[Beta] Asuswrt-Merlin 382.1 Beta is available

[RMerlin]

As well as the following error-syslog entries:

Squashfs errors would indicate errors in the flash. Try reflashing again.

 

[tsunami2311]

@RMerlin I understand ac66u will be highly likely be dropped out off support.. How about ac66u b1? Will it still be supported by you when Asus drop 382 source?


Sent from my iPad using Tapatalk

this would make me sad, i still using the AC66U, I been thinking about getting the AC86U but change routers is of low priority atm

 

[hervon]

I just cancelled my order for a AC86U! I'll wait to make sure Merlin can support it as he wishes. I completely understand that those obstacles are beyond his will. Without Merlin, Asus routers are just overpriced. Their specs are available from other vendors.

 

[isr25]

Any new setting that's longer than 255 characters won't be saved. This will limit the number of Policy rules to 255 chars total for instance - around 5-7 rules, depending on the length of your description field.

Any settings which I made larger (like the DHCP static reservations) won't be able to store more settings than allowed by the original buffer length. The DHCP reservation list will be limited even more, as I add an extra hostname field to the stored data in nvram.

I was also forced to move SSH keys to JFFS during early development - didn't know at the time the cause, now I do. Anything stored to jffs won't be included in settings backups.

Dang... I hope Asus can assist or revert their changes on this. I already have the AC86U to replace my current AC66U - hopefully you won’t drop the support

 

[Kewjoe]

Dang... I hope Asus can assist or revert their changes on this. I already have the AC86U to replace my current AC66U - hopefully you won’t drop the support

I'm in the same boat. Just bought a AC86u to replace my AC66u. Asus better realize what their stupid decisions are doing. Look at how many people have downloaded the alpha/beta 382. A lot of customers are only customers because of Merlin. They need to have an open dialogue with him and stop closing everything off.

 

[RMerlin]

I reached out to Asus on the issue, we'll see if they can offer a solution.

 

[SeaConn]

I'm in the same boat. Just bought a AC86u to replace my AC66u.

Did the same thing. I'm just hoping I don't end up with two non-Merlin Asus routers now. As is, the Beta is working great for me. I only use it for OpenVPN and policy based routing, of which, I only use three policies:

1. Route all to tunnel
2. Route the router to WAN
3. Route Synology DS916+ to WAN

Even with the NVRAM issue, I get by okay since I only statically route a few devices on my network. Works for me!

 

[matthew_eli]

Hi @RMerlin,

I've updated my RT-AC88U from 380.68_4 to 382.1_beta1 and all went smooth (no factory restore), excepting for the following:

• My CTF Hw Acceleration changed from disabled (380.68_4) to automatic (382.1_beta1): I reverted it to disabled, due to inconsistency with QOS management
• Samba shares have been limited by the QOS engine (set in Adaptive QOS or Bandwidth limiter modes): if you restart the QOS engine the problem is fixed, but it returns with the next router reboot
• I had to clear my browser cash with CTRL+F5 in order to correctly visualize the webUI
• I had a lot of these message in the syslog:

Oct 20 09:09:11 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:24 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:24 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:31 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:31 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:13:10 kernel: dhd_prot_flow_ring_create: Send Flow Create Req flow ID 262 for peer XX:XX:XX:XX:XX:XX prio 0 ifindex 0
Oct 20 09:13:10 kernel: dhd_prot_flow_ring_create_response_process: Flow Create Response status = 0 Flow 262
Oct 20 09:13:10 kernel: CONSOLE: 027513.964 flow_create : bitmap_size=512  maxitems=512
Oct 20 09:13:10 kernel: CONSOLE: 027513.975 wl0.0: wlc_send_bar: seq 0x1 tid 0
Oct 20 09:13:17 kernel: dhd_flow_rings_delete_for_peer: ifindex 0
Oct 20 09:13:17 kernel: dhd_prot_flow_ring_delete: Send Flow Delete Req RING ID 262 for peer XX:XX:XX:XX:XX:XX prio 0 ifindex 0
Oct 20 09:13:17 kernel: dhd_prot_flow_ring_delete_response_process: Flow Delete Response status = 0
Oct 20 09:13:17 kernel: CONSOLE: 027520.974 wl0: Proxy STA XX:XX:XX:XX:XX:XX link is already gone !!??
Oct 20 09:18:40 kernel: dhd_prot_flow_ring_create: Send Flow Create Req flow ID 262 for peer XX:XX:XX:XX:XX:XX prio 0 ifindex 0
Oct 20 09:18:40 kernel: dhd_prot_flow_ring_create_response_process: Flow Create Response status = 0 Flow 262
Oct 20 09:18:40 kernel: CONSOLE: 027842.960 flow_create : bitmap_size=512  maxitems=512
Oct 20 09:18:40 kernel: CONSOLE: 027842.971 wl0.0: wlc_send_bar: seq 0x1 tid 0
Oct 20 09:18:48 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:18:48 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:18:51 kernel: dhd_flow_rings_delete_for_peer: ifindex 0
Oct 20 09:18:51 kernel: dhd_prot_flow_ring_delete: Send Flow Delete Req RING ID 262 for peer XX:XX:XX:XX:XX:XX prio 0 ifindex 0
Oct 20 09:18:51 kernel: dhd_prot_flow_ring_delete_response_process: Flow Delete Response status = 0
Oct 20 09:18:51 kernel: CONSOLE: 027854.439 wl0: Proxy STA XX:XX:XX:XX:XX:XX link is already gone !!??

Is that normal?

With OpenVPN I can say it's all ok: I also tried to connecting my Android phone from outside the LAN and all went smooth as in 380.68_4

EDIT: I removed the explicit MAC address from the log

EDIT2: added the Samba and QOS bug

EDIT3: for temporarily fixing the Samba Bug, I add the following command to service-start script:

service restart_qos

 

[pattiri]

Hi @RMerlin,

I've updated my RT-AC88U from 380.68_4 to 382.1_beta1 and all went smotth, excepting for the following:

• My CTF Hw Acceleration changed from disabled (380.68_4) to automatic (382.1_beta1): I reverted it to disabled, due to inconsistency with QOS management
• I had to clear my browser cash with CTRL+F5 in order to correctly visualize the webUI
• I had a lot of these message in the syslog:

Oct 20 09:09:11 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:24 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:24 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:31 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:09:31 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:13:10 kernel: dhd_prot_flow_ring_create: Send Flow Create Req flow ID 262 for peer 70:ee:50:12:3c:86 prio 0 ifindex 0
Oct 20 09:13:10 kernel: dhd_prot_flow_ring_create_response_process: Flow Create Response status = 0 Flow 262
Oct 20 09:13:10 kernel: CONSOLE: 027513.964 flow_create : bitmap_size=512  maxitems=512
Oct 20 09:13:10 kernel: CONSOLE: 027513.975 wl0.0: wlc_send_bar: seq 0x1 tid 0
Oct 20 09:13:17 kernel: dhd_flow_rings_delete_for_peer: ifindex 0
Oct 20 09:13:17 kernel: dhd_prot_flow_ring_delete: Send Flow Delete Req RING ID 262 for peer 70:ee:50:12:3c:86 prio 0 ifindex 0
Oct 20 09:13:17 kernel: dhd_prot_flow_ring_delete_response_process: Flow Delete Response status = 0
Oct 20 09:13:17 kernel: CONSOLE: 027520.974 wl0: Proxy STA 70:ee:50:12:3c:86 link is already gone !!??
Oct 20 09:18:40 kernel: dhd_prot_flow_ring_create: Send Flow Create Req flow ID 262 for peer 70:ee:50:12:3c:86 prio 0 ifindex 0
Oct 20 09:18:40 kernel: dhd_prot_flow_ring_create_response_process: Flow Create Response status = 0 Flow 262
Oct 20 09:18:40 kernel: CONSOLE: 027842.960 flow_create : bitmap_size=512  maxitems=512
Oct 20 09:18:40 kernel: CONSOLE: 027842.971 wl0.0: wlc_send_bar: seq 0x1 tid 0
Oct 20 09:18:48 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:18:48 kernel: dhd_prot_ioctl: status ret value is -17
Oct 20 09:18:51 kernel: dhd_flow_rings_delete_for_peer: ifindex 0
Oct 20 09:18:51 kernel: dhd_prot_flow_ring_delete: Send Flow Delete Req RING ID 262 for peer 70:ee:50:12:3c:86 prio 0 ifindex 0
Oct 20 09:18:51 kernel: dhd_prot_flow_ring_delete_response_process: Flow Delete Response status = 0
Oct 20 09:18:51 kernel: CONSOLE: 027854.439 wl0: Proxy STA 70:ee:50:12:3c:86 link is already gone !!??

Is that normal?

With OpenVPN I can say it's all ok: I also tried to connecting my Android phone from outside the LAN and all went smooth as in 380.68_4

I saw similar logs on my RT88u expect "dhd_prot_ioctl: status ret value is -17" I don't know what is this but I think the logs like "dhd_prot_flow_ring" are related with the wireless client connection. "xx:xx:xx:xx:xx:xx" this MAC address possibly belong one of your Wi-Fi clients. It seems now we can see more logs about the Wi-Fi clients, when they connected and disconnected.

 

[nash16]

Hi @RMerlin,

Could you add in Merlin firmware the Movistar+ Triple VLAN? It is mandatory for IPTV and VoIP in Movistar (Spain ISP).

Here you could get the Beta FW from Asus for RT-AC86U with the Movistar+ Triple VLAN in preview: https://www.asuswebstorage.com/navigate/s/17AD71220AC246CB90478625CE46022CY

It is supported by other Asus Routers:

Thanks in advance,
BR

 

[LouisvilleUK]

As is, the Beta is working great for me

Same here, OpenVPN performance is fantastic!!

 

[dr_lucas]

Same here, OpenVPN performance is fantastic!!

Cool.
Can you define "fantastic"?

 

[RMerlin]

I had a lot of these message in the syslog:

Same with stock firmware, Asus left debugging logging enabled in the wireless driver.

Here you could get the Beta FW from Asus for RT-AC86U with the Movistar+ Triple VLAN in preview: https://www.asuswebstorage.com/navigate/s/17AD71220AC246CB90478625CE46022CY

It has to be added by Asus to the source code. Without it, there's nothing I can do.

 

[LouisvilleUK]

Cool.
Can you define "fantastic"?

My ISP service is 200 down, 20 up. With my primary VPN service provider, PrivateTunnel, via Chicago server and AES-128-GCM cipher I get full throughput, 200/20. I tried Mullvad VPN service (3-hour trial) via Chicago server last weekend out of curiosity with AES-256-GCM cipher and was still able to achieve full throughput, 200/20.

Additional OpenVPN performance numbers from RMerlin…

https://www.snbforums.com/threads/openvpn-performance-of-the-rt-ac86u.41217/

 

[demesmaeker]

Cool.
Can you define "fantastic"?

I also just bought one and I am very happy with openvpn speeds. I have max 100mbit and I actually get the whole speed with the vpn client ativated. It can do better if I paid for a faster internet Connection

The only problem I have is that (two times in one day) I cannot connect to VPN. It says the CA-key is missing. If then look it up, it is there. I just give a enter behind the key, saved, and then tried again. Then it all worked. Very strange...

 

[dr_lucas]

My ISP service is 200 down, 20 up. With my primary VPN service provider, PrivateTunnel, via Chicago server and AES-128-GCM cipher I get full throughput, 200/20. I tried Mullvad VPN service (3-hour trial) via Chicago server last weekend out of curiosity with AES-256-GCM cipher and was still able to achieve full throughput, 200/20.

Additional OpenVPN performance numbers from RMerlin…

https://www.snbforums.com/threads/openvpn-performance-of-the-rt-ac86u.41217/

Wow!
I am on a 200/80 fiber connection and the fastest speed I got to date, using the WRT3200ACM (with its 1.8 GHz CPU) & DD-WRT OpenVPN is ~130/80, while on the PC I can get the full 200/80, so it seems this router is THE router for OpenVPN performance!

 

[dvohwinkel]

@RMerlin : Can you let us know if there is anything we can do to apply pressure or help you get things changed to make supporting the RT-AC86U and other models easier. I know you are waiting to hear back from ASUS.. but please let us know if we can help.

Regards

-Dave

 

[RMerlin]

@RMerlin : Can you let us know if there is anything we can do to apply pressure or help you get things changed to make supporting the RT-AC86U and other models easier. I know you are waiting to hear back from ASUS.. but please let us know if we can help.

Thanks for offering, but there isn't much users can do about it. Ultimately, it will depend on whether there's any way for Asus to provide me a way around this without having to break their licensing agreement with Broadcom and/or investing a significant amount of engineer time working on something that ultimately is only for a third party.

Asus's R&D probably already has their hands quite full with the recent WPA2 issues and something like 30-ish router models (plus all the wifi clients) in need of update, so it might take some time before I hear back from them one way or another.

At least I have a potential way around that I want to investigate in the coming days (splitting large nvram values into multiple small ones).

 

[dvohwinkel]

Thanks for offering, but there isn't much users can do about it. Ultimately, it will depend on whether there's any way for Asus to provide me a way around this without having to break their licensing agreement with Broadcom and/or investing a significant amount of engineer time working on something that ultimately is only for a third party.

Asus's R&D probably already has their hands quite full with the recent WPA2 issues and something like 30-ish router models (plus all the wifi clients) in need of update, so it might take some time before I hear back from them one way or another.

At least I have a potential way around that I want to investigate in the coming days (splitting large nvram values into multiple small ones).

Is the nvram limitation in the SDK or is it specific to the RT-AC86U only? seems like a lib like that would be shared code among different models or anything using the 382 codebase. A kind of a "why reinvent the wheel".

 

[Razor237]

I was going to try this out on my RT-AC3100 , since this is a big change i should to a fresh setup instead of just updating keeping my settings i have already . Is my thinking on this correct ?