[Fork] Asuswrt-Merlin 374.43 LTS releases (Archive)

[john9527]

Still the older problem with DNSMasq and ab-solution: DNSMasq is using almost 100% CPU on my RT-N16.

Sorry, but I'm not willing to invest any time here unless you can recreate the problem without your 3/4 million host entries.

 

[jrmwvu04]

Flashed 26BA, upnp behavior is back to the way I was used to in 25 and before. The port mappings are persisting even 16+ hours after the boxes are powered off.

 

[john9527]

Last call for any 'show stopper' issues on the 26 Beta
Thanks to everyone who gave it a whirl....

 

[jrmwvu04]

A new option to clear existing miniupnpd leases

Where is this? I haven't seen it in my searching. Other than that, no issues on my setup. I'll admit I've been quite busy though and haven't tested/tinkered much lately. Thanks for your effort.

 

[ColinTaylor]

Where is this? I haven't seen it in my searching.

WAN > Internet Connection > Basic Config > UPNP: Clear existing leases

 

[jrmwvu04]

WAN > Internet Connection > Basic Config > UPNP: Clear existing leases

Ah there it is. Many thanks.

 

[Chazall1]

Updated to 3.0.0.4.374.43_2-26B6j9527 a few days ago. no issues on my end. Thanks

 

[jrmwvu04]

upnp release works, but is on a really long timer, close to the full reboot timer. After it got to 20% I saw my lights were normal and I reloaded the GUI with no problem. Is there a reason for the long timer?

 

[ColinTaylor]

Ah there it is. Many thanks.

Although I've just tried it for the first time and it just seems to restart the WAN interface rather than doing anything clever. Correct @john9527 ? The WAN outage time on my system was ~10 seconds, although the web interface has an 80 second timer.

No issues found though.

 

[jrmwvu04]

Although I've just tried it for the first time and it just seems to restart the WAN interface rather than doing anything clever. Correct @john9527 ? The WAN outage time on my system was ~10 seconds, although the web interface has an 80 second timer.

No issues found though.

Mine was 40s all told having run it a couple of times. It's fine like it is, I think. In any event it's certainly not a show stopper. Thanks John.

 

[john9527]

Mine was 40s all told having run it a couple of times. It's fine like it is, I think. In any event it's certainly not a show stopper. Thanks John.

Programmers dilema I wanted to keep it with the rest of the upnp options, so it ends up on a page that restarts the wan (and sets a flag to delete the leases file on the restart). It's overkill, but in the slimmed down webserver it's a real pain in the *ss to try and determine what mix of options have been selected and to change the action to be taken.

 

[RMerlin]

Programmers dilema I wanted to keep it with the rest of the upnp options, so it ends up on a page that restarts the wan (and sets a flag to delete the leases file on the restart). It's overkill, but in the slimmed down webserver it's a real pain in the *ss to try and determine what mix of options have been selected and to change the action to be taken.

I figured out a simple way to do this recently. Basically:

if ( document.form.some_settings.value != "<% nvram_get("some_settings"); %>" ) {
   document.form.restart += ";restart_thisservice";
}

This works nicely because the nvram_get() is processed before the HTML page is sent to the browser, so it will always contain the original value.

 

[john9527]

I figured out a simple way to do this recently. Basically:

That works well when you have one (or a couple) of items that require a 'bigger' reset procedure. When you have one that can be done with a 'smaller' reset, you would still need to check all the other possible selections to see if there is one that would override the small action.

In this case, hopefully it's not going to be a heavily used option.

The other possibility is to make it an 'exclusive' option, with a warning that any other selections would be ignored/lost.

 

[john9527]

In search of a brave man (or woman)

After a weeks worth of work, I have a development fork release that should support the AC68U rev C1 and E1. It's loaded on my AC68P router right now as I type this message. Unfortunately, I don't have a later rev AC68 to test.

So, looking for someone with a later rev AC68 to give it a try to see if I should continue. You will need to load the code via the Firmware Recovery tool just like the current fork supported routers, followed by a factory reset and reconfigure. You should also be prepared to use the FW recovery tool to get back up and running on your previous code if something goes wrong (If it doesn't work, it's likely the router will not complete boot or will enter a boot loop).

Please respond via PM if you are willing and able to help. Thanks!

 

[cybrnook]

In search of a brave man (or woman)

After a weeks worth of work, I have a development fork release that should support the AC68U rev C1 and E1. It's loaded on my AC68P router right now as I type this message. Unfortunately, I don't have a later rev AC68 to test.

So, looking for someone with a later rev AC68 to give it a try to see if I should continue. You will need to load the code via the Firmware Recovery tool just like the current fork supported routers, followed by a factory reset and reconfigure. You should also be prepared to use the FW recovery tool to get back up and running on your previous code if something goes wrong (If it doesn't work, it's likely the router will not complete boot or will enter a boot loop).

Please respond via PM if you are willing and able to help. Thanks!

RT-AC1900P work for this test?

 

[john9527]

RT-AC1900P work for this test?

Yes.....it should also cover the AC1900P. You game? Will send you a link to the code just in case.

 

[cybrnook]

You game?

Yep

 

[ColinTaylor]

I have just discovered a MASSIVE bug. I don't know how I didn't spot it before. I'm amazed my router works at all.

On the Wireless Log page if the Connected time >= 100 hours the Flags column is misaligned.

 

[john9527]

On the Wireless Log page if the Connected time >= 100 hours the Flags column is misaligned.

OH NO!!!

I'll tweak it to allow for 9999 hours....416 days (I'm an optimist on the potential fork uptime )

 

[snapieee]

john9527, I would like to thank you for your build. All reviews on it is rock solid.

This is my first post on this forum so forgive me If I am posting in the wrong thread.

The issue I am having: I have Asus TM-AC1900, I bought it new last black friday when it was on sale for $60. I already converted it to RT-AC68U running bootloader 1.0.2. I had merlin's build on it since I got it. I kept up with Merlin's update. I run VPN on that router and I found the most stable version with merlin's is 380.66_4. With 380.66_4 I can keep the VPN connection alive for extended period of time, with later version of merlin's The VPN connections drops, even though it is showing connected in the VPN status page after I login to the router. I also lose internet access, nothing works at all. In another word, I don't have WAN access and no internet surfing. LAN access is fine. If I turned the VPN off/on without rebooting the router, the WAN gets restored again. On the other hand, I have another TM-AC1900, bought on the same deal, setup as the main router with the latest Merlin's 380.66_6. On that other main router I have enabled Trend Micro network protection.


After reading about your fork, I pulled the plug and I installed V25E8 on the other router that I have strictly for VPN. It worked fine, nice wireless coverage and speed. I always get my ISP advertised speed 55+ to 60 mb when wired. If I am connected via wifi to the VPN router the speed sometimes matches my ISP, and other times I get less than 20mb. This VPN router is overclocked to 1400,800 "CPU, Memory respectively" with active cooling. Temp when it is really overloaded with vpn traffic, watching YouTube and downloading in the background it never went above 62F, normal temp is 56f-58f. That was issue one with V25E8.


Issue two with your fork V25E8, if I logged on to the router and minimized the window, then forget about it for longer time It never logout due to inactivity even though it is enabled in the Administration>System for 30min. I changed to 20 min then to 15 min , and I still have the same issue.


I pulled the plug again and installed V26BA hoping to resolve the inactivity logout issue. Still having same issue. With V26BA, If I logged out manually, that page will freeze. Even after closing all opened windows of the browser "Chrome", reopening brand new window, I cannot get to the login page again even after clearing cache/cookies. Even tried to login to the router another workstation/ipad/iphone it just won't let me in. No errors, it is just blank page. I just enter 192.168.2.1, since this is a secondary router connected via Ethernet to the main one, progress bar shows the web page is trying to load but nothing loads. I have to wait a few minutes then it will let me in.


So, on the vpn router now I reverted back to merlin build 380.65_4 after I read that it is the most stable build. If I could how to resolve the issue with V25E8 or V26BA, I will definitely give it another try. By the way, even with merlin's 380.65._4 the logout due to inactivity still there but I installed just for stability issue and no vpn/wan drops.


Just to summarize how everything is connected: Cable modem > cat6 to the main asus tm-ac1900 router with merlin's 380.66_4>shielded cat6a cable to gigabit switch>then cat6 from the switch to the second asus tm-ac1900 "vpn router" currently with merlin's 380.65_4. Bother routers are overclocked at 1400,800 "cpu, memory respectively" with active cooling on both of them.


Any thoughts on that? I would really appreciate your input and help.