Beta Asuswrt-Merlin 386.7 Beta is now available

[learning_curve]

I ran a dirty upgrade from Alpha to Beta 1. No errors, no problems & all good for me so far. Re Your IPv6 related feedback requests @RMerlin As per the Alpha release: All works perfectly for me, but FWIW My IPv6 Caveats;

There's still no DDNS IPv6 Support - Expected at this stage
There's still no full IPv6, OpenVPN Client Support - Expected at this stage
My RT-AC68U is a Non-HND Model, so I can't add any feedback / info on the IPv6 DNS Filter item
I don't use OpenVPN IPv6 Server at all (only Clients) so I can't add any feedback / info on this item either.

 

[ComputerSteve]

Anyone else see issues with amtm scripts (Diversion and unbound, possibly others) that don't want to accept any new updates, or even a forced re-install (uu updates)? This is on an RT-AC86U, and has been happening since the alpha release, as well as on the newer beta release. Jffs is mounted, scripts run, but something's not right: I seem to always get something similar to "Error: Address /opt/share/diversion/webui/process.div is invalid", even if the scripts themselves run. Similar with unbound, but it exits out of the update process too fast to get a screenshot.

I was having this problem. Script appeared working but the iOS shortcut stopped functioning.

 

[jsbeddow]

I was having this problem. Script appeared working but the iOS shortcut stopped functioning.

And....you repaired it by doing....?
A full format and reinstall of amtm and all scripts? Sure, I know that will work, but was trying to avoid it if possible.

 

[SwampKracker]

They're not providing me release code, they're providing me development snapshots as I need to have the same codebase for all models.

Seriously guys, this is complaining about something totally irrelevant. Has any of you ever seen a Red Hat 7 system log, or even a Windows Event log? These are 100X more verbose than your router's log.

Be glad I am able to get something that works.

This criticism is directed toward Asus, not you, RMerlin.

With regard to RH7 or Windows, they have far greater resources (CPU, RAM, storage) to exploit than does a simple Asus router. Having a simple check box for enable/disable debug logging would be a considerate thing to do for the customer.

 

[PeterR]

DDNS Custom Server - I'm having a problem entering a pipe symbol '|' in the host name field - it's not accepted (I'm using duckdns.org with a custom script 'host|token'in the field)

 

[pinkfloyd1173]

Dirty up grade like a Tijuana cantina, all is well.

 

[RMerlin]

I was able to reproduce the RT-AX86U JFFS problem by reverting back to 386.5_2, then flashing 386.7 on top of it. It probably didn't happen before as I had the newer cferom already flashed, so it could be a problem when mixing the older cferom with the newer firmware content.

I'll remove the RT-AX86U firmware for now, I'll need to do more investigation tomorrow.

 

[RMerlin]

DDNS Custom Server - I'm having a problem entering a pipe symbol '|' in the host name field - it's not accepted (I'm using duckdns.org with a custom script 'host|token'in the field)

Which is expected, as that character is not valid for a hostname.

 

[RMerlin]

Having a simple check box for enable/disable debug logging would be a considerate thing to do for the customer.

The problem is that this is a compile-time flag, not something that can be enabled or disabled at runtime. To make it user-configurable would require Broadcom to rewrite their code to allow that.

 

[GoNz0]

On the main page DDNS isn't showing as connected but on the DDNS page is shows registered when you click apply. AX86 router, clean install with manual setup.

I had to reboot the router and now have the padlock and DDNS is working, the homepage still shows GO instead of my domain.
View attachment 41743

@RMerlin
I think my issues got lost with all those deciding it's ok to spam the thread with crap about blood pressure monitors.

 

[hasan]

I did a dirty update from last stable release on my ac88u to this build.

About 12 hours after updating. It takes an extremely long time for the router interface to load after logging, if it all. This isn't an issue I faced with the last stable build however I did enable upnp/secure upnp post update which was not previously enabled.

I'm yet to bounce the router, will probably do it later this evening when everyone is offline to see if its something repeatable.

 

[maghuro]

LOL, that particular model the RT-AC86U is such a PITA with the updates NOT initiating properly...
I went to the following method & haven't had a problem since.
You need to change the variables in the top section, save the script & make it executable... of course.
It ALMOST makes it idiot-proof.

#!/bin/bash

DATA_DIRECTORY="/home/pi/Downloads"
USERNAME="admin"
ROUTER_IP="12.34.56.78"

clear
cd $DATA_DIRECTORY
echo "sha256sum -c sha256sum.sha256"
sha256sum -c sha256sum.sha256
WORKDIR=$(pwd)
echo "Please Note: the working directory = $WORKDIR"
FIRMWARE=$(ls -t RT-AC86U_*.w . | head -n 1)
echo "And the FIRMWARE to be flashed, will be... "
echo "$FIRMWARE"
echo "( Before proceeding )"
echo "FIRST verify the FIRMWARE-version is CORRECT "
echo "Then..."
echo "CUT, PASTE, & ENTER each of the following command lines: "
echo "( One-line, after the other )"
echo "----------------------------------------------------------"
echo "cd $WORKDIR"
echo "scp $FIRMWARE $USERNAME@$ROUTER_IP:/tmp"
echo "ssh $USERNAME@$ROUTER_IP"
echo "cd /tmp"
echo "/sbin/hnd-write $FIRMWARE"
echo "reboot now"
echo "----------------------------------------------------------"

Very very noob-proof indeed!
Thank you very much for the script!
I was preparing myself to trying it. However, in an ultimate desperate attempt, I went again to firmware page and uploaded the firmware (even with USB mounted and no preparation at all), and guess what... It flashed correctly
Can you explain? Lol

 

[charlie2alpha]

Upgraded my AX86U last night, came back running fine. Did not encounter the jffs issue.

 

[FLA_NL]

I was able to reproduce the RT-AX86U JFFS problem by reverting back to 386.5_2, then flashing 386.7 on top of it. It probably didn't happen before as I had the newer cferom already flashed, so it could be a problem when mixing the older cferom with the newer firmware content.

I'll remove the RT-AX86U firmware for now, I'll need to do more investigation tomorrow.

Just to notice, I've never flashed the cferom version, so I must have a mix of the older cferom and the newer firmware, all is well here. Also didn't have the traffic analyser on, I only use the jffs for adding a dnsmasq.conf.add to configure some local domains. The tar backup is around 4 mbyte. On the tools page the size is indicated as 2.95 / 48.00 MB.

 

[bits]

I upgraded my ax86u from 386.5_2 to 386.7 alpha no cferom then upgraded to 386.7 beta.

So only skipped the alpha with cferom update.

No brick or jffs issue.
By utter luck and chance traffic analysis was disabled and cleared on 386.5_2 and was reenabled at 386.7 beta.

Unit details
hw 1.0
Production year 2020
S/n L6IF6000xxx

 

[spykos]

Yep, this is what I found and used to get it back going.

Did you upload the firmware with Rescue tool ? I am asking cause I had the same issue and because Rescue tool was not working I had to revert to TFTPD client.

 

[Ubimo]

I'm noticing a "not proper logout after timeout" in the GUI.
When I'm logged in and wait some hours, the GUI should take me to the login-screen due to timeout, but it does not. Instead I still see the menu and can click around. Some submenus load, some dont'.
The correct way should be, that the GUI would take me to the login screen, but it does not. I have to manually click on the logout button to see the login screen. Then the submenus will load again.
I hope I could express my findings understandably. I haven't set a timeout under Administration->System->timeout inactivity.

AC86U 386.7 beta

 

[alan6854321]

When I'm logged in and wait some hours, the GUI should take me to the login-screen due to timeout, but it does not. Instead I still see the menu and can click around. Some submenus load, some dont'.

I've noticed this for several releases now.

After a while, when you should have been 'timed out', you can still explore the UI, in particular view the network map all log pages.

But once you try to view a page that 'does something' E.g. scMerlin page, you get thrown out to the login page.

 

[alan6854321]

I've noticed a new message appeared in the syslog ...

kernel: sched: RT throttling activated

Anything to worry about?

 

[JGrana]

The problem is that this is a compile-time flag, not something that can be enabled or disabled at runtime. To make it user-configurable would require Broadcom to rewrite their code to allow that.

If these debug messages (and others) annoy users, I would recommend installing scribe via the built in amtm command. It will install syslog-ng which allows all kinds of filters. Scribe better integrates syslog-ng with Asuswrt-merlin firmware and also provides a number of examples to get you going. You can send them to seperarte files or simply “drop” them.

Dropping messages​

log: Filter and route log messages using log paths, flags, and filters > Dropping messages
To skip the processing of a message without sending it to a destination, create a log statement with the appropriate filters, but do not include any destination in the statement, and use the final flag.

Example: Skipping messages​

The following log statement drops all debug level messages without any further processing.
filter demo_debugfilter { level(debug); };
log { source(s_all); filter(demo_debugfilter); flags(final); };