Beta Asuswrt-Merlin 3006.102.5 Beta is now available

[GWTechTalk]

After full reset this version seems stable. Dirty flashing not recommended. BSD and Smart Connect NVRAM parameters seem to make this build unstable so I'm no longer doing that. Default logging is debug, for newbies sakes might want to change that. So far nothing abnormal noticed in logging however, there are normal operation errors showing that some might mistake for issues. IOT devices appear fully stable now after reset where previously they were having major connection and auth issues when dirty flashed. New ASUS firmware also dropped fixing an issue with 5Ghz on the BE96U not sure if that's in this base.

 

[RMerlin]

New ASUS firmware also dropped fixing an issue with 5Ghz on the BE96U not sure if that's in this base.

Beta - Asuswrt-Merlin 3006.102.5 Beta is now available

Code yes but what about the wireless drivers version I believe they are the same in the latest ASUS release and the 3006.102.5 Beta 2 correct? CC Version means nothing if patches are applied on top of existing code.

www.snbforums.com

 

[RMerlin]

Ports are enumerated in the wrong order on the RT-BE92U:

10G, 4, 1, 2, 3

Instead of

10G, 1, 2, 3, 4

People using the 2.5 Gbpos WAN should swap ports 1 and 2 on their router.

I'll see next week if this is something that Asus can fix without requiring a new GPL (this was a regression from when they migrated the RT-BE92U from the dedicated code branch to the common Wifi 7 code branch).

 

[Rhialto]

Ports are enumerated in the wrong order on the RT-BE92U:

10G, 4, 1, 2, 3

I have to ask, does it looks like bad QA (Quality Assurance) to you? I mean how can they miss this? They don't have a private group to check bits they could have missed? ASUS is a huge company with lots of ressources.

Oh well, you're alone improving some parts of their firmware and you have many fans here willing to alpha/beta test so was wondering about ASUS. Maybe they read and monitor you so you can't speak much... lol just kidding.

 

[dave14305]

I have to ask, does it looks like bad QA (Quality Assurance) to you? I mean how can they miss this? They don't have a private group to check bits they could have missed? ASUS is a huge company with lots of ressources.

To be fair, Asus has never publicly released this 102_38757 code that 102.5 is based on. Maybe they’re QA’ing it as we speak?

ASUS RT-BE92U｜WiFi 7｜ASUS USA

RT-BE92U achieves blazing-fast WiFi speeds of up to 9,700 Mbps, ensuring seamless streaming and data-intensive tasks, which is also flexible, safe, and easy to manage.

www.asus.com

 

[RMerlin]

I have to ask, does it looks like bad QA (Quality Assurance) to you?

No, because I am using unreleased code. Their RT-BE92U releases are all from a dedicated branch, they merged it to the branch that I'm using to make my life easier, and sent me the result like a day or two after they merged it. It was untested beyond probably just making sure the code compiled and ran.

 

[Ev0]

Beta2 seems to have a bug with RT-BE92U.

My iPhone 15 Pro Max will connect via Wifi, but there is no internet connection no matter what I do.

I did see some mention that limit ip tracking is the problem, but it doesn't change anything if I turn on or off.

Also using 2.4 / 5 or 6Ghz did not change anything.

In an attempt to fix the issue, I did a settings backup, then factory reset and then restored the backup again.

Nothing changed.

Flashed back to Beta1 and it's working perfectly again.

 

[Comoman]

Running a BE88U with Beta 2. No matter what I do, the wifi channels wont stay fixed. I have manually selected 36 with 20/40/80/160MHz bandwidth. After a few hours it always changes to another channel at 80MHz although on the logs 36/160 is clear. Not sure if there's anything that can be done as I assume it's related to the wireless drivers....

 

[bbunge]

Yesterday I flashed 3006.102.4 over Asus firmware. Factory reset and manually configured. Added a thumb drive. Set up a swap file. Added Entware and Diversion. Flashed 3006.102.5 beta 2. Added a guest IoT WIFI. Worked great! After a couple of hours I turned on DNS Director and added Skynet then noticed that the connected clients count went past the "normal" 32 in Network Map. I discovered that the Dish Hopper with Sling was showing 41 connected clients and as I watch that number clmbed to over 70! The Hopper is connected to a Joey via MOCA. The Hopper is connected to the LAN via its upper Ethernet port and normally the Hopper and Joey each get an IP address. As it was getting late and I felt something was very wrong I flashed back to Asus firmware and restored from a saved settings file.
Today, I have gone back to the 3006.102.5 beta 2 but have not added Skynet nor turned on DNS Director. All is working well with no strange events.
Am reporting this just-in-case...

 

[bennor]

... then noticed that the connected clients count went past the "normal" 32 in Network Map. I discovered that the Dish Hopper with Sling was showing 41 connected clients and as I watch that number clmbed to over 70! ...

Am reporting this just-in-case...

Seems to be an issue a few are reporting (climbing connected client counts) for what ever reason. Some earlier posts in the discussion about it:
https://www.snbforums.com/threads/asuswrt-merlin-3006-102-5-beta-is-now-available.95115/#post-960904
https://www.snbforums.com/threads/a...eta-is-now-available.95115/page-8#post-962914

 

[MDM]

Running a BE88U with Beta 2. No matter what I do, the wifi channels wont stay fixed. I have manually selected 36 with 20/40/80/160MHz bandwidth. After a few hours it always changes to another channel at 80MHz although on the logs 36/160 is clear. Not sure if there's anything that can be done as I assume it's related to the wireless drivers....

Only related to DFS...

 

[NiklasO]

Yesterday I flashed 3006.102.4 over Asus firmware. Factory reset and manually configured. Added a thumb drive. Set up a swap file. Added Entware and Diversion. Flashed 3006.102.5 beta 2. Added a guest IoT WIFI. Worked great! After a couple of hours I turned on DNS Director and added Skynet then noticed that the connected clients count went past the "normal" 32 in Network Map. I discovered that the Dish Hopper with Sling was showing 41 connected clients and as I watch that number clmbed to over 70! The Hopper is connected to a Joey via MOCA. The Hopper is connected to the LAN via its upper Ethernet port and normally the Hopper and Joey each get an IP address. As it was getting late and I felt something was very wrong I flashed back to Asus firmware and restored from a saved settings file.
Today, I have gone back to the 3006.102.5 beta 2 but have not added Skynet nor turned on DNS Director. All is working well with no strange events.
Am reporting this just-in-case...

Looks like the same problem I (and another person) posted about earlier in this thread. Haven’t seen any response to it yet.
I hope it can be fixed
I haven’t noticed any negative effects from it, apart from the client count being useless.

 

[bbunge]

Seems to be an issue a few are reporting (climbing connected client counts) for what ever reason. Some earlier posts in the discussion about it:
https://www.snbforums.com/threads/asuswrt-merlin-3006-102-5-beta-is-now-available.95115/#post-960904
https://www.snbforums.com/threads/a...eta-is-now-available.95115/page-8#post-962914

Thanks. Good to know the Dish box was not acting up. However, there were a couple of my IoT clients that did not like the WPA2-Personal in this release. Had to drop back to WPA/WPA2-Personal to get them to connect. I feel that my test is complete and that there is no real advantage for me to use the Merlin firmware. Back to Asus firmware.

 

[Marcus Yansen]

After full reset this version seems stable. Dirty flashing not recommended. BSD and Smart Connect NVRAM parameters seem to make this build unstable so I'm no longer doing that. Default logging is debug, for newbies sakes might want to change that. So far nothing abnormal noticed in logging however, there are normal operation errors showing that some might mistake for issues. IOT devices appear fully stable now after reset where previously they were having major connection and auth issues when dirty flashed. New ASUS firmware also dropped fixing an issue with 5Ghz on the BE96U not sure if that's in this base.

View attachment 66936

Hi, when full reset is done does it/do you refer to soft or hard reset of the router? Also, does a hard reset include soft reset by default (therefore making doing the soft reset redfundant)

Thanks

View attachment 66937

View attachment 66938

 

[RMerlin]

Flashed back to Beta1 and it's working perfectly again.

Zero change related to wifi between beta 2 and beta 1. Your issue lies elsewhere.

 

[RMerlin]

No, because I am using unreleased code. Their RT-BE92U releases are all from a dedicated branch, they merged it to the branch that I'm using to make my life easier, and sent me the result like a day or two after they merged it. It was untested beyond probably just making sure the code compiled and ran.

An Asus engineer looked into it, and the issue is specific to Asuswrt-Merlin. The problem is that the GT-BE98U_Pro and the RT-BE92U both use a closed source Realtek network driver, and that driver is compiled differently per model. Asuswrt-Merlin's kernel tree contained the GT-BE98U_Pro version, which is incompatible. I will need to keep separate copies for each model, and copy the appropriate one at compile time.