1: iperf3: A TCP, UDP, and SCTP network bandwidth measurement tool
2: ================================================================
3:
4: Summary
5: -------
6:
7: iperf is a tool for active measurements of the maximum achievable
8: bandwidth on IP networks. It supports tuning of various parameters
9: related to timing, protocols, and buffers. For each test it reports
10: the bandwidth, loss, and other parameters.
11:
12: This version, sometimes referred to as iperf3, is a redesign of an
13: original version developed at NLANR/DAST. iperf3 is a new
14: implementation from scratch, with the goal of a smaller, simpler code
15: base, and a library version of the functionality that can be used in
16: other programs. iperf3 also a number of features found in other tools
17: such as nuttcp and netperf, but were missing from the original iperf.
18: These include, for example, a zero-copy mode and optional JSON output.
19: Note that iperf3 is *not* backwards compatible with the original iperf.
20:
21: Primary development for iperf3 takes place on CentOS Linux, FreeBSD,
22: and MacOS X. At this time, these are the only officially supported
23: platforms, however there have been some reports of success with
24: OpenBSD, NetBSD, Android, Solaris, and other Linux distributions.
25:
26: iperf3 is principally developed by ESnet / Lawrence Berkeley National
27: Laboratory. It is released under a three-clause BSD license.
28:
29: Note that at this point, ESnet plans to support iperf3 in "maintenance
30: mode". At this point, there are no definite plans for further iperf3
31: releases, and ESnet will be providing a very limited amount of
32: resources for support and development, going forward. However, ESnet
33: could issue new iperf3 releases to deal with security issues or
34: high-impact bug fixes.
35:
36: For more information see: http://software.es.net/iperf
37:
38: Source code and issue tracker: https://github.com/esnet/iperf
39:
40: Obtaining iperf3
41: ----------------
42:
43: Downloads of iperf3 are available at:
44:
45: http://downloads.es.net/pub/iperf/
46:
47: To check out the most recent code, clone the git repository at:
48:
49: https://github.com/esnet/iperf.git
50:
51: Building iperf3
52: ---------------
53:
54: ### Prerequisites: ###
55:
56: None.
57:
58: ### Building ###
59:
60: ./configure; make; make install
61:
62: (Note: If configure fails, try running `./bootstrap.sh` first)
63:
64: Invoking iperf3
65: ---------------
66:
67: iperf3 includes a manual page listing all of the command-line options.
68: The manual page is the most up-to-date reference to the various flags and parameters.
69:
70: For sample command line usage, see:
71:
72: http://fasterdata.es.net/performance-testing/network-troubleshooting-tools/iperf-and-iperf3/
73:
74: Using the default options, iperf is meant to show typical well
75: designed application performance. 'Typical well designed application'
76: means avoiding artificial enhancements that work only for testing
77: (such as splice()'ing the data to /dev/null). iperf does also have
78: flags for 'extreme best case' optimizations but they must be
79: explicitly activated.
80:
81: These flags include:
82:
83: -Z, --zerocopy use a 'zero copy' sendfile() method of sending data
84: -A, --affinity n/n,m set CPU affinity
85:
86: Bug Reports
87: -----------
88:
89: Before submitting a bug report, please make sure you're running the
90: latest version of the code, and confirm that your issue has not
91: already been fixed. Then submit to the iperf3 issue tracker on
92: GitHub:
93:
94: https://github.com/esnet/iperf/issues
95:
96: In your issue submission, please indicate the version of iperf3 and
97: what platform you're trying to run on (provide the platform
98: information even if you're not using a supported platform, we
99: *might* be able to help anyway). Exact command-line arguments will
100: help us recreate your problem. If you're getting error messages,
101: please include them verbatim if possible, but remember to sanitize any
102: sensitive information.
103:
104: If you have a question about usage or about the code, please do *not*
105: submit an issue. Please use one of the mailing lists for that.
106:
107: Changes from iperf 2.x
108: ----------------------
109:
110: New options:
111:
112: -V, --verbose more detailed output than before
113: -J, --json output in JSON format
114: -Z, --zerocopy use a 'zero copy' sendfile() method of sending data
115: -O, --omit N omit the first n seconds (to ignore slowstart)
116: -T, --title str prefix every output line with this string
117: -F, --file name xmit/recv the specified file
118: -A, --affinity n/n,m set CPU affinity (Linux and FreeBSD only)
119: -k, --blockcount #[KMG] number of blocks (packets) to transmit (instead
120: of -t or -n)
121: -L, --flowlabel set IPv6 flow label (Linux only)
122:
123: Changed flags:
124:
125: -C, --linux-congestion set congestion control algorithm (Linux only)
126: (-Z in iperf2)
127:
128:
129: Deprecated options:
130:
131: Not planning to support these iperf2 flags. If you really miss these
132: options, please submit a request in the issue tracker:
133:
134: -d, --dualtest Do a bidirectional test simultaneously
135: -r, --tradeoff Do a bidirectional test individually
136: -T, --ttl time-to-live, for multicast (default 1)
137: -x, --reportexclude [CDMSV] exclude C(connection) D(data) M(multicast)
138: S(settings) V(server) reports
139: -y, --reportstyle C report as a Comma-Separated Values
140:
141: Also deprecated is the ability to set the options via environment
142: variables.
143:
144: Known Issues
145: ------------
146:
147: * UDP performance: Some problems have been noticed with iperf3 on the
148: ESnet 100G testbed at high UDP rates (above 10Gbps). The symptom is
149: that on any particular run of iperf3 the receiver reports a loss rate
150: of about 20%, regardless of the -b option used on the client side.
151: This problem appears not to be iperf3-specific, and may be due to the
152: placement of the iperf3 process on a CPU and its relation to the
153: inbound NIC. In some cases this problem can be mitigated by an
154: appropriate use of the CPU affinity (-A) option. (Issue #55)
155:
156: * The -Z flag sometimes causes the iperf3 client to hang on OSX.
157: (Issue #129)
158:
159: * When specifying the socket buffer size using the "-w" flag on Linux, Linux
160: doubles the value you pass in. (You can see this using iperf3's debug flag).
161: But then the CWND does not actually ramp up to the doubled value, but only
162: to about 75% of the doubled value. This appears to be by design.
163:
164: * Although the "-w" flag is documented as setting the (TCP) window
165: size, it is also used to set the socket buffer size. This has been
166: shown to be helpful with high-bitrate UDP tests.
167:
168: * On some platforms, it might be necessary to invoke "ldconfig"
169: manually after doing a "make install" before the iperf3 executable can
170: find its shared library. (Issue #153)
171:
172: * The results printed on the server side at the end of a test do not
173: correctly reflect the client-side measurements. This is due to the
174: ordering of computing and transferring results between the client
175: and server. (Issue #293)
176:
177: * The server could have a very short measurement reporting interval at
178: the end of a test (particularly a UDP test), containing few or no
179: packets. This issue is due to an artifact of timing between the
180: client and server. (Issue #278)
181:
182: Links
183: -----
184:
185: This section lists links to user-contributed Web pages regarding
186: iperf3. ESnet and Lawrence Berkeley National Laboratory bear no
187: responsibility for the content of these pages.
188:
189: * Installation instructions for Debian Linux (by Cameron Camp
190: <cameron@ivdatacenter.com>):
191:
192: http://cheatsheet.logicalwebhost.com/iperf-network-testing/
193:
194: Copyright
195: ---------
196:
197: iperf, Copyright (c) 2014-2016, The Regents of the University of
198: California, through Lawrence Berkeley National Laboratory (subject
199: to receipt of any required approvals from the U.S. Dept. of
200: Energy). All rights reserved.
201:
202: If you have questions about your rights to use or distribute this
203: software, please contact Berkeley Lab's Technology Transfer
204: Department at TTD@lbl.gov.
205:
206: NOTICE. This software is owned by the U.S. Department of Energy.
207: As such, the U.S. Government has been granted for itself and others
208: acting on its behalf a paid-up, nonexclusive, irrevocable,
209: worldwide license in the Software to reproduce, prepare derivative
210: works, and perform publicly and display publicly. Beginning five
211: (5) years after the date permission to assert copyright is obtained
212: from the U.S. Department of Energy, and subject to any subsequent
213: five (5) year renewals, the U.S. Government is granted for itself
214: and others acting on its behalf a paid-up, nonexclusive,
215: irrevocable, worldwide license in the Software to reproduce,
216: prepare derivative works, distribute copies to the public, perform
217: publicly and display publicly, and to permit others to do so.
218:
219: This code is distributed under a BSD style license, see the LICENSE
220: file for complete information.
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to README.md CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / iperf