Beta Asuswrt-Merlin 386.9 beta is now available for AC models

[Tech9]

@ugandy, I see you put a Like on my post with a question. Do you see the same thing I'm seeing? It's like page load delay or parts of the page load with visible delay. Not always, sometimes. It's hard to reproduce due to DNS caching and you have to clear it or load different pages all the time, but with DoT disabled everything is smooth and consistent and it never happens. I was thinking server load because Quad9 is sometimes slow here, but OpenDNS and Google have fast and many servers around. They respond instantly with or without DoT.

 

[RMerlin]

t seems that these are the culprit.

Those highlighted are barely using any CPU. The acsd daemon is the one that seem to be stuck in that second screenshot. That's Broadcom' automatic channel selection daemon.

 

[Subkulture]

Hi guys, running the new BETA on my AC86 since 4:00pm yesterday though notice my Skynet was not running stating the following errors in log -

/jffs/scripts/firewall: line 1: arithmetic syntax error​

&
Lock File Detected (could not find exact line of error)

Using SHH found Skynet menu would not load either so reinstalled which fixed menu but still brought errors so deleted and re-created SWAP file which fixed issue.

 

[nathanb]

Hi everyone, There has been something strange happening since I installed this beta. A process is running a search on the /tmp folder, which spins up my connected disk and searches its contents continuously. It is searching for "crrt*". The command being run is "/usr/bin/find /tmp -iname crrt*"

I have carried out a full clean install, but the process still shows up and continuously searches the disk. I did not experience this on 386.7. Does anyone have the same issue?

I have attached a screenshot of the running processes.

 

[Frank Monroe]

I had to revert back to version 386.3_2 as 386.9 Beta also has the ARP/ICMP issue. As it stands today, every release after 386.3_2 has the issue.

 

[Clark Griswald]

AC86U in AP mode with zero issues!

 

[5stringdeath]

RT-AC3100 working fine, running quiet. I'll be sad when this router is EOL'd it has truly been the best I've owned.

 

[Wrkdbf_Guy]

RT-AC3100 working fine, running quiet. I'll be sad when this router is EOL'd it has truly been the best I've owned.

I feel the same way about my venerable RT-AC68U router(s). Like the Energizer Bunny ads.

 

[skeal]

RT-AC3100 working fine, running quiet. I'll be sad when this router is EOL'd it has truly been the best I've owned.

Can you point me to the post or what ever that points out the units scheduled for EOL please?

 

[Martinski]

... A process is running a search on the /tmp folder, which spins up my connected disk and searches its contents continuously. It is searching for "crrt*". The command being run is "/usr/bin/find /tmp -iname crrt*"
...

I have attached a screenshot of the running processes.
...

Based on your screenshot of the "top" command, it seems that "/usr/bin/find /tmp -iname crrt*" is a child process of the ASUS "Security Daemon" (asd).

See some info about "asd" on this thread here.

What is "asd" process?

Hi guys, I am wondering what is the "asd" process, which produces the following log in /jffs/asd.log: 1525496708[register_feature] chknvram registered 1525496708[register_feature] misc registered 1525496708[init] Load existing signature files. 1525496708[update_signature_in_feature] Update sig...

www.snbforums.com

 

[ugandy]

@ugandy, I see you put a Like on my post with a question. Do you see the same thing I'm seeing? It's like page load delay or parts of the page load with visible delay. Not always, sometimes. It's hard to reproduce due to DNS caching and you have to clear it or load different pages all the time, but with DoT disabled everything is smooth and consistent and it never happens. I was thinking server load because Quad9 is sometimes slow here, but OpenDNS and Google have fast and many servers around. They respond instantly with or without DoT.

i'm experiencing this too. using 9.9.9.11
doesn't happen often enough to annoy me, but it didn't happen until i enabled DoT. it does not load the page for a bit, and it looks like it's resolving the new address.
also didn't happen when i was using unbound

 

[Smokey613]

For the past two days, I have been running my network through one of my RT-AC86U with this Beta firmware loaded.

After loading the Beta, I did a full reset and setup everything manually, except for the YazDHCP entries.

I restored those from a YazDHCP backup.

I am running it in Dual WAN Load Balancing mode with a 1 to 1 load balance.

I have created routing rules to make sure my Plex server goes out my 100/100 fiber and receives traffic over the same fiber connection.

All other devices are utilizing the load balancing.

I am running Skynet, scribe, scMerlin, uiScribe, YazDHCP and AdGuard Home

So far, it has been running without any issues and I am still amazed how this "obsolete" model has better wifi coverage than anything else I have tried except the GT-AX6000. But to be honest, I think it is a very small difference.



Now to be clear, the TP-Link ER605, the 2 eero Pro and the FWB+ are still up and operational in case I need to fallback to that setup.

To switch back I only need to move 1 network cable, change the eero wifi SSID back to my network SSID and turn off the wifi radios on the RT-AC86U.

This setup on the AC86U had been working so well, I postponed sending the AX86U back. I decided to load 388.1 on it and recreate my AC86U setup on the AX86U to see how it performs. After 24 hours I must say I am pleasently surprised. It has been working without any issues.

I still have until Jan 31st to return it, if I want to.

BTW, the AX86S was shipped back.

 

[capncybo]

This setup on the AC86U had been working so well, I postponed sending the AX86U back. I decided to load 388.1 on it and recreate my AC86U setup on the AX86U to see how it performs. After 24 hours I must say I am pleasently surprised. It has been working without any issues.

I still have until Jan 31st to return it, if I want to.

BTW, the AX86S was shipped back.

Sooooo, In your opinion is there "currently" more (bugs/gremlins) with the AX86S hardware/software combo than the older more mature AX86U? Perhaps many of the recent issues are actually more exclusive to the newer hardware?

 

[n3cr0shark]

Just to report, I was able to dirty flash my meshed RT-AC5300 to the 386.9 Beta1 without issue and it's staying stable.

When I tried the alpha, dirty flashing it would say it was successful but it never actually uploaded the new code or changed the reported release version. I ultimately had to remove the RT-AC5300 from the mesh, default it, update it, physically power cycle it, then readd it to the mesh.

So positive report on the beta here.

 

[Tech9]

After 24 hours I must say I am pleasently surprised. It has been working without any issues.

AX86U with 388.1 does not pass my Wi-Fi tests. I've tried twice with clean start. I wish it was working as well as 386.9 on AC86U.

Perhaps many of the recent issues are actually more exclusive to the newer hardware?

AX86S has the same radios, the same drivers and runs the same firmware. I would expect the same behavior on both AX86U/S.

 

[capncybo]

I feel bad for all those who are experiencing unreliability with their AX86U/S... twasn't very long ago that there was "almost" a consensus within these forums & that WAS...
The AX86U was probably the most reliable & best value for the $$
Hopefully, it will get resolved, soon.

EDIT: It sucks when you do your (research/homework) & technology still screws you over.

 

[Tech9]

Best value if you don't know anything better and can live with perpetual beta firmware.

 

[capncybo]

Well (as you know all too well) there is no IDEAL one-solution for ALL. However, there used to be some typical usage for home networking environments. Ex) Less than 25 clients & maybe Max ISP speeds of say 250Mbps. But alas as these routers try to handle more & more clients @ faster & faster ISP speeds...
They are ill-equipped for the task.
Thanks to the IOT a household can easily have like 50+ clients & with 1G-ISP... these poor lil routers LOL

 

[bennor]

Folks lets keep things on topic and not drag the discussion into routers not supported by this beta firmware. Plenty of other threads to discuss the problems with the AX86U/S and it's firmware.

 

[JohnD5000]

Folks lets keep things on topic and not drag the discussion into routers not supported by this beta firmware. Plenty of other threads to discuss the problems with the AX86U/S and it's firmware.

OK, here are some lines from the log file that I'm not sure of running on the beta. Anything that needs to be looked into?

Jan 3 04:31:11 kernel: nf_conntrack: automatic helper assignment is deprecated and it will be removed soon. Use the iptables CT target to attach helpers instead.
Jan 3 08:36:24 dnsmasq[22283]: possible DNS-rebind attack detected: googlecm.hit.gemius.pl
Jan 3 14:00:13 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 3 20:50:48 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 3 21:33:58 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 3 23:06:28 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 4 00:22:53 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 4 08:16:48 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 4 09:25:14 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 4 14:42:48 dnsmasq[22283]: possible DNS-rebind attack detected: pool.york.iponweb.net
Jan 4 14:42:48 dnsmasq[22283]: possible DNS-rebind attack detected: pool.york.iponweb.net
Jan 4 17:08:50 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232
Jan 4 17:58:36 dnsmasq[22283]: reducing DNS packet size for nameserver 127.0.1.1 to 1232