[Beta] Asuswrt-Merlin 384.16 beta (and 384.13_5) are available

[Ozymandias]

RT-AX56U owners, what does the following command return?

 nvram get led_pwr_gpio

4112

 

[GaselK]

RT-AX56U owners, what does the following command return?

 nvram get led_pwr_gpio

4112

 

[Hawk]

Eric, I noticed strange behaviour with guest network, whenever new clients join it show that it is not associated with guest network, though it shows under list. When new client joins then it shows 1 at the end of Flags column.

In this screenshot first X.X.X.164 joined at that time there was no 1 in the flags column, then came 92 and 164 started showing flags, however 92 was not showing till 62 and with 62 92 shows flags correctly.
Basically in simple term, which ever latest device to join guest network, it doesn't show correct flags.

 

[eddiez]

Well, at least the .16 beta resolved my IPV6 issue (IPV6 address visible but no connection to be made)

Running in Media Bridge mode.

 

[Simon W]

Possible issue with the firmware update check: "Unable to connect to the update server."

(AX88U)

 

[maxbraketorque]

I just realized that something about beta1 is preventing my computer from making a Cisco AnyConnect VPN connection to my employer network. The AnyConnect client times out with a message saying that it can't find the VPN server. I was able to login immediately by tethering my laptop to my cell phone.

 

[vdemarco]

I just realized that something about beta1 is preventing my computer from making a Cisco AnyConnect VPN connection to my employer network. The AnyConnect client times out with a message saying that it can't find the VPN server. I was able to login immediately by tethering my laptop to my cell phone.

under

Wan->"NAT Passthrough"

there are settings for PPTP,L2TP, IPSec passthrough (mine are all set to enabled)

I have a VPN connection to work and its good.

 

[RMerlin]

Eric, I noticed strange behaviour with guest network, whenever new clients join it show that it is not associated with guest network, though it shows under list. When new client joins then it shows 1 at the end of Flags column.

In this screenshot first X.X.X.164 joined at that time there was no 1 in the flags column, then came 92 and 164 started showing flags, however 92 was not showing till 62 and with 62 92 shows flags correctly.
Basically in simple term, which ever latest device to join guest network, it doesn't show correct flags.

It does show as being under the guest client list however. The flag is supposed to be removed once a client is shown, so the issue is probably that the flag doesn't properly get removed.

 

[RMerlin]

Possible issue with the firmware update check: "Unable to connect to the update server."

(AX88U)

Check your DNS settings. There was no change to the update check code in a long time.

 

[jsbeddow]

I just realized that something about beta1 is preventing my computer from making a Cisco AnyConnect VPN connection to my employer network. The AnyConnect client times out with a message saying that it can't find the VPN server. I was able to login immediately by tethering my laptop to my cell phone.

Any luck resolving this yet? (without rolling back to a previous release?) I also use this (Cisco AnyConnect) and don't want to take the risk if 384.16 Beta 1 is truly causing issues (no problems here on Alpha2). Keep us posted.

 

[maxbraketorque]

Any luck resolving this yet? (without rolling back to a previous release?) I also use this (Cisco AnyConnect) and don't want to take the risk if 384.16 Beta 1 is truly causing issues (no problems here on Alpha2). Keep us posted.

After logging in from my cell phone network, I can now log in via my home network, so all ok for now. Perhaps it was a hiccup in my AC86U DNSSEC setup (everything set to as strict of conditions as possible).

 

[RMerlin]

Debugging support for the RT-AX56U is next to impossible without having a router in hand, so I'll order one (graciously offered by a generous donor). I should get it in a couple of days.

 

[Simon W]

Check your DNS settings. There was no change to the update check code in a long time.

Seems this is on the right lines, though I think it's highlighted a bug elsewhere, as follows:

I briefly enabled/disabled IPv6, a number of times using a variety of options, including static and the setting of a manual DNS server (all via the interface I should add) = quickly realised IPv6 was a non-starter because my ISP doesn't currently offer an IPv6 service! So it was then set back to disabled.

Despite this, your host (asuswrt-merlin.net) is still being served up an IPv6 address from/via the routers internal DNS server, demonstrated by both the built-in Network Tools nslookup (both IPv4 and IPv6 addresses listed, whereas a handful of other random domain names I try just give IPv4 addresses) and also doing a nslookup from an attached client (which errors with a SERVFAIL message, but again only for your host).

Disabling DoT (Cloudflare) allows your host to be resolved from a client, as does re-enabling DoT using another provider (Google). Yet in neither case is the router able to access the upgrade server via the interface, nor does adjusting the WAN DNS server settings from Cloudflare to Google. Further, reselecting DoT via Cloudflare reintroduces the error back to the clients again.

EDIT: actually, do Google's DNS servers support DoT? Perhaps not, in which case it's a more straight-forward "if DoT is enabled it doesn't work from a client".. though regardless it doesn't work from the update interface which got me looking at this in the first place.

TLDR; seems IPv6 hasn't been fully disabled somewhere

 

[SheikhSheikha]

RT-AC5300: System Log does not Auto Refresh

 

[Rik Stigter]

Anybody else seeing short spikes on the CPU-usage front? My AX88 has short spikes on core-4. Top/Htop tells me its samba related (process is /usr/sbin/smbd -D -s /etc/smbd.conf). They occur every ~25 secs.

It's not a big issue, but I wasn't seeing this on the AX58 in the 'same' configuration but that might be because on a three core system these spikes don't become noticeable because they are mixed with other processes in the core usage.

 

[XIII]

I just realized that something about beta1 is preventing my computer from making a Cisco AnyConnect VPN connection to my employer network.

I have to work from home due to the Corona virus, but for me Cisco AnyConnect VPN works fine with 384.16 Beta 1.

 

[FTC]

Anybody else seeing short spikes on the CPU-usage front? My AX88 has short spikes on core-4. Top/Htop tells me its samba related (process is /usr/sbin/smbd -D -s /etc/smbd.conf). They occur every ~25 secs.

It's not a big issue, but I wasn't seeing this on the AX58 in the 'same' configuration but that might be because on a three core system these spikes don't become noticeable because they are mixed with other processes in the core usage.

No spikes at all here and also using samba in my ax88.

 

[Simon W]

Seems this is on the right lines, though I think it's highlighted a bug elsewhere, as follows:

<snip>

TLDR; seems IPv6 hasn't been fully disabled somewhere

Several hours later = clients regained access (IPv4 name resolution) of asuswrt-merlin.net even with DoT (Cloudfare) enabled.
Several further hours later = router now can successfully check for updates.

= a) coincidental resolution problem? or b) more likely a case of a cached record somewhere ultimately ageing (I had tried to prompt force this with multiple reboots though to nil effect).

Had planned to reset and manually reconfigure the router today (not being confident there's not something left over, and with no need for IPv6 for sometime according to my ISP) though now everything seems ok I might hold off until a later beta/release.

 

[Rik Stigter]

No spikes at all here and also using samba in my ax88.

It stopped all by itself.
Just a glitch then.

One more thing: I had to go back to stock on my AX58U to get it to accept that I wanted it to be an AIMesh-node. Again: no big issue there but be prepared if you want to do the same and have the stock-firmware at hand (although you'd have the AIMesh-router available to download it, of course!).
I'll leave it at that for a while before I upgrade both to 384.16, as soon as that one is available of course.

 

[Makaveli]

It stopped all by itself.
Just a glitch then.

One more thing: I had to go back to stock on my AX58U to get it to accept that I wanted it to be an AIMesh-node. Again: no big issue there but be prepared if you want to do the same and have the stock-firmware at hand (although you'd have the AIMesh-router available to download it, of course!).
I'll leave it at that for a while before I upgrade both to 384.16, as soon as that one is available of course.

I believe this is known.

Someone else running AI mesh can chime in.

From what i've read here only one of the routers should be using merlin firmware and the rest should be on stock when using an AI mesh setup.