[384.16_Alpha Builds] Testing all variants

[Makaveli]

This is a Alpha testing thread am i wrong ?? Bug reports will be posted here..

Are those log entries affecting your usage and performance of the router?

 

[Frankflash]

it's because people sit in front computers all day looking at their router log's and then complain here if something does not look right

 

[Sanna1967]

What I see, it will be repaired at the next build. Correct me

https://github.com/RMerl/asuswrt-merlin.ng/commit/12a946dec73149319a6d4b419fbb3cd61439142c

 

[RMerlin]

Anyone brave enough to test a completely untested RT-AX56U build? (don't have one to test it).

 

[JWoo]

Are those log entries affecting your usage and performance of the router?

I have one of the RT-AX58U units.

The system performance seems not impacted by the excessive entries and as someone posted, there is already a commit from @RMerlin to correct. The main consequence is that the log files fill up on /jffs and cycle to a new file.

One odd thing that I noticed (a Linux thing) is that when I use command "top" via SSH to check the CPU load, the normal Linux load averages look wrong. The RT-AX58U load typically is less then 6% per core.

Using SSH and "top" command and press number "1" to see all the cores individually:

Mem: 215508K used, 296752K free, 1020K shrd, 0K buff, 16076K cached
CPU0: 0.3% usr 1.1% sys 0.0% nic 98.2% idle 0.0% io 0.0% irq 0.1% sirq
CPU1: 0.5% usr 3.7% sys 0.0% nic 94.6% idle 0.0% io 0.0% irq 0.9% sirq
CPU2: 0.0% usr 1.1% sys 0.0% nic 98.8% idle 0.0% io 0.0% irq 0.0% sirq
Load average: 0.46 0.44 0.45 1/125 11189

If I look at all the cores averaged

CPU: 0.5% usr 2.1% sys 0.0% nic 97.0% idle 0.0% io 0.0% irq 0.2% sirq

The 1, 5, 15 minute load averages seem incorrect. The system is not running at 46% load and it is not running at .46% load either. Each core is running at around 6% or less all the time, and the average of all cores combined is less than 3%. So not sure what the 1, 5, 15 minute load averages are showing me.

 

[a5m]

Anyone brave enough to test a completely untested RT-AX56U build? (don't have one to test it).

@Kal-EL you're up.

 

[Makaveli]

@Kal-EL you're up.

That is a different router he has the RT-AX58U

 

[a5m]

That is a different router he has the RT-AX58U

You're right. I should read better.

 

[GaselK]

Anyone brave enough to test a completely untested RT-AX56U build? (don't have one to test it).

Fire it up ! i have a RT-AX56U waiting !

 

[themiron]

Fire it up ! i have a RT-AX56U waiting !

soon

 

[RMerlin]

Fire it up ! i have a RT-AX56U waiting !

Thanks, I was able to get it tested by themiron.

Sorted out one last issue with it, so RT-AX56U support will be available in 384.16.

 

[Phil Outram]

Thanks, I was able to get it tested by themiron.

Sorted out one last issue with it, so RT-AX56U support will be available in 384.16.

Great news, I've also got an AX56U if you need anyone else to test.

 

[JGrana]

I have one of the RT-AX58U units.

...
One odd thing that I noticed (a Linux thing) is that when I use command "top" via SSH to check the CPU load, the normal Linux load averages look wrong. The RT-AX58U load typically is less then 6% per core.

Using SSH and "top" command and press number "1" to see all the cores individually:

Mem: 215508K used, 296752K free, 1020K shrd, 0K buff, 16076K cached
CPU0: 0.3% usr 1.1% sys 0.0% nic 98.2% idle 0.0% io 0.0% irq 0.1% sirq
CPU1: 0.5% usr 3.7% sys 0.0% nic 94.6% idle 0.0% io 0.0% irq 0.9% sirq
CPU2: 0.0% usr 1.1% sys 0.0% nic 98.8% idle 0.0% io 0.0% irq 0.0% sirq
Load average: 0.46 0.44 0.45 1/125 11189

If I look at all the cores averaged

CPU: 0.5% usr 2.1% sys 0.0% nic 97.0% idle 0.0% io 0.0% irq 0.2% sirq

The 1, 5, 15 minute load averages seem incorrect. The system is not running at 46% load and it is not running at .46% load either. Each core is running at around 6% or less all the time, and the average of all cores combined is less than 3%. So not sure what the 1, 5, 15 minute load averages are showing me.

I just purchased an AX58U to use as an Aimesh node to an AX88U. I have been evaluating this combination since yesterday. I am presently running stock Asus with Entware side loaded (so I can run iperf3, htop, shellinabox etc).
The AX88U is running RMerlin 384.16_alpha1-gfcd95af01 and Entware, etc.

I ran htop on both. To me, the load average looked typical. The AX88U is running 69 tasks, the AX58U 56. The CPU load on the cores (on both) bounce between almost 0 to an average of 3% with an occasional bounce to 6% or 7%.

I ran some stress tests on the AX58U. As expected, something like "while true; do true; done" swamps a CPU and pegs it to ~100%. If I run 2 instances, it pegs 2 cores.

So, it seems to scale and when "idle" the CPU load looks close to what the AX88U shows.

My take is that with a lot of the packet handling done in silicon, these cores are usually sitting around ;-).

BTW, so far I am very impressed with the AX58U. Nice compact device and price point.

 

[swejuggalo]

Maybe this can make some sense into the speed issues on wifi on AX88U. Adding some extra rows just to show when the error comes in a better way.

"Mar 7 17:42:37 WLCEVENTD: Assoc ¥¥¥¥¥¥
Mar 7 17:42:37 hostapd: eth7: STA ¥¥¥¥¥¥ IEEE 802.11: associated
Mar 7 17:42:37 kernel: CFG80211-ERROR) wl_cfg80211_change_station : WLC_SCB_AUTHORIZE sta_flags_mask not set
Mar 7 17:42:37 hostapd: eth7: STA ¥¥¥¥¥¥ RADIUS: starting accounting session ¥¥¥¥¥¥
Mar 7 17:42:37 hostapd: eth7: STA ¥¥¥¥¥¥ WPA: pairwise key handshake completed (RSN)"

and

"Mar 7 18:30:07 hostapd: eth7: STA ¥¥¥¥¥¥ WPA: pairwise key handshake completed (RSN)
Mar 7 18:35:05 kernel: CFG80211-ERROR) wl_cfg80211_sta_info : GET STA INFO failed, -21
Mar 7 18:35:08 hostapd: eth6: STA ¥¥¥¥¥¥ IEEE 802.11: disassociated due to inactivity"

 

[RMerlin]

Great news, I've also got an AX56U if you need anyone else to test.

Thanks. Still a lot of work to do, it's currently not working properly, and since I don't have a unit here debugging it will be slow and painful.

 

[GaselK]

56U Gpl just emerged.
"Version 3.0.0.4.384.82532020/03/06 1.76 GBytes
GPL of ASUS RT-AX56U for firmware 3.0.0.4.384.8253"
https://dlcdnets.asus.com/pub/ASUS/wireless/RT-AX56U/GPL_RT_AX56U_30043848253.zip

 

[RMerlin]

56U Gpl just emerged.
"Version 3.0.0.4.384.82532020/03/06 1.76 GBytes
GPL of ASUS RT-AX56U for firmware 3.0.0.4.384.8253"
https://dlcdnets.asus.com/pub/ASUS/wireless/RT-AX56U/GPL_RT_AX56U_30043848253.zip

I already merged that nearly a week ago.

commit 91b0ba80f16edab98ac68caeba582a89740b76b6
Author: Eric Sauvageau <merlin@asuswrt-merlin.net>
Date:   Tue Mar 3 17:01:23 2020 -0500
    Merged with GPL 384_8253 (from RT-AX56U)

 

[Billy Chaney]

I'm having a problem when using the dhcp-option with my VPN client. When I use that option my wireless Laptop can't seem to automatically connect when booting up. When using an ethernet connection on my other Laptop I don't have this problem. When I revert back to 384.15 everything works as it should.

 

[Toink]

Reverted back to 384.15 as well.

I have to manually toggle the 'internet connection' multiple times to get my internet connection back - even after manually power-cycling the router if I made changes in the WAN settings or just merly rebooting/power cycling the router. This happened in both RT-AX88U_384.16_alpha1-gfcd95af015 and RT-AX88U_384.16_alpha1-g2a2f50e2fa alpha(s). It does not manifest in 384.15.

SETUP:
ONU in bridge mode > AX88U
Aside from VPN and scripts to access my ONU from the router, there is no other installed script.

 

[swejuggalo]

I'm having a problem when using the dhcp-option with my VPN client. When I use that option my wireless Laptop can't seem to automatically connect when booting up. When using an ethernet connection on my other Laptop I don't have this problem. When I revert back to 384.15 everything works as it should.

I use OpenVPN with no apperant issue. I only use it on my phone though and autoconnect with VPN Client Pro on wifi connections that is not on my whitelist. (AX88U)

Sent from my GM1913 using Tapatalk