[Release] Asuswrt-Merlin 384.7 is now available

[RMerlin]

UPnP issues probably come from this miniupnpd commit, if you are behind a double NAT setup:

https://github.com/miniupnp/miniupnp/commit/8e10a1aeab9b8cd4d3b2e964b02e9ad409cf3aaa

I opened an issue with the author, but in the mean time I will probably have to revert that change.

Ping @john9527

 

[e_lasman]

Make sure you use your username and password, not your email or DDNS key.

Don't get me wrong but maybe then it is worth to update field titles

User Name or E-mail Address
Password or DDNS Key

or they're shared across providers?

 

[john9527]

UPnP issues probably come from this miniupnpd commit, if you are behind a double NAT setup:

https://github.com/miniupnp/miniupnp/commit/8e10a1aeab9b8cd4d3b2e964b02e9ad409cf3aaa

I opened an issue with the author, but in the mean time I will probably have to revert that change.

Ping @john9527

I went to do a revert, and found they changed that patch. New error test is

        } else if (addr_is_reserved(&addr)) {
            syslog(LOG_INFO, "Reserved / private IP address %s on ext interface %s: Port forwarding is impossible", if_addr, ext_if_name);
            syslog(LOG_INFO, "You are probably behind NAT, enable option ext_perform_stun=yes to detect public IP address");
            disable_port_forwarding = 1;

So maybe a fix is to include that config opt?

 

[ColinTaylor]

UPnP issues probably come from this miniupnpd commit, if you are behind a double NAT setup:

https://github.com/miniupnp/miniupnp/commit/8e10a1aeab9b8cd4d3b2e964b02e9ad409cf3aaa

This does seem to be an amazingly narrow minded and presumptuous change. Whilst probably OK in 99% of use cases, apart from the double-NAT scenario, it's entirely possible to run multiple local networks with port forwarding and never have a public IP address (i.e. a test lab setup). To flat-out say that port forwarding without an public IP address is impossible is simply wrong. I'd prefer it if they just put a warning in syslog and leave it enabled.

 

[RMerlin]

Don't get me wrong but maybe then it is worth to update field titles

User Name or E-mail Address
Password or DDNS Key

or they're shared across providers?

They're indeed shared accross providers. I will have to research all of them to determine which supports what, and make the label dynamic. I just haven't gotten around yet to doing so.

 

[RMerlin]

I went to do a revert, and found they changed that patch. New error test is

        } else if (addr_is_reserved(&addr)) {
            syslog(LOG_INFO, "Reserved / private IP address %s on ext interface %s: Port forwarding is impossible", if_addr, ext_if_name);
            syslog(LOG_INFO, "You are probably behind NAT, enable option ext_perform_stun=yes to detect public IP address");
            disable_port_forwarding = 1;

So maybe a fix is to include that config opt?

disable_forwarding is not controlable through config, tho it probably could.

I've opened an issue on Github, will see which way Thomas decides to go to handle that issue. Personally I'd just revert it (and any related patch) since they are more problem than they're worth. And IMHO, it's not the UPNP daemon's job to determine whether a forward works or not, its only job is to configure the forward. Anything else is the user's responsability (whether they use DMZ, CGNAT, forwarding a port range on the upstream router, etc...)

 

[john9527]

disable_forwarding is not controlable through config, tho it probably could.

I was talking about the 'ext_perform_stun=yes' suggestion.....

 

[RMerlin]

I was talking about the 'ext_perform_stun=yes' suggestion.....

Not sure how reliable that would be. Some paranoid people for instance would block access to Google's STUNs, which might falsely cause miniupnpd to fail.

I don't see the need for such a change. The port forward shouldn't require to know the actual public WAN IP AFAIK? I would just either revert back to how it worked before all those stun-related changes, or remove the check for the disable_forwarding flag where appropriate.

 

[john9527]

I would just either revert back to how it worked before all those stun-related changes

I did the revert patch (but left in the ext_perform_stun option.....having that available may be useful for custom configs)

 

[Levente]

@RMerlin
The USB drive can format from the ui, but only can do ntfs/fat/hfs, but not the native linux filesystem: ext.
From SSH I can do it, with the mkfs.ext3 /dev/sda1 command. But mkfs.ext4 not working (-sh: mkfs.ext4: not found), I had also problem with GTP (MBR was ok). Can you update somehow the mke2fs 1.42.13 (17-May-2015) package in your firmware?

Thank you!!

 

[ColinTaylor]

But mkfs.ext4 not working (-sh: mkfs.ext4: not found)

https://www.snbforums.com/threads/r...hows-strangely-in-the-logs.47562/#post-415554

 

[loveleeyoungae]

UPnP issues probably come from this miniupnpd commit, if you are behind a double NAT setup:

https://github.com/miniupnp/miniupnp/commit/8e10a1aeab9b8cd4d3b2e964b02e9ad409cf3aaa

I opened an issue with the author, but in the mean time I will probably have to revert that change.

Ping @john9527

Thanks for finding out the reason. It's a relief to know that the upnp error was not at my setup, and my careful tests were not wasted

 

[Zastoff]

Noticed Parental Controls Time Scheduling again stuck in am & pm time cant change to 24h format and somehow it`s one hour off..System time is correct
Turns off internet one hour before time

 

[eastavin]

384.7_2 rt-ac68u c1 AI protection incorrectly reports UPNP is enabled when it is not. https://www.snbforums.com/threads/384-7_2-bug-upnp-setting-ai-protection-disagree.49600/#post-441336

 

[Kronyx]

The latest GPL code for the RT-AC3200 was released on June 29th 2018, only a few months ago.

Do you think they are done with the 3200 ? I bough this router 2 years ago and it's still a good router, i hope they are not stopping to support it

 

[JIPG]

Noticed Parental Controls Time Scheduling again stuck in am & pm time cant change to 24h format and somehow it`s one hour off..System time is correct
Turns off internet one hour before time

I can select both formats (am/pm - 24h) without problem, and I have not heard any complain from my children about the allowed hours ...

Do you have NAT acceleration enabled? In the GUI it is stated : "Please disable NAT Acceleration for more precise scheduling control".

 

[Zastoff]

I can select both formats (am/pm - 24h) without problem, and I have not heard any complain from my children about the allowed hours ...

Do you have NAT acceleration enabled? In the GUI it is stated : "Please disable NAT Acceleration for more precise scheduling control".

NAT acceleration is off
And thank you for checking your settings @JIPG

edit: Weird part..when checking from computer(W7 64bit) i get am/pm time cant change and when i log in from my phone(android 9) i get 24hour format i use Brave browser both on phone and computer.
Have to check the off time later

edit: reboot seems to have fixed off time issue
am/pm time issue seems to be a Brave browser issue on Win7..

 

[RMerlin]

Do you think they are done with the 3200 ? I bough this router 2 years ago and it's still a good router, i hope they are not stopping to support it

It's not on their public EOL page, that's all I can tell.

 

[redvers]

Having been using an RT-AC66U for several years without issue, last weekend the USB drive seemed to fail... this was the push for me to finally upgrade, so I bought myself an RT-AC86U along with a new USB3 Sandisk Ultra Flair drive...

The first thing I did was to upgrade straight to 384.7.2, and start setting things up based on the configuration on my existing router. The only configuration that was actually imported was my list of manually assigned IP addresses.

After getting everything set up, there were a few niggling issues, but most of these are now resolved. I've read through the whole thread this morning, and of my three main issues left, one has already been mentioned (devices showing as MAC address rather than name) as being an Asus problem. And my issue with getting the PIA VPN client working is probably just my setup rather than anything inherent with the firmware

Which leaves us with my biggest problem... As I have come straight from stock I don't know if this is a 384.7.2 issue, or something I've screwed up with my configuration, but I've played around with settings and cannot find anything to fit it myself!

I have a WHS server sharing folders using SMB, and my laptop mounts these on boot up. However, half the time I boot they're not available, and I cannot mount them manually. It's not just an SMB issue, as I cannot connect to the web services running on the same PC, although I can connect to the AC86U. If I disable/reenable wifi, I can then mount the shares and connect to the web services. There are no issues for devices connecting to the server over ethernet (not via AC86U).

After a few minutes of being connected, I get a pop-up on my laptop telling me that the server has disconnected. If I wait long enough, the server becomes accessible again, but nothing shows up in the system logs (default log settings) so I am at a loss as to what it caused the issue. Alternatively, I can restart wifi and everything becomes accessible immediately.

I'm hugely confused!

I thought it might have been down to the dual band smart connect, so I disabled it, but still have the problem. I encounter the same issue with my phone; so it's not purely a laptop issue. Funnily enough, I can connect to the web services via my DDNS address at the same time as being unable to connect via the direct IP address.

Does anyone have any thoughts / suggestions as to how to track this down?

 

[loveleeyoungae]

I bought myself an RT-AC68U along with a new USB3 Sandisk Ultra Flair drive...

I thought it might have been down to the dual band smart connect, so I disabled it

Just curious, if its AC68U, not AC86U, how can you have that Smart Connect feature?