TPLink Archer C2 Internet speed slower since upgrade from 22.03.5 to 23.05.2

[robdavies]

Hi

I have a TPLink Archer C2 v1 router. I used to run openwrt 22.03.5 and got 650Mb download (checked with speedtest.net) with my cable broadband. I upgraded to 23.05.2 and the download reduced to 220Mb (same configuration/settings).

I installed 22.03.5 to see if it was the upgrade, its back to 650Mb download again.

Is there something I can do or configure with 23.05.2 to increase the speed or diagnose? I already have the flow offloading enabled. I checked. Could it be that this upgrade be too much for my router? I'm not sure how I can find that out.

Thanks,
Rob

[DGdodo]

robdavies,

Best way to check differences is to watch both kernel and system logs.
Maybe the newer firmware needs some other adjustments / settings (maybe due to DSA network changes?).

Better way to check speeds is locally, as external speed tests can differ every time you run those.
Package iperf3 is a tool to test speeds locally (between PC and router, or between routers).

Check how much free memory the overview page of OpenWrt displays, gives a roughly indication of free memory on both versions.
On cli command: free

DG.

[robdavies]

@DGdodo

Fantastic, thank you, I will run these checks/tests on 22.05.5 and 23.03.2 to see how each run on my router.

Rob

[munkey]

Did you ever manage to resolve this? I have the same issue on an Archer C7 v5, after upgrade to latest version WAN speed is only a fraction of what it used to be.

Only thing I haven't tried yet is start fresh and reconfigure everything manually.

[robdavies]

Hi @munkey

I haven't resolved this yet,. As @DGdodo suggester, I performed the following tests on fresh installs, with the same config changes and no extra packages installed.

When running idle, CPU usage is the same at around 10%. RAM usage differs.
22.03.5 : 16000 used, 25112 free
23.05.2 : 23032 used 14552 free

LAN <-> LAN has no extra load on RAM or CPU, I think this router is using a switch. I get 940Mb/s before and after the upgrade.

LAN <-> WAN RAM (above idle), CPU usage is up. I couldn't get iperf to work on this test, so I used speedtest.
22.03.5 RAM increase usage to 17500 on load. CPU goes to 100% usage, Using speedtest, I got 655Mb/s, 90% sirq on CPU usage (75% ksoftirqd, 12% kworker)
23.05.2 RAM increase is small on load. CPU goes to 100% usage. Using speedtest I got 250Mb/s, 92% sirq (85% ksoftirqd, 13% kworker).

I didn't see any errors in the kernel or system logs.

I have tried setting up LAN <-> WAN speed test using iperf, I haven't figured out a way to get this working yet.

@DGdodo could it be that the CPU it the limiting factor here, with irq loads (sirq/ksoftirqd)? Its a single CPU with 1 core.

Thanks,
Rob

[robdavies]

@Dgdodo I just found a way to do LAN <-> WAN iperf test, I used a second router.

22.03.6 : iperf3 320Mb/s (which is strange, as I got better results using speedtest)
23.05.2 : iperf3 350Mb/s

I don't know why I was getting 320Mb/s on 22.03.6 using iperf?

How can I check if the firmware need adjustments/settings? How can I check if its migrated to DSA?

Thanks,
Rob

[DGdodo]

robdavies,

The LAN - WAN test is only between 2 devices / routers (the physical LAN on 1, and WAN port of other device). What you mean is LAN - Internet which is in all situations different. (Your first speedtest.)

The use of a second router just slows down the whole link, I guess.
Or you have iperf3 installed on that second router?

I found some External iperf3 servers to test your internet speed.
Maybe better then a speedtest website as that does test: PC - and all network in between - internet

GitHub - R0GGER/public-iperf3-servers: A list of public iPerf3 servers...

Or do the speedtest (more then twice to get an avarage value) with and without your second router?
Also make sure to use the best quality cables, I've seen tests dramatically go down using cheap ethernet-cables. Even I have 1 GBit connection that only runs 100mBit, due to length & quality of the cable.

Locally you can test iperf3 from both sides, if both routers have iperf3 installed.
Testing WAN-LAN and LAN-WAN.

        ------------------            ------------------
       | Router 1         |          | Router 2         |
PC --- | LAN          WAN |----------| LAN          WAN |----- Internet  
       |                  |          |                  |
        ------------------            ------------------

If you have iperf3 installed on your PC it is different!
It will test PC to Router 2 LAN (and not the WAN of Router 2).

I presume you ssh/telnet into Router 1 (with iperf3) and test to iperf3 on Router 2.
iperf3 does only test: WAN ---------- LAN between 2 routers (devices).

If both routers have iperf3 installed, and run from Router 1 (Router 2 is iperf3 server).
A iperf3 test from Router2 (to Router 1) does test: LAN-WAN (used iperf3 port must be open on Router 1).

I have 2 older routers (and a MESH network in between) but the weird thing is that WAN-LAN is : 84Mb/s
The other way around, LAN-WAN is : 176 Mb/s

Testing from Router 2 to Router 1 is in my situation 2 times faster then testing Router 1 to Router 2, which is weird according to me.
As the data must also go through (in my case 2) Firewall(s), the other way around is normally 'open'.

Testing with a large file discovered indeed a difference between up/downloading from PC to NAS, although 23.5MB/s vs 28.5MB/s
This could mean my MESH does a little is this, but not the 50% we see in iperf3 performance.

Maybe you have also to try it both ways?
Don't forget to open the firewall(s), as I was forgotten 1, it took over 1,5 hours to discover

About your q: How can I check if the firmware need adjustments/settings? How can I check if its migrated to DSA?
This should be updated on the device page of OpenWrt: https://openwrt.org/toh/tp-link/archer_c2_ac750

Example in my home situation:

Router 1 - TorRouter      (OpenWrt v22.03.5, 192.168.100.1)
Router 2 - TorRouter-DG03 (OpenWrt v22.03.2, 192.168.1.1)
Mesh Network              (AVM, 192.168.1.254) 

root@TorRouter:~# iperf3 -c 192.168.1.1 -V
iperf 3.11
Linux TorRouter 5.10.176 #0 SMP Thu Apr 27 20:28:15 2023 mips
Control connection MSS 1448
Time: Mon, 11 Mar 2024 13:38:02 UTC
Connecting to host 192.168.1.1, port 5201
      Cookie: cotyrcgmy2mzp6p2wteeptxzqwyxk4tmqmp7
      TCP MSS: 1448 (default)
[  5] local 192.168.117.162 port 52226 connected to 192.168.1.1 port 5201
Starting Test: protocol: TCP, 1 streams, 131072 byte blocks, omitting 0 seconds, 10 second test, tos 0
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  10.4 MBytes  87.3 Mbits/sec    5    154 KBytes       
[  5]   1.00-2.00   sec  9.63 MBytes  80.8 Mbits/sec    0    197 KBytes       
[  5]   2.00-3.01   sec  9.63 MBytes  80.3 Mbits/sec    0    230 KBytes       
[  5]   3.01-4.00   sec  9.76 MBytes  82.3 Mbits/sec    1    192 KBytes       
[  5]   4.00-5.00   sec  10.4 MBytes  87.6 Mbits/sec    0    230 KBytes       
[  5]   5.00-6.00   sec  9.45 MBytes  79.2 Mbits/sec    0    260 KBytes       
[  5]   6.00-7.00   sec  9.82 MBytes  82.4 Mbits/sec    1    215 KBytes       
[  5]   7.00-8.00   sec  10.2 MBytes  85.5 Mbits/sec    2    136 KBytes       
[  5]   8.00-9.00   sec  10.7 MBytes  89.7 Mbits/sec    1    150 KBytes       
[  5]   9.00-10.00  sec  10.9 MBytes  91.2 Mbits/sec    1    150 KBytes       
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec   101 MBytes  84.6 Mbits/sec   11             sender
[  5]   0.00-10.01  sec   100 MBytes  84.1 Mbits/sec                  receiver
CPU Utilization: local/sender 17.3% (0.5%u/16.8%s), remote/receiver 72.9% (7.2%u/65.6%s)
snd_tcp_congestion cubic
rcv_tcp_congestion cubic

root@TorRouter-DG03:~# iperf3 -c 192.168.1.254 -V
iperf 3.11
Linux TorRouter-DG03 5.10.146 #0 SMP Fri Oct 14 22:44:41 2022 mips
Control connection MSS 1448
Time: Mon, 11 Mar 2024 13:40:18 UTC
Connecting to host 192.168.1.254, port 5201
      Cookie: arha572kq3jemefgszvk5vjuzreq54waaupg
      TCP MSS: 1448 (default)
[  5] local 192.168.1.1 port 44272 connected to 192.168.1.254 port 5201
Starting Test: protocol: TCP, 1 streams, 131072 byte blocks, omitting 0 seconds, 10 second test, tos 0
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  20.6 MBytes   172 Mbits/sec    0    352 KBytes       
[  5]   1.00-2.00   sec  21.2 MBytes   179 Mbits/sec    0    403 KBytes       
[  5]   2.00-3.05   sec  22.5 MBytes   179 Mbits/sec    0    458 KBytes       
[  5]   3.05-4.04   sec  21.2 MBytes   181 Mbits/sec    0    458 KBytes       
[  5]   4.04-5.06   sec  20.6 MBytes   171 Mbits/sec    1    484 KBytes       
[  5]   5.06-6.04   sec  21.2 MBytes   181 Mbits/sec    0    484 KBytes       
[  5]   6.04-7.02   sec  21.2 MBytes   181 Mbits/sec    0    484 KBytes       
[  5]   7.02-8.01   sec  21.2 MBytes   181 Mbits/sec    0    484 KBytes       
[  5]   8.01-9.00   sec  21.2 MBytes   180 Mbits/sec    0    484 KBytes       
[  5]   9.00-10.03  sec  21.2 MBytes   174 Mbits/sec    0    484 KBytes       
- - - - - - - - - - - - - - - - - - - - - - - - -
Test Complete. Summary Results:
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.03  sec   212 MBytes   178 Mbits/sec    1             sender
[  5]   0.00-10.04  sec   212 MBytes   178 Mbits/sec                  receiver
CPU Utilization: local/sender 97.2% (0.1%u/97.1%s), remote/receiver 82.4% (3.5%u/78.9%s)
snd_tcp_congestion cubic
rcv_tcp_congestion cubic

My external iperf3 test result in Europe (speedtest.fra1.de.leaseweb.net):

root@TorRouter:~# iperf3 -c speedtest.fra1.de.leaseweb.net -p 5201-5210
Connecting to host speedtest.fra1.de.leaseweb.net, port 5201
[  5] local 192.168.117.162 port 57504 connected to 178.162.226.176 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  3.93 MBytes  33.0 Mbits/sec    7    146 KBytes       
[  5]   1.00-2.00   sec  3.67 MBytes  30.8 Mbits/sec    2    117 KBytes       
[  5]   2.00-3.00   sec  3.54 MBytes  29.7 Mbits/sec    0    139 KBytes       
[  5]   3.00-4.00   sec  3.73 MBytes  31.3 Mbits/sec    2    113 KBytes       
[  5]   4.00-5.00   sec  3.54 MBytes  29.7 Mbits/sec    0    136 KBytes       
[  5]   5.00-6.00   sec  3.67 MBytes  30.8 Mbits/sec    0    153 KBytes       
[  5]   6.00-7.00   sec  3.73 MBytes  31.3 Mbits/sec    1    130 KBytes       
[  5]   7.00-8.00   sec  3.60 MBytes  30.2 Mbits/sec    0    147 KBytes       
[  5]   8.00-9.00   sec  3.60 MBytes  30.2 Mbits/sec    2    120 KBytes       
[  5]   9.00-10.00  sec  3.79 MBytes  31.8 Mbits/sec    0    141 KBytes       
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  36.8 MBytes  30.9 Mbits/sec   14             sender
[  5]   0.00-10.05  sec  36.4 MBytes  30.4 Mbits/sec                  receiver

iperf Done.

DG.

[robdavies]

@DGdodo Thanks for your help, I will follow up with this.

Rob

[steinhof5]

I seem to be having a similar issue on Archer C7 v4 on 23.05.2.

root@OpenWrt:~# iperf3 -c a209.speedtest.wobcom.de
Connecting to host a209.speedtest.wobcom.de, port 5201
[ 5] local 2a02:2f0e:550d:d100::1 port 33580 connected to 2a01:581:8::15 port 5201
[ ID] Interval Transfer Bitrate Retr Cwnd
[ 5] 0.00-1.00 sec 11.6 MBytes 97.2 Mbits/sec 0 1.96 MBytes
[ 5] 1.00-2.00 sec 15.4 MBytes 129 Mbits/sec 0 1.96 MBytes
[ 5] 2.00-3.00 sec 15.4 MBytes 129 Mbits/sec 0 1.96 MBytes
[ 5] 3.00-4.00 sec 14.8 MBytes 123 Mbits/sec 0 1.96 MBytes
[ 5] 4.00-5.00 sec 15.1 MBytes 127 Mbits/sec 0 1.96 MBytes
[ 5] 5.00-6.00 sec 15.4 MBytes 129 Mbits/sec 0 1.96 MBytes
[ 5] 6.00-7.00 sec 15.2 MBytes 128 Mbits/sec 0 1.96 MBytes
[ 5] 7.00-8.00 sec 15.6 MBytes 131 Mbits/sec 0 1.96 MBytes
[ 5] 8.00-9.00 sec 14.8 MBytes 124 Mbits/sec 0 1.96 MBytes
[ 5] 9.00-10.00 sec 15.5 MBytes 130 Mbits/sec 0 1.96 MBytes

---

[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 149 MBytes 125 Mbits/sec 0 sender
[ 5] 0.00-10.05 sec 149 MBytes 124 Mbits/sec receiver

On stock TP-Link I was getting 600-700+ Mbps.

The result is with Software flow offloading enabled.

[Klingon]

Take a look here

Distribute DSA across 2 different cores MR8300 SNAPSHOT

Hey Yall, I've been looking at upgrading to 910Mbps Fiber and Im worried that I will have to move back to stock firmware to get the speeds I'm paying for. At the moment I know that the current DSA implementation is limited to 1 core for wan to lan connections which means that speed is limited from wan to lan... I have an MR8300 which is running OpenWrt 23.05-SNAPSHOT r23263-b7fac1bf8c. I got about 400Mbps when using iperf3 from wan to my pc, but while doing iperf3 from my pc to the router I …

The same is happening in other devices, is related to DSA.

[robdavies]

Thanks @Klingon , I will keep it on 22.03.5 until I replace my router.