax, or using hosts defined in \fB~/.ssh/config\fP). |
| 956 |
This will attempt to run remote commands on these hosts to gather the |
| 957 |
required statistics, so passwordless login has to be enabled for. Multiple |
| 958 |
hostnames can be specified, separated by commas. |
| 959 |
.sp |
| 960 |
CPU stats and netstat output is global to the machine being connected to. The |
| 961 |
qdisc and WiFi stats need extra parameters to work. These are |
| 962 |
\fBqdisc_stats_interfaces\fP, \fBwifi_stats_interfaces\fP and |
| 963 |
\fBwifi_stats_stations\fP\&. The two former specify which interfaces to gather |
| 964 |
statistics from. These are paired with the hostnames, and so must contain the |
| 965 |
same number of elements (also comma\-separated) as the \fB_hosts\fP variables. |
| 966 |
To specify multiple interfaces on the same host, duplicate the hostname. The |
| 967 |
\fBwifi_stats_stations\fP parameter specifies MAC addresses of stations to |
| 968 |
gather statistics for. This list is the same for all hosts, but only stations |
| 969 |
present in debugfs on each host are actually captured. |
| 970 |
.sp |
| 971 |
The qdisc stats gather statistics output from \fBtc \-s\fP, while the WiFi stats |
| 972 |
gather statistics from debugfs. These are gathered by looping in a shell |
| 973 |
script; however, for better performance, the \fBtc_iterate\fP and |
| 974 |
\fBwifistats_iterate\fP programmes available in the \fBmisc/\fP directory of the |
| 975 |
source code tarball can be installed. On low\-powered systems this can be |
| 976 |
critical to get correct statistics. The helper programmes are packaged for |
| 977 |
LEDE/OpenWrt in the \fBflent\-tools\fP package. |
| 978 |
.UNINDENT |
| 979 |
.INDENT 0.0 |
| 980 |
.TP |
| 981 |
.B ping_hosts |
| 982 |
.UNINDENT |
| 983 |
.INDENT 0.0 |
| 984 |
.TP |
| 985 |
.B ping_local_binds |
| 986 |
.UNINDENT |
| 987 |
.INDENT 0.0 |
| 988 |
.TP |
| 989 |
.B ping_labels |
| 990 |
These are used to define one or more extra host names that will receive a |
| 991 |
ping flow while a test is run. The \fBping_hosts\fP variable simply specifies |
| 992 |
hostnames to ping (several can be specified by separating them with commas). |
| 993 |
The \fBping_local_binds\fP variable sets local IP address(es) to bind to for |
| 994 |
the extra ping flows. If specified, it must contain the same number of local |
| 995 |
addresses as the number of ping hosts. The same local address can be |
| 996 |
specified multiple times, however. The \fBping_labels\fP variable is used to |
| 997 |
label each of the ping flows; if not specified, Flent will create a default |
| 998 |
label based on the target hostname for each flow. |
| 999 |
.UNINDENT |
| 1000 |
.INDENT 0.0 |
| 1001 |
.TP |
| 1002 |
.B voip_host |
| 1003 |
.UNINDENT |
| 1004 |
.INDENT 0.0 |
| 1005 |
.TP |
| 1006 |
.B voip_local_bind |
| 1007 |
.UNINDENT |
| 1008 |
.INDENT 0.0 |
| 1009 |
.TP |
| 1010 |
.B voip_control_host |
| 1011 |
.UNINDENT |
| 1012 |
.INDENT 0.0 |
| 1013 |
.TP |
| 1014 |
.B voip_marking |
| 1015 |
Similar to the ping variants above, these parameters specify a hostname that |
| 1016 |
will receive a VoIP test. However, unlike the ping parameters, only one |
| 1017 |
hostname can be specified for VoIP tests, and that end\-host needs to have |
| 1018 |
either D\-ITG (and the control server) or the IRTT server running. The marking |
| 1019 |
setting controls which DiffServ marking is applied to the VoIP flow and |
| 1020 |
defaults to no marking being set. |
| 1021 |
.UNINDENT |
| 1022 |
.INDENT 0.0 |
| 1023 |
.TP |
| 1024 |
.B control_hosts |
| 1025 |
Hostnames to use for the control connections for the \fBrtt_fair*\fP tests. |
| 1026 |
Comma\-separated. If specified, it must contain as many hostnames as the |
| 1027 |
number of target hostnames specified for the test. |
| 1028 |
.UNINDENT |
| 1029 |
.INDENT 0.0 |
| 1030 |
.TP |
| 1031 |
.B markings |
| 1032 |
.UNINDENT |
| 1033 |
.INDENT 0.0 |
| 1034 |
.TP |
| 1035 |
.B labels |
| 1036 |
Flow markings to use for each of the flows in the \fBrtt_fair*\fP tests. |
| 1037 |
Comma\-separated values of markings understood by Netperf (such as \(dqCS0\(dq). |
| 1038 |
Only supports setting the same marking on both the upstream and downstream |
| 1039 |
packets of each flow (so no \(dqCS0,CS0\(dq setting as can be used for Netperf). If |
| 1040 |
not set, defaults to CS0 (best effort). If set, each value corresponds to a |
| 1041 |
flow, and any extra flows will be set to CS0. By default each flow will be |
| 1042 |
labelled according to its marking; to override these labels, use the |
| 1043 |
\fBlabels\fP parameter. |
| 1044 |
.UNINDENT |
| 1045 |
.INDENT 0.0 |
| 1046 |
.TP |
| 1047 |
.B stream_delays |
| 1048 |
Specify a per\-stream delay (in seconds) for the different streams started up |
| 1049 |
by a test. Use commas to separate values for the different streams. This can |
| 1050 |
be used to create tests with staggered start times, for example to test TCP |
| 1051 |
flow startup convergence times. What exactly constitutes a stream depends on |
| 1052 |
the test. For example, the rtt_fair* tests considers each hostname a stream, |
| 1053 |
whether or not there is one or two flows going to that host. |
| 1054 |
.UNINDENT |
| 1055 |
.SH THE FLENT GUI |
| 1056 |
.sp |
| 1057 |
Flent comes equipped with a GUI to browse and plot previously captured datasets. |
| 1058 |
The GUI requires PyQt4; if this is installed, it can be launched with the |
| 1059 |
\fB\-\-gui\fP parameter, or by launching the \fBflent\-gui\fP binary. |
| 1060 |
Additionally, if Flent is launched without parameters and without a controlling |
| 1061 |
terminal, the GUI will be launched automatically. |
| 1062 |
.sp |
| 1063 |
The GUI can be used for interactively plotting previously captured datasets, and |
| 1064 |
makes it easy to compare results from several test runs. It presents a tabbed |
| 1065 |
interface to graphs of data files, allows dynamic configuration of plots, and |
| 1066 |
includes a metadata browser. For each loaded data file, additional data files |
| 1067 |
can be loaded and added to the plot, similar to what happens when specifying |
| 1068 |
multiple input files for plotting on the command line. A checkbox controls |
| 1069 |
whether the added data files are added as separate entries to the plot, or |
| 1070 |
whether they are used for scaling the output (mirroring the |
| 1071 |
\fB\-\-scale\-mode\fP ) command line switch. |
| 1072 |
.sp |
| 1073 |
The GUI also incorporates matplotlib’s interactive browsing toolbar, enabling |
| 1074 |
panning and zooming of the plot area, dynamic configuration of plot and axis |
| 1075 |
parameters and labels and saving the plots to file. The exact dynamic features |
| 1076 |
supported depends on the installed version of matplotlib. |
| 1077 |
.SH CONFIGURATION FILES |
| 1078 |
.SS The RC file |
| 1079 |
.sp |
| 1080 |
Some of the command line options can be specified in an rc file. By default, |
| 1081 |
flent looks for this in \fB~/.flentrc\fP, but an alternative location can be |
| 1082 |
specified with the \fB\-\-rcfile\fP command line option. |
| 1083 |
.sp |
| 1084 |
The rc file allows options to be specified globally, an optionally overridden |
| 1085 |
for specific tests. For an explanation of the options, refer to the annotated |
| 1086 |
example rc file, by default installed to |
| 1087 |
\fB/usr/share/doc/flent/flentrc.example\fP\&. |
| 1088 |
.SS Batch Files |
| 1089 |
.sp |
| 1090 |
Flent supports reading batch files to automate running several tests and |
| 1091 |
do setup/teardown of test environment etc. This greatly aids |
| 1092 |
reproducibility of tests. |
| 1093 |
.sp |
| 1094 |
The batch file is parsed as an ini file, and can have three types of sections: |
| 1095 |
batches, commands and args. Each section also has a name; type and name are |
| 1096 |
separated with two colons. \(aqBatches\(aq are named tests that can be selected from |
| 1097 |
the command line, \(aqcommands\(aq are system commands to be run before or after each |
| 1098 |
test run, and \(aqargs\(aq are used in the looping mechanism (which allows repeating |
| 1099 |
tests multiple times with different parameters). |
| 1100 |
.sp |
| 1101 |
Variables in sections control the operation of Flent and can be modified in |
| 1102 |
several ways: Sections of the same type can inherit from each other and the |
| 1103 |
variables in an \(aqarg\(aq section will be interpolated into the batch definition on |
| 1104 |
each iteration of a loop. In addition, variable contents can be substituted into |
| 1105 |
other variables by using the ${varname} syntax. These three operations are |
| 1106 |
resolved in this order (inheritance, arg interpolation and variable |
| 1107 |
substitution). |
| 1108 |
.sp |
| 1109 |
An annotated example batchfile is distributed with the source code, and is by |
| 1110 |
default installed to \fB/usr/share/doc/flent/batchfile.example\fP\&. |
| 1111 |
.SH THE DATA FILE FORMAT |
| 1112 |
.sp |
| 1113 |
The aggregated test data is saved in a file called |
| 1114 |
\fB<test_name>\-<date>.<title>.flent.gz\fP (the title part is omitted if no title is |
| 1115 |
specified by the \fB\-t\fP parameter). This file contains the data points |
| 1116 |
generated during the test, as well as some metadata. |
| 1117 |
.SS The top\-level object keys |
| 1118 |
.INDENT 0.0 |
| 1119 |
.TP |
| 1120 |
.B version |
| 1121 |
The file format version as an integer. |
| 1122 |
.UNINDENT |
| 1123 |
.INDENT 0.0 |
| 1124 |
.TP |
| 1125 |
.B x_values |
| 1126 |
An array of the x values for the test data (typically the time |
| 1127 |
values for timeseries data). |
| 1128 |
.UNINDENT |
| 1129 |
.INDENT 0.0 |
| 1130 |
.TP |
| 1131 |
.B results |
| 1132 |
A JSON object containing the result data series. The keys are the |
| 1133 |
data series names; the value for each key is an array of y values |
| 1134 |
for that data series. The data array has the same length as the |
| 1135 |
\fBx_values\fP array, but there may be missing data points |
| 1136 |
(signified by null values). |
| 1137 |
.UNINDENT |
| 1138 |
.INDENT 0.0 |
| 1139 |
.TP |
| 1140 |
.B metadata |
| 1141 |
An object containing various data points about the test run. The |
| 1142 |
metadata values are read in as configuration parameters when the |
| 1143 |
data set is loaded in for further processing. Not all tests use all |
| 1144 |
the parameters, but they are saved anyway. |
| 1145 |
.UNINDENT |
| 1146 |
.INDENT 0.0 |
| 1147 |
.TP |
| 1148 |
.B raw_values |
| 1149 |
An array of objects for each data series. Each element of the array |
| 1150 |
contains the raw values as parsed from the test tool corresponding |
| 1151 |
to that data series. |
| 1152 |
.UNINDENT |
| 1153 |
.SS Metadata keys |
| 1154 |
.INDENT 0.0 |
| 1155 |
.TP |
| 1156 |
.B NAME |
| 1157 |
The test name. |
| 1158 |
.UNINDENT |
| 1159 |
.INDENT 0.0 |
| 1160 |
.TP |
| 1161 |
.B TITLE |
| 1162 |
Any extra title specified by the \fB\-\-title\-extra\fP parameter |
| 1163 |
when the test was run. |
| 1164 |
.UNINDENT |
| 1165 |
.INDENT 0.0 |
| 1166 |
.TP |
| 1167 |
.B HOSTS |
| 1168 |
List of the server hostnames connected to during the test. |
| 1169 |
.UNINDENT |
| 1170 |
.INDENT 0.0 |
| 1171 |
.TP |
| 1172 |
.B LOCAL_HOST |
| 1173 |
The hostname of the machine that ran the test. |
| 1174 |
.UNINDENT |
| 1175 |
.INDENT 0.0 |
| 1176 |
.TP |
| 1177 |
.B LENGTH |
| 1178 |
Test length in seconds, as specified by the \fB\-\-length\fP parameter. |
| 1179 |
.UNINDENT |
| 1180 |
.INDENT 0.0 |
| 1181 |
.TP |
| 1182 |
.B TOTAL_LENGTH |
| 1183 |
Actual data series length, after the test has added time to the |
| 1184 |
\fBLENGTH\fP\&. |
| 1185 |
.UNINDENT |
| 1186 |
.INDENT 0.0 |
| 1187 |
.TP |
| 1188 |
.B STEP_SIZE |
| 1189 |
Time step size granularity. |
| 1190 |
.UNINDENT |
| 1191 |
.INDENT 0.0 |
| 1192 |
.TP |
| 1193 |
.B TIME |
| 1194 |
ISO timestamp of the time the test was initiated. |
| 1195 |
.UNINDENT |
| 1196 |
.INDENT 0.0 |
| 1197 |
.TP |
| 1198 |
.B NOTE |
| 1199 |
Arbitrary text as entered with the \fB\-\-note\fP switch when the |
| 1200 |
test was run. |
| 1201 |
.UNINDENT |
| 1202 |
.INDENT 0.0 |
| 1203 |
.TP |
| 1204 |
.B FLENT_VERSION |
| 1205 |
Version of Flent that generated the data file. |
| 1206 |
.UNINDENT |
| 1207 |
.INDENT 0.0 |
| 1208 |
.TP |
| 1209 |
.B IP_VERSION |
| 1210 |
IP version used to run test (as specified by command line |
| 1211 |
parameters, or auto\-detected from \fBgetaddrinfo()\fP if unspecified). |
| 1212 |
.UNINDENT |
| 1213 |
.INDENT 0.0 |
| 1214 |
.TP |
| 1215 |
.B KERNEL_NAME |
| 1216 |
The kernel name as reported by \fBuname \-s\fP\&. |
| 1217 |
.UNINDENT |
| 1218 |
.INDENT 0.0 |
| 1219 |
.TP |
| 1220 |
.B KERNEL_RELEASE |
| 1221 |
The kernel release as reported by \fBuname \-r\fP\&. |
| 1222 |
.UNINDENT |
| 1223 |
.INDENT 0.0 |
| 1224 |
.TP |
| 1225 |
.B MODULE_VERSIONS |
| 1226 |
The sha1sum of certain interesting Linux kernel modules, if |
| 1227 |
available. Can be used to match test data to specific code versions, |
| 1228 |
if the kernel build is instrumented to, e.g., set the build ID to a |
| 1229 |
git revision. |
| 1230 |
.UNINDENT |
| 1231 |
.INDENT 0.0 |
| 1232 |
.TP |
| 1233 |
.B SYSCTLS |
| 1234 |
The values of several networking\-related sysctls on the host (if |
| 1235 |
available; Linux only). |
| 1236 |
.UNINDENT |
| 1237 |
.INDENT 0.0 |
| 1238 |
.TP |
| 1239 |
.B EGRESS_INFO |
| 1240 |
Interface name, qdisc, offload, driver and BQL configuration of the |
| 1241 |
interface used to reach the test target. This requires that the |
| 1242 |
\fBip\fP binary is present on Linux, but can be extracted from |
| 1243 |
\fBroute\fP on BSD. Qdisc information requires the |
| 1244 |
\fBtc\fP binary to be present, and offload information requires |
| 1245 |
\fBethtool\fP\&. |
| 1246 |
.sp |
| 1247 |
If the \fB\-\-remote\-metadata\fP is used, the extended metadata |
| 1248 |
info is gathered for each of the hostnames specified. This is |
| 1249 |
gathered under the \fBREMOTE_METADATA\fP key in the metadata |
| 1250 |
object, keyed by the hostname values passed to |
| 1251 |
\fB\-\-remote\-metadata\fP \&. Additionally, the |
| 1252 |
\fBREMOTE_METADATA\fP object will contain an object called |
| 1253 |
\fBINGRESS_INFO\fP which is a duplicate of |
| 1254 |
\fBEGRESS_INFO\fP, but with the destination IP exchanged for the |
| 1255 |
source address of the host running flent. The assumption here is |
| 1256 |
that \fB\-\-remote\-metadata\fP is used to capture metadata of a |
| 1257 |
router known to be in the test path, in which case |
| 1258 |
\fBINGRESS_INFO\fP will contain information about the reverse |
| 1259 |
path from the router (which is ingress from the point of view of the |
| 1260 |
host running flent). If the host being queried for remote metadata |
| 1261 |
is off the path, the contents of \fBINGRESS_INFO\fP will |
| 1262 |
probably be the same as that of \fBEGRESS_INFO\fP . |
| 1263 |
.UNINDENT |
| 1264 |
.SS Extended metadata |
| 1265 |
.sp |
| 1266 |
If the \fB\-\-extended\-metadata\fP switch is turned on, the following |
| 1267 |
additional values are collected and stored (to the extent they are available |
| 1268 |
from the platform): |
| 1269 |
.INDENT 0.0 |
| 1270 |
.TP |
| 1271 |
.B IP_ADDRS |
| 1272 |
IP addresses assigned to the machine running flent. |
| 1273 |
.UNINDENT |
| 1274 |
.INDENT 0.0 |
| 1275 |
.TP |
| 1276 |
.B GATEWAYS |
| 1277 |
IP addresses of all detected default gateways on the system, and the |
| 1278 |
interfaces they are reachable through. Only available if the |
| 1279 |
\fBnetstat\fP binary is present on the system. |
| 1280 |
.UNINDENT |
| 1281 |
.INDENT 0.0 |
| 1282 |
.TP |
| 1283 |
.B EGRESS_INFO |
| 1284 |
In the \fBEGRESS_INFO\fP key, the IP address of the next\-hop |
| 1285 |
router and the interface MAC address are added if extended metadata |
| 1286 |
is enabled. |
| 1287 |
.UNINDENT |
| 1288 |
.SH OUTPUT FORMATS |
| 1289 |
.sp |
| 1290 |
The following output formats are currently supported by Flent: |
| 1291 |
.SS Plot output (\fB\-f plot\fP ) |
| 1292 |
.sp |
| 1293 |
Output test data as one of a series of graphical plots of timeseries data or |
| 1294 |
summarised as a CDF plot. Each test supplies a number of different plots; the |
| 1295 |
list of plots for a given test is output by the \fB\-\-list\-plots\fP switch |
| 1296 |
(which must be supplied along with a test name). |
| 1297 |
.sp |
| 1298 |
The plots are drawn by matplotlib, and can be displayed on the screen |
| 1299 |
interactively (requires a graphical display), or output to a file in svg, pdf, |
| 1300 |
ps and png formats. Using the \fB\-o\fP switch turns out file output (the |
| 1301 |
file format is inferred from the file name), while not supplying the switch |
| 1302 |
turns on the interactive plot viewer. |
| 1303 |
.SS Tabulated output (\fB\-f csv\fP and \fB\-f org_table\fP ) |
| 1304 |
.sp |
| 1305 |
These formats output the numeric data in a tabulated format to be consumed by |
| 1306 |
other applications. The \fIcsv\fP output format is a comma\-separated output that can |
| 1307 |
be imported into e.g. spreadsheets, while \fIorg_table\fP outputs a tabulated output |
| 1308 |
in the table format supported by Emacs org mode. The data is output in text |
| 1309 |
format to standard output, or written to a file if invoked with the \fB\-o\fP |
| 1310 |
parameter. |
| 1311 |
.SS Statistics output (\fB\-f stats\fP ) |
| 1312 |
.sp |
| 1313 |
This output format outputs various statistics about the test data, such as total |
| 1314 |
bandwidth consumed, and various statistical measures (min/max/mean/median/std |
| 1315 |
dev/variance) for each data source specified in the relevant test (this can |
| 1316 |
include some data sources not includes on plots). The data is output in text |
| 1317 |
format to standard output, or written to a file if invoked with the \fB\-o\fP |
| 1318 |
parameter. |
| 1319 |
.SS Metadata output (\fB\-f metadata\fP ) |
| 1320 |
.sp |
| 1321 |
This output format outputs the test metadata as pretty\-printed JSON (also |
| 1322 |
suitable for human consumption). It is output as a list of objects, where each |
| 1323 |
object corresponds to the metadata of one test. Mostly useful for inspecting |
| 1324 |
metadata of stored data files. |
| 1325 |
.SH MISC INFO |
| 1326 |
.SS Running tests with the D\-ITG tool |
| 1327 |
.sp |
| 1328 |
This version of flent has experimental support for running and parsing the |
| 1329 |
output of the D\-ITG test tool (see |
| 1330 |
\%<http://\:traffic\:.comics\:.unina\:.it/\:software/\:ITG/>). Flent supports parsing the |
| 1331 |
one\-way delay as measured by D\-ITG. However, in order to do so, the data needs |
| 1332 |
to be collected at the receiver end, statistics extracted, and the result passed |
| 1333 |
back to flent on the sending side. |
| 1334 |
.sp |
| 1335 |
To perform this function, flent supports a control server which will listen to |
| 1336 |
XML\-RPC requests, spawn an appropriate ITGRecv instance and, after the test is |
| 1337 |
done, parse its output and make it available for flent to retrieve. This control |
| 1338 |
server is available as a Python file that by default is installed in |
| 1339 |
\fB/usr/share/doc/flent/misc\fP\&. It currently requires a patched version of |
| 1340 |
D\-ITG v2.8.1. The patch is also included in the same directory. |
| 1341 |
.sp |
| 1342 |
Note that the D\-ITG server is finicky and not designed with security in mind. |
| 1343 |
For this reason, the control server includes HMAC authentication to only allow |
| 1344 |
authenticated clients to run a test against the server; however there is |
| 1345 |
currently no support for enforcement of this in e.g. firewall rules. Please bear |
| 1346 |
this in mind when running a publicly reachable ITGRecv instance (with or without |
| 1347 |
the control server). Another security issue with the control server is that the |
| 1348 |
Python XML\-RPC library by default is vulnerable to XML entity expansion attacks. |
| 1349 |
For this reason, it is highly recommended to install the \fBdefusedxml\fP |
| 1350 |
library (available at \%<https://\:pypi\:.python\:.org/\:pypi/\:defusedxml/>) on the host |
| 1351 |
running the control server. The server will try to find the library on startup |
| 1352 |
and refuse to run if it is not available, unless explicitly told otherwise. |
| 1353 |
.sp |
| 1354 |
Due to the hassle of using D\-ITG, it is recommended to install \fBirtt\fP |
| 1355 |
instead and use that for VoIP tests. |
| 1356 |
.SS Bugs |
| 1357 |
.sp |
| 1358 |
Under some conditions (such as severe bufferbloat), the UDP RTT measurements |
| 1359 |
done by netperf can experience packet loss to the extent that the test aborts |
| 1360 |
completely, which can cause missing data points for some measurement series. |
| 1361 |
The \-\-socket\-timeout feature can alleviate this, but requires a recent SVN |
| 1362 |
version of netperf to work. Flent tries to detect if netperf supports this |
| 1363 |
option and enables it for the UDP measurements if it does. Using \fBirtt\fP |
| 1364 |
for UDP measurements is a way to alleviate this; Flent will automatically detect |
| 1365 |
the availability of irtt and use it if available. |
| 1366 |
.sp |
| 1367 |
Probably many other bugs. Please report any found to |
| 1368 |
\%<https://\:github\:.com/\:tohojo/\:flent/\:issues> and include the output of \fBflent |
| 1369 |
\-\-version\fP in the report. A debug log (as obtained with \fBflent |
| 1370 |
\-\-log\-file\fP ) is also often useful. |
| 1371 |
.SH Author |
| 1372 |
Toke Høiland-Jørgensen |
| 1373 |
.SH Copyright |
| 1374 |
2012-2024, Toke Høiland-Jørgensen and contributors. Source code is GPLv3. Documentation is CC-BY-SA. Neither license affects any output produced by Flent |
| 1375 |
.\" End of generated man page. |