1IPERF(1) User Manuals IPERF(1)
2
6 iperf - perform network throughput tests
7
9 iperf -s [options]
10 iperf -c server [options]
12 iperf -u -s [options]
14 iperf -u -c server [options]
16
19 iperf is a tool for performing network throughput measurements. It can
20 test either TCP or UDP throughput. To perform an iperf test the user
21 must establish both a server (to discard traffic) and a client (to gen‐
22 erate traffic).
23
25 -b, --bandwidth
26 set the target bandwidth and optional standard devation per
27 <mean>,[<stdev>] (See NOTES for suffixes)
28 -e, --enhanced
30 Display enhanced output in reports otherwise use legacy report
31 (ver 2.0.5) formatting (see notes)
32 -f, --format [abkmgBKMG]
34 format to report: adaptive, bits, Bytes, Kbits, Mbits, Gbits,
35 KBytes, MBytes, GBytes (see NOTES for more)
36 -h, --help
38 print a help synopsis
39 -i, --interval n
41 pause n seconds between periodic bandwidth reports
42 -l, --len n[kmKM]
44 set read/write buffer size (TCP) or length (UDP) to n (TCP
45 default 128K, UDP default 1470)
46 --l2checks
48 perform layer 2 length checks on received UDP packets (requires
49 systems that support packet sockets, e.g. Linux)
50 -m, --print_mss
52 print TCP maximum segment size (MTU - TCP/IP header)
53 -o, --output filename
55 output the report or error message to this specified file
56 -p, --port n
58 set server port to listen on/connect to to n (default 5001)
59 -u, --udp
61 use UDP rather than TCP
62 -w, --window n[kmKM]
64 TCP window size (socket buffer size)
65 -z, --realtime
67 Request real-time scheduler, if supported.
68 -B, --bind host
70 bind to host, ip address or multicast address and optional port
71 (see notes)
72 -C, --compatibility
74 for use with older versions does not sent extra msgs
75 -M, --mss n
77 set TCP maximum segment size (MTU - 40 bytes)
78 -N, --nodelay
80 set TCP no delay, disabling Nagle's Algorithm
81 -v, --version
83 print version information and quit
84 -x, --reportexclude [CDMSV]
86 exclude C(connection) D(data) M(multicast) S(settings) V(server)
87 reports
88 -y, --reportstyle C|c
90 if set to C or c report results as CSV (comma separated values)
91
93 -b, --bandwidth n[kmgKMG]
94 set target read rate to n bits/sec. TCP only for the server.
95 -s, --server
97 run in server mode
98 --udp-histogram[=binwidth[u],bincount,[lowerci],[upperci]]
100 output UDP latency histograms, bin width (default 1 millisecond,
101 append u for microseconds,) bincount is total bins (default
102 1000), ci is confidence interval between 0-100% (default lower
103 5%, upper 95%)
104 -B, --bind ip | ip%device
106 bind src ip addr and optional src device for receiving
107 -D, --daemon
109 run the server as a daemon. On Windows this will run the speci‐
110 fied command-line under the IPerfService, installing the service
111 if necessary. Note the service is not configured to auto-start
112 or restart - if you need a self-starting service you will need
113 to create an init script or use Windows "sc" commands.
114 -H, --ssm-host host
116 Set the source host (ip addr) per SSM multicast, i.e. the S of
117 the S,G
118 -R, --remove
120 remove the IPerfService (Windows only).
121 -U, --single_udp
123 run in single threaded UDP mode
124 -V, --ipv6_domain
126 Enable IPv6 reception by setting the domain and socket to
127 AF_INET6 (Can receive on both IPv4 and IPv6)
128
130 -b, --bandwidth n[kmgKMG] | npps
131 set target bandwidth to n bits/sec (default 1 Mbit/sec) or n
132 packets per sec. This may be used with TCP or UDP. For vari‐
133 able loads use format mean,standard deviation
134 -c, --client host
136 run in client mode, connecting to host
137 -d, --dualtest
139 Do a bidirectional test simultaneously.
140 --fq-rate n[kmgKMG]
142 Set a rate to be used with fair-queueing based socket-level pac‐
143 ing, in bytes or bits per second. Only available on platforms
144 supporting the SO_MAX_PACING_RATE socket option. (Note: Here the
145 suffixes indicate bytes/sec or bits/sec per use of uppercase or
146 lowercase, respectively)
147 --incr-dstip
149 increment the destination ip address when using the parallel
150 (-P) option
151 --ipg n
153 set the interpacket gap to n (units of milliseconds) for packets
154 within an isochronous frame (burst), requires --isochronous
155 --isochronous[=fps:mean,stdev]
157 send isochronous traffic with frequency frames per second and
158 load defined by mean and standard deviation using a log normal
159 distribution, defaults to 60:20m,0. (Note: Here the suffixes
160 indicate bytes/sec or bits/sec per use of uppercase or lower‐
161 case, respectively)
162 -n, --num n[kmKM]
164 number of bytes to transmit (instead of -t)
165 -r, --tradeoff
167 Do a bidirectional test individually - client-to-server, fol‐
168 lowed by a reversed test, server-to-client
169 -t, --time n
171 time in seconds to listen for new traffic connections, receive
172 traffic or transmit traffic (Defaults: transmit is 10 secs while
173 listen and receive are indefinite)
174 --trip-time
176 request the server to report the total trip time, i.e from the
177 client's 3WHS done to client's (fin, fin-ack or socket close)
178 (requires synchronized clocks)
179 --txstart-time n.n
181 set the txstart-time to n.n using unix or epoch time format
182 (supports nanonsecond resolution, e.g 1536014418.839992457)
183 -B, --bind ip | ip:port | ipv6 -V | [ipv6]:port -V
185 bind src ip addr and optional port as the source of traffic (see
186 notes)
187 -F, --fileinput name
189 input the data to be transmitted from a file
190 -I, --stdin
192 input the data to be transmitted from stdin
193 -L, --listenport n
195 port to recieve bidirectional tests back on
196 -P, --parallel n
198 number of parallel client threads to run
199 -R, --reverse
201 reverse the traffic flow after header exchange, useful for test‐
202 ing through firewalls
203 -S, --tos
205 set the socket's IP_TOS (byte) field
206 -T, --ttl n
208 time-to-live, for multicast (default 1) -V, --ipv6_domain Set
209 the domain to IPv6 (send packets over IPv6)
210 -X, --peerdetect
212 run server version detection prior to traffic.
213 -Z, --linux-congestion algo
215 set TCP congestion control algorithm (Linux only)
216
218 TCP tests (client)
219 iperf -c <host> -e -i 1
221 ------------------------------------------------------------
222 Client connecting to <host>, TCP port 5001 with pid 5149
223 Write buffer size: 128 KByte
224 TCP window size: 340 KByte (default)
225 ------------------------------------------------------------
226 [ 3] local 45.56.85.133 port 49960 connected with 45.33.58.123 port
227 5001 (ct=3.23 ms)
228 [ ID] Interval Transfer Bandwidth Write/Err Rtry
229 Cwnd/RTT NetPwr
230 [ 3] 0.00-1.00 sec 126 MBytes 1.05 Gbits/sec 1006/0 0
231 56K/626 us 210636.47
232 [ 3] 1.00-2.00 sec 138 MBytes 1.15 Gbits/sec 1100/0 299
233 483K/3884 us 37121.32
234 [ 3] 2.00-3.00 sec 137 MBytes 1.15 Gbits/sec 1093/0 24
235 657K/5087 us 28162.31
236 [ 3] 3.00-4.00 sec 126 MBytes 1.06 Gbits/sec 1010/0 284
237 294K/2528 us 52366.58
238 [ 3] 4.00-5.00 sec 117 MBytes 980 Mbits/sec 935/0 373
239 487K/2025 us 60519.66
240 [ 3] 5.00-6.00 sec 144 MBytes 1.20 Gbits/sec 1149/0 2
241 644K/3570 us 42185.36
242 [ 3] 6.00-7.00 sec 126 MBytes 1.06 Gbits/sec 1011/0 112
243 582K/5281 us 25092.56
244 [ 3] 7.00-8.00 sec 110 MBytes 922 Mbits/sec 879/0 56
245 279K/1957 us 58871.89
246 [ 3] 8.00-9.00 sec 127 MBytes 1.06 Gbits/sec 1014/0 46
247 483K/3372 us 39414.89
248 [ 3] 9.00-10.00 sec 132 MBytes 1.11 Gbits/sec 1054/0 0
249 654K/3380 us 40872.75
250 [ 3] 0.00-10.00 sec 1.25 GBytes 1.07 Gbits/sec 10251/0 1196
251 -1K/3170 us 42382.03
252 where (per -e,)
255 ct= TCP connect time (or three way handshake time 3WHS)
256 Write/Err Total number of successful socket writes. Total number
257 of non-fatal socket write errors
258 Rtry Total number of TCP retries
259 Cwnd/RTT (*nix only) TCP congestion window and round trip time
260 (sampled)
261 NetPwr (*nix only) Network power defined as (throughput / RTT)
262 TCP tests (server)
265 iperf -s -e -i 1 -l 8K
267 ------------------------------------------------------------
268 Server listening on TCP port 5001 with pid 13430
269 Read buffer size: 8.00 KByte
270 TCP window size: 85.3 KByte (default)
271 ------------------------------------------------------------
272 [ 4] local 45.33.58.123 port 5001 connected with 45.56.85.133 port
273 49960
274 [ ID] Interval Transfer Bandwidth Reads
275 Dist(bin=1.0K)
276 [ 4] 0.00-1.00 sec 124 MBytes 1.04 Gbits/sec 22249
277 798:2637:2061:767:2165:1563:589:11669
278 [ 4] 1.00-2.00 sec 136 MBytes 1.14 Gbits/sec 24780
279 946:3227:2227:790:2427:1888:641:12634
280 [ 4] 2.00-3.00 sec 137 MBytes 1.15 Gbits/sec 24484
281 1047:2686:2218:810:2195:1819:728:12981
282 [ 4] 3.00-4.00 sec 126 MBytes 1.06 Gbits/sec 20812
283 863:1353:1546:614:1712:1298:547:12879
284 [ 4] 4.00-5.00 sec 117 MBytes 984 Mbits/sec 20266
285 769:1886:1828:589:1866:1350:476:11502
286 [ 4] 5.00-6.00 sec 143 MBytes 1.20 Gbits/sec 24603
287 1066:1925:2139:822:2237:1827:744:13843
288 [ 4] 6.00-7.00 sec 126 MBytes 1.06 Gbits/sec 22635
289 834:2464:2249:724:2269:1646:608:11841
290 [ 4] 7.00-8.00 sec 110 MBytes 921 Mbits/sec 21107
291 842:2437:2747:592:2871:1903:496:9219
292 [ 4] 8.00-9.00 sec 126 MBytes 1.06 Gbits/sec 22804
293 1038:1784:2639:656:2738:1927:573:11449
294 [ 4] 9.00-10.00 sec 133 MBytes 1.11 Gbits/sec 23091
295 1088:1654:2105:710:2333:1928:723:12550
296 [ 4] 0.00-10.02 sec 1.25 GBytes 1.07 Gbits/sec 227306
297 9316:22088:21792:7096:22893:17193:6138:120790
298 where (per -e,)
300 Reads Total number of socket reads
301 Dist(bin=size) Eight bin histogram of the socket reads returned
302 byte count. Bin width is set per size. Bins are separated by a
303 colon. In the example, the bins are 0-1K, 1K-2K, .., 7K-8K.
304 UDP tests (client)
307 iperf -c <host> -e -i 1 -u -b 10m
309 ------------------------------------------------------------
310 Client connecting to <host>, UDP port 5001 with pid 5169
311 Sending 1470 byte datagrams, IPG target: 1176.00 us (kalman adjust)
312 UDP buffer size: 208 KByte (default)
313 ------------------------------------------------------------
314 [ 3] local 45.56.85.133 port 32943 connected with 45.33.58.123 port
315 5001
316 [ ID] Interval Transfer Bandwidth Write/Err PPS
317 [ 3] 0.00-1.00 sec 1.19 MBytes 10.0 Mbits/sec 852/0 851 pps
318 [ 3] 1.00-2.00 sec 1.19 MBytes 10.0 Mbits/sec 850/0 850 pps
319 [ 3] 2.00-3.00 sec 1.19 MBytes 10.0 Mbits/sec 850/0 850 pps
320 [ 3] 3.00-4.00 sec 1.19 MBytes 10.0 Mbits/sec 851/0 850 pps
321 [ 3] 4.00-5.00 sec 1.19 MBytes 10.0 Mbits/sec 850/0 850 pps
322 [ 3] 5.00-6.00 sec 1.19 MBytes 10.0 Mbits/sec 850/0 850 pps
323 [ 3] 6.00-7.00 sec 1.19 MBytes 10.0 Mbits/sec 851/0 850 pps
324 [ 3] 7.00-8.00 sec 1.19 MBytes 10.0 Mbits/sec 850/0 850 pps
325 [ 3] 8.00-9.00 sec 1.19 MBytes 10.0 Mbits/sec 851/0 850 pps
326 [ 3] 0.00-10.00 sec 11.9 MBytes 10.0 Mbits/sec 8504/0 850 pps
327 [ 3] Sent 8504 datagrams
328 [ 3] Server Report:
329 [ 3] 0.00-10.00 sec 11.9 MBytes 10.0 Mbits/sec 0.047 ms 0/ 8504
330 (0%) 0.537/ 0.392/23.657/ 0.497 ms 850 pps 2329.37
331 where (per -e,)
333 Write/Err Total number of successful socket writes. Total number
334 of non-fatal socket write errors
335 PPS Transmit packet rate in packets per second
336 UDP tests (server)
339 iperf -s -e -i 1 -u
341 ------------------------------------------------------------
342 Server listening on UDP port 5001 with pid 13496
343 Receiving 1470 byte datagrams
344 UDP buffer size: 208 KByte (default)
345 ------------------------------------------------------------ [ 3]
346 local 45.33.58.123 port 5001 connected with 45.56.85.133 port 32943
347 [ ID] Interval Transfer Bandwidth Jitter Lost/Total
348 Latency avg/min/max/stdev PPS NetPwr
349 [ 3] 0.00-1.00 sec 1.19 MBytes 10.0 Mbits/sec 0.057 ms 0/ 851
350 (0%) 0.475/ 0.408/ 1.898/ 0.090 ms 851 pps 2633.56
351 [ 3] 1.00-2.00 sec 1.19 MBytes 10.0 Mbits/sec 0.039 ms 0/ 851
352 (0%) 0.669/ 0.405/16.256/ 1.375 ms 850 pps 1869.32
353 [ 3] 2.00-3.00 sec 1.19 MBytes 10.0 Mbits/sec 0.038 ms 0/ 850
354 (0%) 0.795/ 0.395/23.657/ 2.138 ms 850 pps 1572.05
355 [ 3] 3.00-4.00 sec 1.19 MBytes 10.0 Mbits/sec 0.045 ms 0/ 850
356 (0%) 0.475/ 0.403/ 3.477/ 0.148 ms 850 pps 2628.58
357 [ 3] 4.00-5.00 sec 1.19 MBytes 10.0 Mbits/sec 0.043 ms 0/ 851
358 (0%) 0.463/ 0.400/ 1.458/ 0.068 ms 850 pps 2699.88
359 [ 3] 5.00-6.00 sec 1.19 MBytes 10.0 Mbits/sec 0.032 ms 0/ 850
360 (0%) 0.486/ 0.404/ 2.658/ 0.154 ms 850 pps 2572.21
361 [ 3] 6.00-7.00 sec 1.19 MBytes 10.0 Mbits/sec 0.055 ms 0/ 850
362 (0%) 0.469/ 0.404/ 2.768/ 0.108 ms 850 pps 2664.82
363 [ 3] 7.00-8.00 sec 1.19 MBytes 10.0 Mbits/sec 0.039 ms 0/ 851
364 (0%) 0.571/ 0.400/12.452/ 0.855 ms 850 pps 2192.68
365 [ 3] 8.00-9.00 sec 1.19 MBytes 10.0 Mbits/sec 0.083 ms 0/ 850
366 (0%) 0.475/ 0.392/ 3.702/ 0.196 ms 850 pps 2628.29
367 [ 3] 9.00-10.00 sec 1.19 MBytes 10.0 Mbits/sec 0.047 ms 0/ 850
368 (0%) 0.493/ 0.396/ 6.010/ 0.343 ms 850 pps 2534.89
369 [ 3] 0.00-10.00 sec 11.9 MBytes 10.0 Mbits/sec 0.047 ms 0/ 8504
370 (0%) 0.537/ 0.392/23.657/ 0.867 ms 850 pps 2329.37
371 where (per -e,)
374 Latency End to end latency in mean/minimum/maximum/standard
375 deviation format (Note: requires the client's and server's sys‐
376 tem clocks to be synchronized to a common reference, e.g. using
377 precision time protocol PTP. A GPS disciplined OCXO is a recom‐
378 mended reference.)
379 PPS Received packet rate in packets per second
380 NetPwr Network power defined as (throughput / latency)
381 Isochronous UDP tests (client)
384 iperf -c 192.168.100.33 -u -e -i 1 --isochronous=60:100m,10m --realtime
386 ------------------------------------------------------------
387 Client connecting to 192.168.100.33, UDP port 5001 with pid 14971
388 UDP isochronous: 60 frames/sec mean= 100 Mbit/s, stddev=10.0 Mbit/s,
389 Period/IPG=16.67/0.005 ms
390 UDP buffer size: 208 KByte (default)
391 ------------------------------------------------------------
392 [ 3] local 192.168.100.76 port 42928 connected with 192.168.100.33
393 port 5001
394 [ ID] Interval Transfer Bandwidth Write/Err PPS
395 frames:tx/missed/slips
396 [ 3] 0.00-1.00 sec 12.0 MBytes 101 Mbits/sec 8615/0 8493 pps
397 62/0/0
398 [ 3] 1.00-2.00 sec 12.0 MBytes 100 Mbits/sec 8556/0 8557 pps
399 60/0/0
400 [ 3] 2.00-3.00 sec 12.0 MBytes 101 Mbits/sec 8586/0 8586 pps
401 60/0/0
402 [ 3] 3.00-4.00 sec 12.1 MBytes 102 Mbits/sec 8687/0 8687 pps
403 60/0/0
404 [ 3] 4.00-5.00 sec 11.8 MBytes 99.2 Mbits/sec 8468/0 8468 pps
405 60/0/0
406 [ 3] 5.00-6.00 sec 11.9 MBytes 99.8 Mbits/sec 8519/0 8520 pps
407 60/0/0
408 [ 3] 6.00-7.00 sec 12.1 MBytes 102 Mbits/sec 8694/0 8694 pps
409 60/0/0
410 [ 3] 7.00-8.00 sec 12.1 MBytes 102 Mbits/sec 8692/0 8692 pps
411 60/0/0
412 [ 3] 8.00-9.00 sec 11.9 MBytes 100 Mbits/sec 8537/0 8537 pps
413 60/0/0
414 [ 3] 9.00-10.00 sec 11.8 MBytes 99.0 Mbits/sec 8450/0 8450 pps
415 60/0/0
416 [ 3] 0.00-10.01 sec 120 MBytes 100 Mbits/sec 85867/0 8574 pps
417 602/0/0
418 [ 3] Sent 85867 datagrams
419 [ 3] Server Report:
420 [ 3] 0.00-9.98 sec 120 MBytes 101 Mbits/sec 0.009 ms 196/85867
421 (0.23%) 0.665/ 0.083/ 1.318/ 0.174 ms 8605 pps 18903.85
422 where (per -e,)
424 frames:tx/missed/slips Total number of isochronous frames or
425 bursts. Total number of frame ids not sent. Total number of
426 frame slips
427 Isochronous UDP tests (server)
430 iperf -s -e -u --udp-histogram=100u,2000 --realtime
432 ------------------------------------------------------------
433 Server listening on UDP port 5001 with pid 5175
434 Receiving 1470 byte datagrams
435 UDP buffer size: 208 KByte (default)
436 ------------------------------------------------------------
437 [ 3] local 192.168.100.33 port 5001 connected with 192.168.100.76 port
438 42928 isoch (peer 2.0.13-alpha)
439 [ ID] Interval Transfer Bandwidth Jitter Lost/Total
440 Latency avg/min/max/stdev PPS NetPwr Frames/Lost
441 [ 3] 0.00-9.98 sec 120 MBytes 101 Mbits/sec 0.010 ms 196/85867
442 (0.23%) 0.665/ 0.083/ 1.318/ 0.284 ms 8585 pps 18903.85 601/1
443 [ 3] 0.00-9.98 sec T8(f)-PDF:
444 bin(w=100us):cnt(85671)=1:2,2:844,3:10034,4:8493,5:8967,6:8733,7:8823,8:9023,9:8901,10:8816,11:7730,12:4563,13:741,14:1
445 (5.00/95.00%=3/12,Outliers=0,obl/obu=0/0)
446 [ 3] 0.00-9.98 sec F8(f)-PDF:
447 bin(w=100us):cnt(598)=15:2,16:1,17:27,18:68,19:125,20:136,21:103,22:83,23:22,24:23,25:5,26:3
448 (5.00/95.00%=17/24,Outliers=0,obl/obu=0/0)
449 where, Frames/lost Total number of frames (or bursts) received. Total
452 number of bursts lost or errored
453 T8-PDF(f) Latency histogram for packets
454 F8-PDF(f) Latency histogram for frames
455
459 Note: The environment variable option settings haven't been maintained
460 well. See the source code if these are of interest.
461
463 Some numeric options support format characters per '<value>c' (e.g.
464 10M) where the c format characters are k,m,g,K,M,G. Lowercase format
465 characters are 10^3 based and uppercase are 2^n based, e.g. 1k = 1000,
466 1K = 1024, 1m = 1,000,000 and 1M = 1,048,576
467 The -b option supports variable offered loads through the <mean>,<stan‐
469 dard deviation> format, e.g. -b 100m,10m on the client. The distribu‐
470 tion used is log normal. Similar for the isochronous option.
471 The -e or --enhanced latency output on the UDP servers assumes the
473 clients' and servers' system clocks are synchronized. Network Time
474 Protocol (NTP) or Precision Time Protocol (PTP) are commonly used for
475 this. The reference clock(s) or oscillator's error will also affect
476 the accuracy of UDP latency measurements.
477 The -B option affects the bind() system call. This is typically used
479 to bind to a particular IP address. Only packets destined to that IP
480 address will be received while any transmitted packets will carry that
481 IP address as their source. The bind() does not control anything about
482 the routing of transmitted packets. So, for example, if the IP address
483 of eth0 is used for -B and the routing table for the destination IP
484 address (per -c) resolves the ouput interface to be eth1, then the host
485 will send the packet out device eth1 with the source IP address of
486 eth0. To affect the physical output interface (e.g. dual homed sys‐
487 tems) the host's routing table(s) need to be configured, e.g. configure
488 policy routing per each -B source address.
489 The TCP connect time (or three way handshake) can be seen on the iperf
491 client when the -e (--enhancedreports) option is set. Look for the
492 ct=<value> in the connected message, e.g.in '[ 3] local 192.168.1.4
493 port 48736 connected with 192.168.1.1 port 5001 (ct=1.84 ms)' shows the
494 3WHS took 1.84 milliseconds.
495 The network power (NetPwr) metric is experimental. It's a convenience
497 function defined as throughput/delay. For TCP, the delay is the sam‐
498 pled RTT times. For UDP the delay is the end/end latency. Don't con‐
499 fuse this with the physics definition of power (delta energy/delta
500 time) but more of a measure of a desireable property divided by an
501 undesireable property. Also note, one must use -i interval with TCP to
502 get this as that's what sets the RTT sampling rate. The metric is
503 scaled to assist with human readability. (Note: if this metric goes
504 beyond the experimental state we'll consider a supporting and RTT sam‐
505 pling rate independent of the -i interval.)
506
508 This section needs to be filled in.
509
511 See https://sourceforge.net/p/iperf2/tickets/
512
514 Iperf2, based from iperf (originally written by Mark Gates and Alex
515 Warshavsky), has a goal of maintainence with some feature enhancement.
516 Other contributions from Ajay Tirumala, Jim Ferguson, Jon Dugan <jdugan
517 at x1024 dot net>, Feng Qin, Kevin Gibbs, John Estabrook <jestabro at
518 ncsa.uiuc.edu>, Andrew Gallatin <gallatin at gmail.com>, Stephen Hem‐
519 minger <shemminger at linux-foundation.org>, Tim Auckland <tim.auckland
520 at gmail.com>, Robert J. McMahon <rjmcmahon at rjmcmahon.com>
521
523 http://sourceforge.net/projects/iperf2/
524NLANR/DAST APRIL 2008 IPERF(1)