iperf3.16 UDP test acceptance issue

[hzywhite]

I tested version 3.16 of iperfs on two Windows 11 computers and during the UDP packet filling process
why the receiving end's rate kept showing 0

server:iperf3 -u -c 192.168.43.100 -i 1 -l 1470 -b 200m -t 999
client:iperf3 -s -i 1

[davidBar-On]

@hzywhite, can you give more details? (Note by the way that you tagged the server command line as "client" and visa versa.)

1. Does the server shows receiving rate 0 from the first interval (fist second) or it happens only sometime during the test?
2. When the server receive rate is 0, does the client continue to send the 200Mbits/sec?
3. Can you run both server and client with the additional options -V --debug=3 and include the first part of this output and a part from when the issue happens?

[hzywhite]

warning: UDP block size 1470 exceeds TCP MSS 1460, may result in fragmentation / drops
Connecting to host 192.168.43.102, port 5201
[ 5] local 192.168.43.10 port 50337 connected to 192.168.43.102 port 5201
[ ID] Interval Transfer Bitrate Total Datagrams
[ 5] 0.00-1.00 sec 23.6 MBytes 197 Mbits/sec 16815
[ 5] 1.00-2.01 sec 24.1 MBytes 200 Mbits/sec 17195
[ 5] 2.01-3.01 sec 23.9 MBytes 200 Mbits/sec 17048
[ 5] 3.01-4.01 sec 23.8 MBytes 200 Mbits/sec 16950
[ 5] 4.01-5.00 sec 23.7 MBytes 200 Mbits/sec 16899
[ 5] 5.00-6.02 sec 24.1 MBytes 200 Mbits/sec 17180
[ 5] 6.02-7.01 sec 23.7 MBytes 200 Mbits/sec 16931
[ 5] 7.01-8.01 sec 23.8 MBytes 200 Mbits/sec 16977
[ 5] 8.01-9.01 sec 23.8 MBytes 200 Mbits/sec 16945
[ 5] 9.01-10.01 sec 24.0 MBytes 200 Mbits/sec 17143
[ 5] 10.01-11.00 sec 23.5 MBytes 200 Mbits/sec 16796
[ 5] 11.00-12.01 sec 24.1 MBytes 200 Mbits/sec 17217
[ 5] 12.01-13.01 sec 23.7 MBytes 200 Mbits/sec 16899
[ 5] 13.01-14.00 sec 23.8 MBytes 200 Mbits/sec 16961

[ ID] Interval Transfer Bitrate Jitter Lost/Total Datagrams
[ 5] 0.00-1.00 sec 19.1 MBytes 160 Mbits/sec 0.115 ms 3163/16815 (19%)
[ 5] 1.00-2.00 sec 488 KBytes 4.01 Mbits/sec 0.017 ms 20/360 (5.6%)
[ 5] 2.00-3.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 3.00-4.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 4.00-5.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 5.00-6.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 6.00-7.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 7.00-8.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 8.00-9.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 9.00-10.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 10.00-11.00 sec 0.00 Bytes 0.00 bits/sec 0.017 ms 0/0 (0%)
[ 5] 11.00-12.00 sec 2.65 MBytes 22.2 Mbits/sec 0.029 ms 170483/172370 (99%)
[ 5] 12.00-13.00 sec 0.00 Bytes 0.00 bits/sec 0.029 ms 0/0 (0%)
[ 5] 13.00-14.00 sec 0.00 Bytes 0.00 bits/sec 0.029 ms 0/0 (0%)

[hzywhite]

Regarding the third point, I tested it and found that adding - V to the client made reception appear normal with slight packet loss. At first, I thought it was caused by a system interrupt, so I bound the CPU, but the effect was that it couldn't handle it, resulting in a large amount of packet loss.
It seems that this slight packet loss is related to the priority of the process...

iperf3 -s -i 1 -V
iperf 3.16
CYGWIN_NT-10.0-19043 HG7111007875 3.4.10-1.x86_64 2023-11-29 12:12 UTC x86_64

Server listening on 5201 (test #1)

Time: Tue, 26 Mar 2024 11:15:35 GMT
Accepted connection from 192.168.43.10, port 50311
Cookie: 67s5vqoc3lzid5oxl3i3a3chaoicwo73zkpc
Target Bitrate: 200000000
[ 5] local 192.168.43.102 port 5201 connected to 192.168.43.10 port 50372
Starting Test: protocol: UDP, 1 streams, 1470 byte blocks, omitting 0 seconds, 999 second test, tos 0
[ ID] Interval Transfer Bitrate Jitter Lost/Total Datagrams
[ 5] 0.00-1.00 sec 23.4 MBytes 196 Mbits/sec 0.011 ms 188/16864 (1.1%)
[ 5] 1.00-2.01 sec 23.5 MBytes 196 Mbits/sec 0.009 ms 258/17053 (1.5%)
[ 5] 2.01-3.00 sec 23.6 MBytes 199 Mbits/sec 0.063 ms 90/16918 (0.53%)
[ 5] 3.00-4.00 sec 23.5 MBytes 197 Mbits/sec 0.082 ms 233/16968 (1.4%)
[ 5] 4.00-5.01 sec 23.7 MBytes 198 Mbits/sec 0.087 ms 202/17134 (1.2%)
[ 5] 5.01-6.00 sec 23.6 MBytes 200 Mbits/sec 0.014 ms 144/16988 (0.85%)
[ 5] 6.00-7.00 sec 23.6 MBytes 198 Mbits/sec 0.073 ms 108/16931 (0.64%)
[ 5] 7.00-8.00 sec 23.5 MBytes 198 Mbits/sec 0.074 ms 179/16967 (1.1%)

bound the CPU：
[ ID] Interval Transfer Bitrate Jitter Lost/Total Datagrams
[ 5] 0.00-1.00 sec 23.1 MBytes 193 Mbits/sec 0.071 ms 348/16791 (2.1%)
[ 5] 1.00-2.00 sec 23.6 MBytes 198 Mbits/sec 0.009 ms 205/17033 (1.2%)
[ 5] 2.00-3.00 sec 23.7 MBytes 199 Mbits/sec 0.063 ms 104/17006 (0.61%)
[ 5] 3.00-4.00 sec 23.7 MBytes 199 Mbits/sec 0.070 ms 105/17021 (0.62%)
[ 5] 4.00-5.00 sec 23.8 MBytes 200 Mbits/sec 0.009 ms 100/17068 (0.59%)
[ 5] 5.00-6.00 sec 23.8 MBytes 199 Mbits/sec 0.009 ms 110/17057 (0.64%)
[ 5] 6.00-7.01 sec 23.5 MBytes 196 Mbits/sec 0.075 ms 176/16973 (1%)
[ 5] 7.01-8.00 sec 23.5 MBytes 199 Mbits/sec 0.066 ms 120/16907 (0.71%)
[ 5] 8.00-9.00 sec 23.7 MBytes 199 Mbits/sec 0.073 ms 92/16965 (0.54%)
[ 5] 9.00-10.00 sec 4.55 MBytes 38.2 Mbits/sec 0.152 ms 12516/15760 (79%)
[ 5] 10.00-11.00 sec 2.42 MBytes 20.3 Mbits/sec 0.340 ms 15966/17693 (90%)
[ 5] 11.00-12.00 sec 2.86 MBytes 24.0 Mbits/sec 1.503 ms 15358/17401 (88%)
[ 5] 12.00-13.00 sec 1.73 MBytes 14.5 Mbits/sec 0.206 ms 15483/16716 (93%)
[ 5] 13.00-14.00 sec 4.11 MBytes 34.5 Mbits/sec 0.018 ms 14560/17493 (83%)
[ 5] 14.00-15.01 sec 19.1 MBytes 159 Mbits/sec 0.087 ms 3459/17097 (20%)
[ 5] 15.01-16.00 sec 19.3 MBytes 162 Mbits/sec 0.013 ms 3294/17042 (19%)
[ 5] 16.00-17.01 sec 18.8 MBytes 158 Mbits/sec 0.086 ms 3520/16950 (21%)
[ 5] 17.01-18.00 sec 6.86 MBytes 57.9 Mbits/sec 0.016 ms 1091/5987 (18%)
[ 5] 18.00-19.00 sec 0.00 Bytes 0.00 bits/sec 0.016 ms 0/0 (0%)
[ 5] 19.00-20.00 sec 0.00 Bytes 0.00 bits/sec 0.016 ms 0/0 (0%)
[ 5] 20.00-21.00 sec 0.00 Bytes 0.00 bits/sec 0.016 ms 0/0 (0%)
[ 5] 20.00-21.00 sec 0.00 Bytes 0.00 bits/sec 0.016 ms 0/0 (0%)

[davidBar-On]

This is strange. I don't know of any reason that -V will make a difference. Try adding -R to see if the issue also happens when the server is sending to the client.

Try running a TCP test without bitrate limit to see what is the network throughput and how it behaves.

Other UDP tests that may help understanding the issue changing to -l 1460 (or even -l 512), so no fragmentation will be done, and try using different bit-rates.

Running Wireshark at the network connection points to the client and the server may help to determine whether the client machine is loosing the packets, whether they are lost in the way, or whether the handling in the server is the problem.