Hello everyone!
I have a significative reduction of the throughput on the network between the WAN and the LAN. This topic might look similar to Turris limited network throughput, although that case was solved re-installing PAKON. However, I have nothing installed beyond the basic settings on my Turris Omnia.
A factory-reset fresh and updated Turris Omnia as Modem/Router (FttH using SFP module) with nothing else than the minimal configuration. All the connections are by wire (no wifi at all). Version is 5.3.1
The throughput between my laptop and a public server is 137 Mbps, about a 23% of what I pay for.
C:\>iperf3 -4 -R -O 10 -V -c lille.testdebit.info -p 9234
iperf 3.1.3
CYGWIN_NT-10.0 laptop2.5.1(0.297/5/3) 2016-04-21 22:14 x86_64
Time: Sat, 04 Dec 2021 16:25:12 GMT
Connecting to host lille.testdebit.info, port 9234
Reverse mode, remote host lille.testdebit.info is sending
Cookie: laptop.1638635111.836118
TCP MSS: 0 (default)
[ 4] local 192.168.laptop port 54259 connected to 89.84.1.234 port 9234
Starting Test: protocol: TCP, 1 streams, 131072 byte blocks, omitting 10 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.00 sec 16.3 MBytes 136 Mbits/sec (omitted)
....(many lines repeating the same)
[ 4] 9.00-10.00 sec 16.6 MBytes 139 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth Retr
[ 4] 0.00-10.00 sec 160 MBytes 134 Mbits/sec 108 sender
[ 4] 0.00-10.00 sec 160 MBytes 134 Mbits/sec receiver
CPU Utilization: local/receiver 2.8% (1.7%u/1.1%s), remote/sender 0.0% (0.0%u/0.0%s)
iperf Done.
The same measurement made on the Turris Omnia against the same public server results on a throughput of 600 Mbps, as expected.
root@turris:~# iperf3 -4 -R -O 10 -V -c lille.testdebit.info -p 9234
iperf 3.7
Linux turris 4.14.254 #0 SMP Tue Nov 23 09:52:21 2021 armv7l
Control connection MSS 1440
Time: Sat, 04 Dec 2021 16:31:35 UTC
Connecting to host lille.testdebit.info, port 9234
Reverse mode, remote host lille.testdebit.info is sending
Cookie: dwuujrxnhlyocfb3hnhvrsgovf63mthauebw
TCP MSS: 1440 (default)
[ 5] local xxx.xxx.xxx.xxx (wan IP) port 44792 connected to 89.84.1.234 port 9234
Starting Test: protocol: TCP, 1 streams, 131072 byte blocks, omitting 10 seconds, 10 second test, tos 0
[ ID] Interval Transfer Bitrate
[ 5] 0.00-1.00 sec 72.6 MBytes 609 Mbits/sec (omitted)
....(many lines repeating the same)
[ 5] 9.00-10.00 sec 73.2 MBytes 614 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.04 sec 720 MBytes 602 Mbits/sec 0 sender
[ 5] 0.00-10.00 sec 717 MBytes 602 Mbits/sec receiver
snd_tcp_congestion bbr
rcv_tcp_congestion cubic
iperf Done.
The Turris Omnia’s CPU does not seem to be burned out.
This is how it looks like the laptop is performing the throughput test with the WAN server
top - 17:13:27 up 59 min, 0 users, load average: 0.00, 0.04, 0.00
Tasks: 111 total, 1 running, 110 sleeping, 0 stopped, 0 zombie
%Cpu0 : 0.7/3.9 5[|||| ]
%Cpu1 : 3.9/12.5 16[||||||||||||| ]
GiB Mem : 11.9/2.0 [ ]
GiB Swap: 0.0/0.0 [ ]
An isolated test between the Turris Omnia and my laptop result on 920 Mbps.
C:\>iperf3 -4 -R -O 10 -V -c 192.168.router
iperf 3.1.3
CYGWIN_NT-10.0 laptop 2.5.1(0.297/5/3) 2016-04-21 22:14 x86_64
Time: Sat, 04 Dec 2021 16:37:20 GMT
Connecting to host 192.168.router, port 5201
Reverse mode, remote host 192.168.router is sending
Cookie: laptop.1638635840
TCP MSS: 0 (default)
[ 4] local 192.168.laptop port 53087 connected to 192.168.router port 5201
Starting Test: protocol: TCP, 1 streams, 131072 byte blocks, omitting 10 seconds, 10 second test
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.00 sec 109 MBytes 912 Mbits/sec (omitted)
...... Many lines repeating the same
[ 4] 9.00-10.00 sec 111 MBytes 934 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval Transfer Bandwidth Retr
[ 4] 0.00-10.00 sec 1.08 GBytes 924 Mbits/sec 735 sender
[ 4] 0.00-10.00 sec 1.07 GBytes 919 Mbits/sec receiver
CPU Utilization: local/receiver 27.4% (15.2%u/12.1%s), remote/sender 12.0% (0.1%u/12.0%s)
iperf Done.
speedtest.net and netmetr confirm the same results.
I would like to ask for some help on identifying the cause of this reduced throughput and hopefully, a solution.
Wishing you all the best.
With best regards
Sebastian