@rygle @giuliomagnifico
Thank you for the help. I will do that, a clean build ........
There is an open bugreport, asking for 160mhz to be reactivated. It was disabled for a few chipsets sometime ago.
All right I would also like to report that I successfully update to the snapshot. For my configuration I only had to reinstall luci and rsync. Everything is still working fine. Great work, thank you. I think slowly I understand much more how everything works.
@carload Now I think if in future you did an attended sysupgrade from the system menu, it will integrate both luci and rsync into the firmware.
The main reason you would want to integrate packages/modules into the firmware (read only) is to save space in the read/write storage area, which is an overlay on top of the firmware. In some routers there is not much storage at all, so the r/w area runs out very quickly, so it's better to pack it into the firmware to save that space. However, if you are only adding in a few packages, and especially because this router has a decent amount of r/w space, it is not as important. If you look at the overview page it will tell you how much space you have free.
Thanks for this good explanation, interesting, I did not know that and now I understand why many build their own firmware
Hey guys.
I have a request for anyone who can and is willing to help me debug something.
I am looking for someone who can check out 21.02.5, apply system config and required patches to get it to build for the WAX206 and supply me with the factory and sysupgrade images.
The reason is this thread: Unable to connect to outside internet with Dumb AP coniguration - #17 by xNUTx
I am running into exactly the same issue as the OP and by setting up my EX6120 with OpenWRT 21.02.5 the problem is gone. I want to make sure I'm not crazy and would like to test this on my WAX206 as well, but I have no build environment for tinkering and testing/retesting.
Does any one of you have a WAX206 and a pfSense or OPNsense router as main dhcp/dns/gateway?
pretty sure the RTL WAN chip wasn't supported by the Linux version used in 21.02.
I don't mind testing it with a newer version of the kernel, as long as it gets configured pretty much the same way as it would have for the 5.4 kernel, including the defaults throughout as if it is a 5.4 kernel getting built.
The thing is, I suspect it is a kernel setting causing this behavior, not a kernel feature or an OpenWRT package as such. I have NO idea if it would even be possible to mingle the 21.02.5 OpenWRT version with a 5.10 kernel though. Again, I do not have any experiencing building images, I have no clue if what am asking is even a remote posibility.
Adding this link here, as I am really not the only one affected, but I do need to find a solution (whatever that may be, except replacing the pfSense appliance).
At present I am running Opnsense on an hp t620 plus with a solarflare 7122 card as main router with dhcp/dns/vlans/etc assignments coming from there. I have an icx7250 as my main switch (lots of vlans and some poe). I've been running an Xiaomi Redmi AX6S for ~10 months as my wireless ap. It uses the mt7622 cpu as the wax206. I've been using SNAPSHOT r19787-e3e9eb31a2 (5.15.45) for a while now on and I've never seen this issue like this. I did not follow the dumb ap guide, I turned off the firewall, dhcp, and dns services and just plugged into lan. I left the Wan interface as is, it's just down.
I'm planning swap in the wax206 now that 2.5gb is working properly. I have been testing it with a laptop and android phone and have not seen this issue either.
Strange i did an attended Upgrade not the first time but after Reboot it was the netgear Firmware.
Any other here with the same problem?
Weird I got this error when I try to compile the latest master
make[3]: Leaving directory '/home/opc/wax206/openwrt/feeds/luci/modules/luci-base'
time: package/feeds/luci/luci-base/host-compile#1.38#0.20#1.86
make[2]: Leaving directory '/home/opc/wax206/openwrt'
make[1]: *** [package/Makefile:110: /home/opc/wax206/openwrt/staging_dir/target-aarch64_cortex-a53_musl/stamp/.package_compile] Error 2
make[1]: *** Waiting for unfinished jobs....
make[4] scripts/config/conf
make[2]: Leaving directory '/home/opc/wax206/openwrt'
make[1]: Leaving directory '/home/opc/wax206/openwrt'
make: *** [/home/opc/wax206/openwrt/include/toplevel.mk:231: world] Error 2
Thanks for sharing this as well as to @Amadeus for the PR.
Be advised though that I've noticed a drop in speeds despite a 2.5G link being active - see attached screenie. Same 2.5G port on the WAX206 was used, just switched around the ports on the switch (GS110emx). Link was auto negotiated in both tests and no anomalies with ports 1 to 4 on the 206 were noticed.
So the retry count is high and speeds drop in half with a 2.5G link active when compared to a 1G active link!
Can someone confirm my findings?
Thanks
W.
I did a test in router mode some days ago and was able to get line speed.
To test it again:
- What OpenWrt revision/commit did you work with?
- WAX206 In AP or Router mode?
- The WAN port is the only one connected/with a route to the iperf3 server?
I'm suspecting it could be an intel (mac pro) vs mediatek (wax206) not playing nice and all but there's a switch in between..
To answer your questions:
-
Latest / master, pulled a couple of hours ago and compiled with the same patch you worked with here
-
Client bridge mode - AP (Asus) > WAX206 with relayd > Switch (Netgear) > MP (Intel) running iperf3
(Tried testing the 2.5G wan port with the snapshot release but couldn't get an active link not even 1G -
Correct, no other devices other than the switch is connected to the wax.
I also tried fiddling around with the switch's settings to see if it was perhaps some odd setting enabled - nothing!
Hmmm... Just tried switching around iperf3 (now running in server mode on the wax206 with an active 2.5G link)
Strange!
In addition, I also tried taking the mac out of the equation and tested with my synology nas (iperf3 running in server mode on the nas) - the speeds are normal but notice the high retry counts
root@Bridge:~# iperf3 -c 10.10.10.10
Connecting to host 10.10.10.10, port 5201
[ 5] local 10.10.10.252 port 35142 connected to 10.10.10.10 port 5201
[ ID] Interval Transfer Bitrate Retr Cwnd
[ 5] 0.00-1.00 sec 112 MBytes 944 Mbits/sec 617 165 KBytes
[ 5] 1.00-2.00 sec 112 MBytes 940 Mbits/sec 512 130 KBytes
[ 5] 2.00-3.00 sec 111 MBytes 934 Mbits/sec 506 139 KBytes
[ 5] 3.00-4.00 sec 112 MBytes 939 Mbits/sec 584 100 KBytes
[ 5] 4.00-5.00 sec 112 MBytes 938 Mbits/sec 638 109 KBytes
[ 5] 5.00-6.00 sec 112 MBytes 938 Mbits/sec 513 143 KBytes
[ 5] 6.00-7.00 sec 112 MBytes 940 Mbits/sec 666 132 KBytes
[ 5] 7.00-8.00 sec 111 MBytes 935 Mbits/sec 703 136 KBytes
[ 5] 8.00-9.00 sec 112 MBytes 938 Mbits/sec 469 156 KBytes
[ 5] 9.00-10.00 sec 112 MBytes 939 Mbits/sec 578 137 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 1.09 GBytes 939 Mbits/sec 5786 sender
[ 5] 0.00-10.00 sec 1.09 GBytes 938 Mbits/sec receiver
iperf Done.
root@Bridge:~# ^C
0 retries and normal speeds were reported when iperf3 is running in server mode on the wax206 and/or the wax's WAN port is connected to a 1G port on the switch. So the issue seems to be with a 2.5G > 2.5G negotiated link and with iperf3 running in server mode anywhere but on the wax unit.
Both the mac and synology nas are connected to the switch via LACP/Lags.
Have to you tried with 2 or 4 streams (iperf3 -P 2 ... | iperf3 -P4 ...)? Is LACP meddling with the traffic?
My setup is simpler but all I can do for now: Intel NUC 8i7 > MikroTik CSS610 > WAX206
Test 1:
- WAX206 in AP mode
- iperf3 server running on the NUC 8i7 with a 10Gbit SONNET TB3 adapter
- Both the NUC and WAX206 are connected to 10Gbit SFPs, NUC by fiber and WAX206 by RJ-45 adapter
- WAX WAN in 2.5G mode
root@OpenWrt:~# iperf3 -c 192.168.169.11 -p 10555
Connecting to host 192.168.169.11, port 10555
[ 5] local 192.168.169.81 port 53428 connected to 192.168.169.11 port 10555
[ ID] Interval Transfer Bitrate Retr Cwnd
[ 5] 0.00-1.00 sec 282 MBytes 2.37 Gbits/sec 0 981 KBytes
[ 5] 1.00-2.00 sec 279 MBytes 2.34 Gbits/sec 0 1.00 MBytes
[ 5] 2.00-3.00 sec 280 MBytes 2.35 Gbits/sec 0 1.06 MBytes
[ 5] 3.00-4.00 sec 280 MBytes 2.35 Gbits/sec 0 1.06 MBytes
[ 5] 4.00-5.00 sec 280 MBytes 2.35 Gbits/sec 0 1.11 MBytes
[ 5] 5.00-6.00 sec 280 MBytes 2.35 Gbits/sec 0 1.11 MBytes
[ 5] 6.00-7.00 sec 280 MBytes 2.35 Gbits/sec 0 1.11 MBytes
[ 5] 7.00-8.00 sec 280 MBytes 2.35 Gbits/sec 0 1.11 MBytes
[ 5] 8.00-9.00 sec 280 MBytes 2.35 Gbits/sec 0 1.11 MBytes
[ 5] 9.00-10.00 sec 277 MBytes 2.33 Gbits/sec 0 1.24 MBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 2.73 GBytes 2.35 Gbits/sec 0 sender
[ 5] 0.00-10.05 sec 2.73 GBytes 2.33 Gbits/sec receiver
iperf Done.
Test 2:
- as above but WAX206 connected to a 1G port on the MikroTik
root@OpenWrt:~# iperf3 -c 192.168.169.11 -p 10555
Connecting to host 192.168.169.11, port 10555
[ 5] local 192.168.169.81 port 47018 connected to 192.168.169.11 port 10555
[ ID] Interval Transfer Bitrate Retr Cwnd
[ 5] 0.00-1.00 sec 114 MBytes 954 Mbits/sec 0 580 KBytes
[ 5] 1.00-2.00 sec 113 MBytes 944 Mbits/sec 0 580 KBytes
[ 5] 2.00-3.00 sec 112 MBytes 940 Mbits/sec 0 608 KBytes
[ 5] 3.00-4.00 sec 112 MBytes 939 Mbits/sec 0 608 KBytes
[ 5] 4.00-5.00 sec 113 MBytes 945 Mbits/sec 0 608 KBytes
[ 5] 5.00-6.00 sec 112 MBytes 940 Mbits/sec 0 747 KBytes
[ 5] 6.00-7.00 sec 112 MBytes 940 Mbits/sec 0 747 KBytes
[ 5] 7.00-8.00 sec 113 MBytes 945 Mbits/sec 0 747 KBytes
[ 5] 8.00-9.00 sec 112 MBytes 938 Mbits/sec 0 747 KBytes
[ 5] 9.00-10.00 sec 113 MBytes 944 Mbits/sec 0 747 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 1.10 GBytes 943 Mbits/sec 0 sender
[ 5] 0.00-10.05 sec 1.10 GBytes 937 Mbits/sec receiver
iperf Done.
Test 3:
- replace the NUC 8i7 with a MacbookPro 2018, connected via 1G using a TB3 hub, to the MikroTik
MBP (TB3 Hub) > MikroTik CSS610 > WAX206
root@OpenWrt:~# iperf3 -c 192.168.169.152 -p 10555
Connecting to host 192.168.169.152, port 10555
[ 5] local 192.168.169.81 port 52362 connected to 192.168.169.152 port 10555
[ ID] Interval Transfer Bitrate Retr Cwnd
[ 5] 0.00-1.00 sec 113 MBytes 947 Mbits/sec 0 488 KBytes
[ 5] 1.00-2.00 sec 113 MBytes 943 Mbits/sec 0 515 KBytes
[ 5] 2.00-3.00 sec 112 MBytes 937 Mbits/sec 0 515 KBytes
[ 5] 3.00-4.00 sec 112 MBytes 942 Mbits/sec 0 570 KBytes
[ 5] 4.00-5.00 sec 112 MBytes 936 Mbits/sec 0 570 KBytes
[ 5] 5.00-6.00 sec 112 MBytes 939 Mbits/sec 0 570 KBytes
[ 5] 6.00-7.00 sec 112 MBytes 941 Mbits/sec 0 570 KBytes
[ 5] 7.00-8.00 sec 112 MBytes 941 Mbits/sec 0 570 KBytes
[ 5] 8.00-9.00 sec 112 MBytes 936 Mbits/sec 0 570 KBytes
[ 5] 9.00-10.00 sec 112 MBytes 941 Mbits/sec 0 570 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 1.09 GBytes 940 Mbits/sec 0 sender
[ 5] 0.00-10.01 sec 1.09 GBytes 939 Mbits/sec receiver
iperf Done.
Looks fine to me, I don't really know what to make of it.
Thanks for carrying out these tests and sharing your results. I'll run additional tests and report back.
Do you mind sharing with me your .config? Any special general or kernel optimizations?
Cheers
W.
I fired up my Synology DS920+ which is connected by LACP (Balance-TCP mode) to the MikroTik switch and ran iperf3 server from it, result in 1G line speed but lots of retries. That would suggest that LACP is causing an issue. And I would have expected more than 1G since the iperf3 clients where WAX206 2.5G port and the 10G NUC.
The internet answers for a query about "iperf3 lacp" with:
- iperf3 issue "Link aggregation testing"
- Why am I seeing retransmissions across the network using iperf3?
root@OpenWrt:~# iperf3 -c 192.168.169.20 -p 10777
Connecting to host 192.168.169.20, port 10777
[ 5] local 192.168.169.81 port 47636 connected to 192.168.169.20 port 10777
[ ID] Interval Transfer Bitrate Retr Cwnd
[ 5] 0.00-1.00 sec 112 MBytes 941 Mbits/sec 573 96.2 KBytes
[ 5] 1.00-2.00 sec 111 MBytes 935 Mbits/sec 522 129 KBytes
[ 5] 2.00-3.00 sec 112 MBytes 940 Mbits/sec 406 115 KBytes
[ 5] 3.00-4.00 sec 111 MBytes 934 Mbits/sec 522 126 KBytes
[ 5] 4.00-5.00 sec 112 MBytes 940 Mbits/sec 462 122 KBytes
[ 5] 5.00-6.00 sec 111 MBytes 935 Mbits/sec 499 130 KBytes
[ 5] 6.00-7.00 sec 112 MBytes 936 Mbits/sec 433 132 KBytes
[ 5] 7.00-8.00 sec 112 MBytes 937 Mbits/sec 472 133 KBytes
[ 5] 8.00-9.00 sec 112 MBytes 938 Mbits/sec 434 120 KBytes
[ 5] 9.00-10.00 sec 112 MBytes 937 Mbits/sec 508 188 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 1.09 GBytes 937 Mbits/sec 4831 sender
[ 5] 0.00-10.00 sec 1.09 GBytes 937 Mbits/sec receiver
iperf Done.
And here my config file, no special kernel setting. Only some additional packages like ethtool and vim-full and some more.
my dotconfig diff file
CONFIG_TARGET_mediatek=y
CONFIG_TARGET_mediatek_mt7622=y
CONFIG_TARGET_mediatek_mt7622_DEVICE_netgear_wax206=y
# CONFIG_FEED_routing is not set
# CONFIG_FEED_telephony is not set
CONFIG_PACKAGE_cgi-io=y
CONFIG_PACKAGE_ethtool=y
# CONFIG_PACKAGE_kmod-lib-crc-ccitt is not set
# CONFIG_PACKAGE_kmod-ppp is not set
CONFIG_PACKAGE_libevent2-core=y
CONFIG_PACKAGE_liblucihttp=y
CONFIG_PACKAGE_liblucihttp-ucode=y
CONFIG_PACKAGE_libncurses=y
CONFIG_PACKAGE_libpcap=y
CONFIG_PACKAGE_librt=y
CONFIG_PACKAGE_libsmartcols=y
CONFIG_PACKAGE_lscpu=y
CONFIG_PACKAGE_luci-app-firewall=y
CONFIG_PACKAGE_luci-app-opkg=y
CONFIG_PACKAGE_luci-base=y
CONFIG_PACKAGE_luci-light=y
CONFIG_PACKAGE_luci-mod-admin-full=y
CONFIG_PACKAGE_luci-mod-network=y
CONFIG_PACKAGE_luci-mod-status=y
CONFIG_PACKAGE_luci-mod-system=y
CONFIG_PACKAGE_luci-proto-ipv6=y
CONFIG_PACKAGE_luci-proto-ppp=y
CONFIG_PACKAGE_luci-ssl=y
CONFIG_PACKAGE_luci-theme-bootstrap=y
# CONFIG_PACKAGE_ppp is not set
CONFIG_PACKAGE_px5g-mbedtls=y
CONFIG_PACKAGE_rpcd=y
CONFIG_PACKAGE_rpcd-mod-file=y
CONFIG_PACKAGE_rpcd-mod-iwinfo=y
CONFIG_PACKAGE_rpcd-mod-luci=y
CONFIG_PACKAGE_rpcd-mod-rrdns=y
CONFIG_PACKAGE_rpcd-mod-ucode=y
CONFIG_PACKAGE_tcpdump=y
CONFIG_PACKAGE_terminfo=y
CONFIG_PACKAGE_ucode-mod-html=y
CONFIG_PACKAGE_ucode-mod-math=y
CONFIG_PACKAGE_uhttpd=y
CONFIG_PACKAGE_uhttpd-mod-ubus=y
CONFIG_PACKAGE_vim-full=y
I am running the latest SNAPSHOT build from the attended sysupgrade, I have DAWN working with 802.11k/v/r but for some reason the WAX206 does not show WPA3-SAE as security option, which my EX6120 does show.
Because of DAWN, wpad (full) is installed, so i can not install wpad-openssl or anything that colides with wpad... does anyone else miss the WPA3 option from the official SNAPSHOT builds?
Appreciate your time and results, thanks!
Good to know about lacp and retransmissions in iperf3 which solves 1 part of the mystery ... thing is, I never had this halving of speeds nor retransmissions with the netgear firmware when I had it setup as a bridge.
The other part is the slowdown/halving and I'm suspecting it has something to do with the client bridge mode/setup under openwrt, specifically with relayd. With the netgear firmware, maxing out my 2.5G link caused a minor bump in CPU utilization while with openwrt and relayd under a client bridge setup, the usage was 80~85% for both cores at ~600Mbps which isn't good. This is regardless of whether I had it connected via 1G or 2.5G, but with 1G I had no halving of speeds (iperf3 or otherwise ie. internet usage).
You have your wax setup as AP which doesn't require relayd and aren't binding interfaces in your network except for the 920+ (I have the 918+, great boxes!)...
So what on earth is causing the 2.5G link to halve speeds when lacp is thrown into the mix?
To be continued . . .
Forgot to add: your .config is similar to mine which reminds of another anomaly I noticed - my wifi negotiated speeds are ~2032Mbps but the highest line speed I was able to record was 1.1Gbit ... is this normal?
Another test, this time all wireless AP (.253) <> WAX206 (.252) ... halving and retransmissions are occurring when wax is in client mode.