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

[john9527]

I had 3 occurrences today so decided to take a baby step to test with. The AC66 (MB) in the middle I loaded Merlin's latest fork, reset completely using the power on with WPS button method, and configured from scratch. So far, so good, and it is neat to be able to configure from the Android App, but time will tell.

Starting with around the 378 code series, ASUS did a LOT of changes to the Media Bridge support particularly in handling multicast traffic. There may be some other change that is helping out in your config, although nothing that I know of specifically with DHCP.

 

[penguin22]

Starting with around the 378 code series, ASUS did a LOT of changes to the Media Bridge support particularly in handling multicast traffic. There may be some other change that is helping out in your config, although nothing that I know of specifically with DHCP.

Interesting, so it is possible that using the Merlin firmware, which is on 380.64 is worthwhile for the media bridge router scenario. It keeps me on the custom firmware, which is beneficial for other needed functionality, and has changes (hopefully improvements, especially with the issue I was experiencing) that may help usage in general.

This is an AC66 running the BCM5300; I don't believe the 378 fork and driver gives benefit over the 380 fork and drivers, at least in MB configuration, but could be incorrect. Anyway, thanks for the support and I'll update in a day or so if the problem did indeed disappear.

 

[adrianmak]

is AC58U not support in Asus Merlin firmware?

 

[joegreat]

is AC58U not support in Asus Merlin firmware?

Answered here!

 

[adrianmak]

Answered here!

I had read this post. And I was curious why just skip this model.

 

[john9527]

I had read this post. And I was curious why just skip this model.

Asuswrt-Merlin only supports models with Broadcom wireless. The AC58U is RALink.
Also, note that this fork only supports a subset of the models supported by the latest Merlin.

 

[maurer]

I had read this post. And I was curious why just skip this model.

if you want alternative firmware try LEDE (openwrt fork):
https://github.com/chunkeey/LEDE-IPQ40XX/releases

 

[Trebuin]

Starting with around the 378 code series, ASUS did a LOT of changes to the Media Bridge support particularly in handling multicast traffic. There may be some other change that is helping out in your config, although nothing that I know of specifically with DHCP.

I use my AC66U in mediabridge with your firmware tying to a AC3200 using Merlin's 380.64_1 firmware & on the lower channels of the 5GHz. I shut off the AC66U's 2.4GHz since that isn't designed to be used in media mode (unless I'm misunderstanding something). I've had instability in the past because my router was an airport, but since I changed over to the ASUS's AC3200, it has improved. Would I gain any benefit going to a newer firmware on the AC66U? I'm looking exclusively at speed & stability.

Eventually, I will move over to a Mesh to cut the number of routers...but that will happen when they get USB 3.0 with decent read/write speed.

 

[penguin22]

I use my AC66U in mediabridge with your firmware tying to a AC3200 using Merlin's 380.64_1 firmware & on the lower channels of the 5GHz. I shut off the AC66U's 2.4GHz since that isn't designed to be used in media mode (unless I'm misunderstanding something). I've had instability in the past because my router was an airport, but since I changed over to the ASUS's AC3200, it has improved. Would I gain any benefit going to a newer firmware on the AC66U? I'm looking exclusively at speed & stability.

Eventually, I will move over to a Mesh to cut the number of routers...but that will happen when they get USB 3.0 with decent read/write speed.

I only had one anomaly in the past day plus, which I am not sure was the same DHCP issue (time will tell). I did also move my AC68P to the Merlin fork and have some benchmark numbers that show an improvement in performance overall. That said, I am still using John's fork on my N66 to boost the signal strength on 2.4GHz 20MHz for maximum range.

John, do you think the AC66 and newer routers have less of a need for the 378 driver compared to the older N66? This is ignoring all the other tweaks and specifically from a performance perspective that I ask.

 

[Chrysalis]

Nice changes John, so you fixed the broken asus QoS code.

I know you didnt like my config, but bear in mind most of my config was to

(a) have working downstream QoS, which you now fixed.
(b) dnscrypt which you also now implemented.

Did you get the idea to fix downstream QoS after you added the extra iptable marks last year?

 

[john9527]

I use my AC66U in mediabridge with your firmware tying to a AC3200 using Merlin's 380.64_1 firmware & on the lower channels of the 5GHz. I shut off the AC66U's 2.4GHz since that isn't designed to be used in media mode (unless I'm misunderstanding something). I've had instability in the past because my router was an airport, but since I changed over to the ASUS's AC3200, it has improved. Would I gain any benefit going to a newer firmware on the AC66U? I'm looking exclusively at speed & stability.

Eventually, I will move over to a Mesh to cut the number of routers...but that will happen when they get USB 3.0 with decent read/write speed.

It may be worthwhile to give the later level a try. When I first dug into this, it was with someone using a router in AP mode and having trouble with Apple devices (there is some shared code between AP and MediaBridge). It worked on the later Merlin level. I started to look at backporting some of the multicast fixes, but it quickly became too involved to really support.

 

[john9527]

John, do you think the AC66 and newer routers have less of a need for the 378 driver compared to the older N66? This is ignoring all the other tweaks and specifically from a performance perspective that I ask.

I assume you are asking about my fork 374 level vs later levels. On the ARM based routers, I periodically check wireless performance on my AC68 (fork vs new), and really don't see any significant difference. On the MIPS based routers (of which the AC66 is one, except for the new AC66U-B1) I haven't done any detailed measurements, but my guess would be that the fork would have an edge on the wireless.

 

[penguin22]

I assume you are asking about my fork 374 level vs later levels. On the ARM based routers, I periodically check wireless performance on my AC68 (fork vs new), and really don't see any significant difference. On the MIPS based routers (of which the AC66 is one, except for the new AC66U-B1) I haven't done any detailed measurements, but my guess would be that the fork would have an edge on the wireless.

You are correct in that I meant 374. My initial tests showed a peak transfer speed of 70Mbps, which was about 15Mbps quicker than my prior tests, however it could have been other factors involved as well. I haven't been able to hit that and am closer to my prior tests since.

I am more concerned with stability for what I am doing as opposed to sheer performance as the signal from the N66 is quite far from the AC66 (thus 20MHz and only 2.4GHz). The tandem of your firmware on the N66, also running AB in JFFS and pixelserv, all feeding the MB + AP means I get distance, stability, and no sacrifices from overlap.

Funny part is that mesh networks are now becoming a thing and I really wonder how they handle the logic to prevent overlap affecting connectivity and performance. I'll sit here and be happy though so long as I keep getting DHCP .

 

[Trebuin]

It may be worthwhile to give the later level a try. When I first dug into this, it was with someone using a router in AP mode and having trouble with Apple devices (there is some shared code between AP and MediaBridge). It worked on the later Merlin level. I started to look at backporting some of the multicast fixes, but it quickly became too involved to really support.

Do you mean move over to the updated firmware?
edit:
Specifically, Merlin's newer firmware over your work?

 

[Santiago C]

Hi @john9527, happy new year! Again your DualWAN (only?) user reporting something weird

wanduck.c, line 327, int do_ping_detect(int wan_unit), is the code that does the ping and checks for the existence of the resulting file as a flag.

Thing is, the ping is done always from the default route, no matter what wan_unit the function had as a parameter

sprintf(cmd, "ping -c 1 %s >/dev/null && touch %s", wandog_target, PING_RESULT_FILE);

line 352

This is transparent in case the main WAN unit is working, but after failover, when the router checks for connectivity on the main unit, ping requests are going to be send from the secondary unit which acts as the default route, am I right?

/bin/ping has the "'-I" parameter to specify the interface to be used, I had no luck trying to enter either wanX_ifname not wanX_pppoe_ifname (my wan1 is pppoe) but it seems to work with wanX_ipaddr although that might need an additional check in the code for cases were the return value is 0.0.0.0 since that makes ping come out of the default route as well

I am also facing an issue where the failback count is not respected. I have wandog_interval=5 and wandog_fb_count=120 which should result in 10 minutes before failing back, but I have seen scenarios were it does fail back much earlier. I think "set_disconn_count(other_wan_unit, S_IDLE)" should be added before failover.

Regards,

S

--

 

[john9527]

Happy New Year to all!
It's been a bit longer for this release, but it contains some significant new functions....a re-written QoS and support for DNSCrypt. Hope everyone finds them useful additions.

Please take the time to read through the first post for details and some usage notes.

Special thanks to @ColinTaylor for helping during the development of the QoS changes.

LATEST RELEASE: Update-22E2
7-January-2017
Merlin fork 374.43_2-22E2j9527
Download http://bit.ly/1YdgUcP
============================

Following are the major changes (full changelog is in the zip files)

Update-22E2 Highlights

• Security Updates including some from ASUS, Merlin backports and OpenVPN
• NEW - A completely rewritten Traditional QoS
  This traditional QoS implementation should work more like one would expect based on the QoS rules, including Download/Upload limits and priority levels
• NEW - Support for DNSCrypt, including integration with OpenVPN clients
• Improvents to OpenVPN start and stop sequences, including a fix to allow use of explicit-exit-notify when specified by the VPN provider
• A new option for strict DNSSEC enforcement
• Improved iptables mark support for the MIPS based routers
• Misc fixes for Dual WAN failover and DDNS update
• Misc backports from ASUSWRT-MERLIN

As always, a reminder to users with MIPS routers to have a backup of /jffs in case the jffs space needs to be reformatted due to increases in firmware size.

SHA256

61eb7529e80295ce73d83d99fb05b1bf680e2622452873024b86a28ae39e4ca9  RT-AC68U_3.0.0.4_374.43_2-22E2j9527.trx
c8470c62a96769724358947d95ecdaee4121327dcd6140bdf7f48bb59b6f65f6  RT-AC56U_3.0.0.4_374.43_2-22E2j9527.trx
287951e0040d1b4a4da37391429d754b737e99e9c5eb6c75a26fcae38debd2bd  RT-N16_3.0.0.4_374.43_2-22E2j9527.trx
e4c6c570d308ce719b7fb295bfbeb80565e8c72aa9695a374e62b86f963307e9  RT-AC66U_3.0.0.4_374.43_2-22E2j9527.trx
330b1e447fd26820c8f4a96d80178e506901cb02f4cdfaae746e344f36f4eb03  RT-N66U_3.0.0.4_374.43_2-22E2j9527.trx

 

[john9527]

Do you mean move over to the updated firmware?
edit:
Specifically, Merlin's newer firmware over your work?

It depends....I always want to try and help everyone to get the best setup for their environment. If you are having disconnect issues or having trouble with some services (particularly Apple) when in AP or Media Bridge mode, Merlin's latest code may be worth a try. It has some enhancements that are just too involved for me to try and implement.

 

[john9527]

Hi @john9527, happy new year! Again your DualWAN (only?) user reporting something weird

Always appreciate good debug help! Especially when I cannot set up the environment. I'll definitely take a look at what you found.

PS - There should be fix in 22E2 for the ping detect issue you had previously found. Should now work regardless of the redirect setting.

 

[jrmwvu04]

I have been using (and testing) traditional qos quite extensively - it's fantastic. Thanks for the great work, just finished putting the final release on.

 

[Trebuin]

It depends....I always want to try and help everyone to get the best setup for their environment. If you are having disconnect issues or having trouble with some services (particularly Apple) when in AP or Media Bridge mode, Merlin's latest code may be worth a try. It has some enhancements that are just too involved for me to try and implement.

Now for the big question, with the new restrictions that is being built in...can I go back to your firmware using the ASUS recovery method if I'm unable to use Merlin's?