Beta Asuswrt-Merlin 388.1 Beta is available for select models

[supe]

Slightly related to this, and while you have the engineers on the line about fc...

Flow cache on the AX86U has caused major issues with QoS (believe stock and with Merlin Addons) for a while now. Some of the original classification bugs appear to have been worked out, but fc actually throttles download speed.

This is reproduceable by enabling QoS, and then comparing *not router-run* speedtests with fc enabled and disabled. Router-run speedtests aren't impacted.

The ax86U does behave wonky with QOS, I notice that WirelessLAN to wiredLAN files transfer speeds drops 25% with QOS enabled. This does not happen on the lower ax68U model.

 

[skeal]

Anyone seen issues where the router will suddenly show the initial config page after reboot?

Router is working perfectly fine, router is then restarted (either manual or via scheduled reboot) then I lose internet but wifi, DHCP... is still configured and allows me to connect to the router.
I can navigate to other subpages manually (doesn't prompt for password).

I literally have to hard reset and reupload the saved config to get it back up and running.

Any thoughts?

Do a reset to defaults and DON'T import that settings file. Hand configure from screenshots.

 

[heysoundude]

It should be disabled. What does

fc status

show?

Here is what it should look like when using wireguard or cake:

fc status
        Flow Timer Interval = 10000 millisecs
        Pkt-HW Activate Deferral rate = 1
        Pkt-HW Idle Deactivate = 0
        Pkt-SW Activate Deferral count = 0
        Acceleration Mode: <L2 & L3>
        MCast Learning <Disabled>
        MCast Acceleration IPv4<Enabled> IPv6<Enabled>
        IPv6 Learning <Enabled>
        GRE Learning <Enabled> Mode<Tunnel>
        4o6 Fragmentation <Enabled>
        TCP Ack Prioritization <Enabled>
        HW Acceleration <Disabled>
        Flow Learning Disabled : Max<16384>, Active<3>, Cummulative [ 8 - 5 ]

HW Acceleration should be <Disabled>

where can we find a description of what the fc parameters are/do? is there a man page somewhere?
(Flow timer interval of 10000ms or 10sec seems a bit long to me...)

 

[heysoundude]

Passthrough, if your Asus is behind ISP modem/router. Native, if your Asus is connected to modem only. Disabled, if you don't need IPv6.

#ThisIsTheWay

 

[ColinTaylor]

where can we find a description of what the fc parameters are/do? is there a man page somewhere?

It's a closed source Broadcom module. I'm not aware of any publicly available documentation.

 

[aex.perez]

I have just updated my GT-AX6000 to Beta2 from 386.7.2

My DDNS (Google Domains) was showing an exclamation mark
I disabled IPv6 Update in DDNS (i have native ipv6) and its working fine now for ipv4
is this a google problem or router problem?

Have the same issue, just with an AX88, with different DDNS provider - same issue and result as you.

 

[sl4fko]

Dirty upgrade on my GT-AX6000, no problems in 3 days uptime, aimesh works (with AC68U), DDNS with Instant Guard working good.

Tried ROG "cosmetics", does not work for me (complexity, low visibility).

 

[samw5]

Do a reset to defaults and DON'T import that settings file. Hand configure from screenshots.

 

[Damit1]

Have the same issue, just with an AX88, with different DDNS provider - same issue and result as you.

Same here. Disabled IPv6 for ddns on AX88U, voila!

Ddns is NoIP

 

[ZebMcKayhan]

Can anyone can confirm if port forwarding works with wireguard VPN Director?
I tested 2 different clients with Torguard VPN, the port still showing closed! if i try the same client with openvpn the port shows open.
any ideas? thx

Im not 100% sure what you are getting at, trying to connect back to a Wireguard server via a Wireguard client which offers port-forward?

Just want to add that a wireguard server will ONLY reply to packets carrying the right encryption. Port scan packets do not carry the right encryption so the port will appear as stealth. While I find this really cool it also makes debugging more difficult.

 

[archiel]

Have the same issue, just with an AX88, with different DDNS provider - same issue and result as you.

I am using the Asuscomm DDNS and although IPv6 didn't work at first after a dirty update (386.8 to beta1) after a reboot it has been fine. Also, no problems after the update to beta2.

 

[Damit1]

That new portmap implementation is quirky on some models. It's closed source so outside of my control. On my GT-AXE16000 for instance it doesn't even appear at all, as the backend fails to generate the required data.

That would explain this, I guess-

Stock firmware for AX88U, I read earlier that your GPL there is having a one newer than the stock one. there they have a fix for these, on the other models (like yourselves AXE16000 and AX56U) it is not distributed yet.

If you have a word with them, maybe worth mention.

 

[iJorgen]

A week with Beta 2 and everything is running smooth, no drop-outs, clean logs and great performance!

 

[Volfi]

Memory consumption on AX86U goes up by 2% every day on Beta2.
This did not happen on previous stable release.

 

[GHammer]

I have just updated my GT-AX6000 to Beta2 from 386.7.2

My DDNS (Google Domains) was showing an exclamation mark
I disabled IPv6 Update in DDNS (i have native ipv6) and its working fine now for ipv4
is this a google problem or router problem?

Set up DNSSEC & DNS security - Google Domains Help

Domain Name System (DNS) translates human-readable domain names like google.com into the machine-readable IP address of a website, like 172.217.3.206. To prevent threats to your domain like cache p

support.google.com

IPv4 and IPv6 are supported, not at the same time though.

 

[Ser9ei]

I added wgserver-start, wgserver-stop, wgclient-start and wgclient-stop (with the unit number passed as argument).

Hi,
Is it available in the 2'd beta?

 

[octopus]

It does auto-refresh for me. Look for any Javascript error in the browser console that may be preventing the refresh from happening.

I can't see any javascript errors on that page.

This page is in near default mode. Page layout may be affected. For default mode use "<!DOCTYPE html>".
Advanced_VPNStatus.asp Some cookies abuse the recommended "SameSite" attribute 3
Cookie "clickedItem_tab" does not have a correct "SameSite" attribute value. Soon cookies without the "SameSite" attribute or with an invalid value will be treated as "Salmon". This means that the cookie is no longer sent in external contexts. If your application depends on this cookie being available in such contexts, add the "SameSite=None" attribute to it. Read https://developer.mozilla.org/docs/Web/HTTP/Headers/Set-Cookie/SameSite to learn more about the "SameSite" attribute state.js:120:67 Cookie "notification_history" does not have a correct "SameSite" attribute value. Soon cookies without the "SameSite" attribute or with an invalid value will be treated as "Salmon". This means that the cookie is no longer sent in external contexts. If your application depends on this cookie being available in such contexts, add the "SameSite=None" attribute to it. Read https://developer.mozilla.org/docs/Web/HTTP/Headers/Set-Cookie/SameSite to learn more about the "SameSite" attribute state.js:128:14 Cookie "clickedItem_tab" does not have a correct "SameSite" attribute value. Soon cookies without the "SameSite" attribute or with an invalid value will be treated as "Salmon". This means that the cookie is no longer sent in external contexts. If your application depends on this cookie being available in such contexts, add the "SameSite=None" attribute to it. Read https://developer.mozilla.org/docs/Web/HTTP/Headers/Set-Cookie/SameSite to learn more about the "SameSite" attribute state.js:120:6
Synchronous XMLHttpRequest on the main thread is deprecated due to its negative impact on the user experience. For more help, see http://xhr.spec.whatwg.org/

 

[nickie]

Memory consumption on AX86U goes up by 2% every day on Beta2.
This did not happen on previous stable release.

when you refer to the previous stable release, do you mean the one based on 386? If so, I'm also seeing this behaviour on my tuf-ax5400 (not supported by @RMerlin ) on 388 stock. I used to have about 110/120mb of free RAM and now it hovers around 56/62MB, with no instability being detected. Maybe something was changed in terms of RAM management/showing with 388 from 386.

 

[Volfi]

when you refer to the previous stable release, do you mean the one based on 386? If so, I'm also seeing this behaviour on my tuf-ax5400 (not supported by @RMerlin ) on 388 stock. I used to have about 110/120mb of free RAM and now it hovers around 56/62MB, with no instability being detected. Maybe something was changed in terms of RAM management/showing with 388 from 386.

Yes, 386.7_2, this one was "RAM stable" for weeks, always same memory percentage used. 388 seems to eat more and more RAM daily. Not sure if it will sustain one month running.

 

[penguin22]

Yes, 386.7_2, this one was "RAM stable" for weeks, always same memory percentage used. 388 seems to eat more and more RAM daily. Not sure if it will sustain one month running.

I do recall a change being made in recent builds to not purge RAM in the same way that several older builds were doing so. RAM usage is not a bad thing as it is maintaining items in memory as required and can overwrite as necessary. This change does mean that usage will remain higher than in some of the older builds, though is by design rather than indicative of an issue.