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

[john9527]

Did you change your router address? The default is 192.168.1.1, not 192.168.0.1. If you did make this change, I strongly recommend that you change back as some of the networkmap code depends on the address being in 192.168.1.1/24.

If the above is not in play, in general as we have learned more about strange web gui issues, it has usually been traced to a browser or addin issue.

- Disable adblockers for the router address
- Disable Password managers for the router address
- Disable off-line caching in the browser for the router address

For Firefox (a common offender) the last point can be found in
Tools-> Options -> Advance -> Network -> "The following websites are allowed to store data for offline use" dialog. Remove the entry for the router.

 

[mei]

Did you change your router address? The default is 192.168.1.1, not 192.168.0.1. If you did make this change, I strongly recommend that you change back as some of the networkmap code depends on the address being in 192.168.1.1/24.

If the above is not in play, in general as we have learned more about strange web gui issues, it has usually been traced to a browser or addin issue.

- Disable adblockers for the router address
- Disable Password managers for the router address
- Disable off-line caching in the browser for the router address

For Firefox (a common offender) the last point can be found in
Tools-> Options -> Advance -> Network -> "The following websites are allowed to store data for offline use" dialog. Remove the entry for the router.

thanks for the reply!
yes sorry I meant 192.168.1.1 sorry to confuse you
yeh its not listed in there
tho the interesting thing is google gives me the IP has changed error, firefox says cant be reached

I found a thread about it
www.smallnetbuilder.com/forums/showthread.php?p=87497

sorry to bug u guys
would be nice to know what caused it
o well

 

[mgkicker]

John, thanks for the quick update.I am away but will update and if I have any issues I'll post back.

Max

 

[threepwood]

thanks for the reply!
yes sorry I meant 192.168.1.1 sorry to confuse you
yeh its not listed in there
tho the interesting thing is google gives me the IP has changed error, firefox says cant be reached

I found a thread about it
www.smallnetbuilder.com/forums/showthread.php?p=87497

sorry to bug u guys
would be nice to know what caused it
o well

Don't know if you use Eset NOD32, but if you do, you need to add your router IP address to the Web exceptions list in NOD32. It has caused me a lot of timeouts and incomplete router webpages, either with AsusWRT or Tomato.

 

[jklepach]

No, sorry....and I actually use the 68R for the fork development and it works fine here.

Is this your first load of the fork code? If you came from a 376.xx release (Merlin's latest) you need to do a factory reset as I know some of the values for the sysinfo page have changed and there may be some remnants that are causing problems.

Also, what browser are you using? Have you tried a different one?

Different browser had the same issue. But a reset fixed this problem. I thought that I could get away without a reset since everything was working fine except for this one page. Lesson learned, do a reset when your supposed to.

 

[il2]

Hello John and participants. Thank you for your work.

I've come from Merlin's 376_47 on RT-N16 to test a few things on entware. I've flashed recent 374.43_2-06j9527 and done settings reset. Now in the process of manual settings restore I cфme across a few things:
1) IPv6 tunnel 6in4 doesn't present "Server IPv6 Address" (it was in 376_47) and needed for Tunnelbroker tunnel. I guess I need
nvram set ipv6_tun_peer="server_ipv6_address"
nvram commit
2) IPv6 "LAN Prefix Length" and "LAN IPv6 Prefix" are not restored when "connection type" is switched from "Tunnel 6in4" to "Native" or "Static". Also I guess "LAN IPv6 Prefix" is not saved to or restored from Settings file (as in 376_47). I'll test this, if you wish to check.
3) "service name" field in WAN,Virtual Server/Port Forwarding filters out underscores "_" (does not allow to enter it) - no clue when this filter come into place. I will happily live with it.

 

[Builder71]

OpenVPN

Not related to a specific FW, if I look into the System Log I see a few [undef] statements when OpenVPN is started.
For sure the incoming TCP connection port is defined.
For me, just the default port 1194.

No big deal, but maybe easy to fix?

Nov 27 21:48:41 openvpn[637]: Listening for incoming TCP connection on [undef]
Nov 27 21:48:41 openvpn[637]: TCPv4_SERVER link local (bound): [undef]
Nov 27 21:48:41 openvpn[637]: TCPv4_SERVER link remote: [undef]

 

[rodenti]

Did you change your router address? The default is 192.168.1.1, not 192.168.0.1. If you did make this change, I strongly recommend that you change back as some of the networkmap code depends on the address being in 192.168.1.1/24.

John, I have been using 192.168.2.1 from the start of getting my n66u. What issues should I be seeing by using that range instead of 1.1?




Sent from my HTC One using Tapatalk

 

[john9527]

John, I have been using 192.168.2.1 from the start of getting my n66u. What issues should I be seeing by using that range instead of 1.1?

I actually simplified things a bit....the real restriction is that networkmap is hardcoded for a /24 netmask, and if someone is changing the 3rd value in the ip address it usually means they are trying to have a larger subnet (like having the dhcp addresses on 192.168.1.x and the static addresses on 192.168.2.x with a netmask of /22).

When the netmask is other than /24, the client status will either fail to find all the clients or get stuck spinning trying to find them. If you stay with a /24 netmask, you'll be OK.

 

[Deleted member 27741]

And by /24 netmask, you mean 255.255.255.0, correct? A lot of people have no idea what /24 netmask means. Including myself.

 

[john9527]

Hello John and participants. Thank you for your work.

I've come from Merlin's 376_47 on RT-N16 to test a few things on entware. I've flashed recent 374.43_2-06j9527 and done settings reset. Now in the process of manual settings restore I cфme across a few things:
1) IPv6 tunnel 6in4 doesn't present "Server IPv6 Address" (it was in 376_47) and needed for Tunnelbroker tunnel. I guess I need
nvram set ipv6_tun_peer="server_ipv6_address"
nvram commit

Well, congratulations.....you found the first case where some of ASUS's continuing work and Merlin's 'big merges' actually makes a functional difference. I went searching for the nvram var 'ipv6_tun_peer' in the fork and it does not exist (so setting it manually won't do anything). So I tracked it down in the latest 376.48 code, found it, then went to the location in the fork. Here is what I found....

/* TODO: add & allow gateway */

I remapped the change back to the fork on my 'development' build (what I use to play with changes before release). If you would like to give it a try, send me a PM.

2) IPv6 "LAN Prefix Length" and "LAN IPv6 Prefix" are not restored when "connection type" is switched from "Tunnel 6in4" to "Native" or "Static". Also I guess "LAN IPv6 Prefix" is not saved to or restored from Settings file (as in 376_47). I'll test this, if you wish to check.

This is the behavior in the fork, if you change the connection type, it clears out 'most' of the variables.

3) "service name" field in WAN,Virtual Server/Port Forwarding filters out underscores "_" (does not allow to enter it) - no clue when this filter come into place. I will happily live with it.

Also confirmed this is the behavior in the fork.

 

[john9527]

And by /24 netmask, you mean 255.255.255.0, correct? A lot of people have no idea what /24 netmask means. Including myself.

Yes, that's correct It's just a shorthand counting the number of bits set most significant to least significant....each value in the 'number netmask' is 8 bits....so 255.255.255.0 = 8+8+8+0 = /24

 

[ika]

Hi everybody!

First of all many thanks for Merlin, john9527, AndreyPopov and for all the other contributors of the great custom content you guys made for the Asus routers. Seriously, thanks a lot, you are awesome!
I'm a long time lurker here (actually I' reading the forum almost from the beginning), I just never made an account until now.

Yes, that's correct It's just a shorthand counting the number of bits set most significant to least significant....each value in the 'number netmask' is 8 bits....so 255.255.255.0 = 8+8+8+0 = /24

I'm using a 255.255.255.224 subnet mask since I have my n66u, I had it with my original fw, with Merlin's one and with your too, but never really got into any trouble with that. (Well minidlna doesn't really work for me with your fw because it rebuffers a lot, but I'm willing to accept that since the WIFI signal is much more stronger with your fork.)

Could you please tell me what kind of problems should I expect by not using 255.255.255.0 ?

 

[il2]

Well, congratulations.....you found the first case where some of ASUS's continuing work and Merlin's 'big merges' actually makes a functional difference. I went searching for the nvram var 'ipv6_tun_peer' in the fork and it does not exist (so setting it manually won't do anything). So I tracked it down in the latest 376.48 code, found it, then went to the location in the fork. Here is what I found....

/* TODO: add & allow gateway */

I remapped the change back to the fork on my 'development' build (what I use to play with changes before release). If you would like to give it a try, send me a PM.

Looks like work on this is unfinished due this TODO.

Actually setting ipv6_tun_peer to my tunnel server address does help: ping6 from the router and behind it returns success and I get 10/10 on test-ipv6.com
Changing ipv6_tun_peer to "" without reboot doesn't broke the tunnel, perhaps, because route -A inet6 still shows the route to the server.
After nvram commit and reboot IPv6 is generally broken.
To conclude, without ipv6_tun_peer set correctly in Tunnel6in4 mode IPv6 is broken.

Thank you, I'd like to try your dev build on RT-N16 and PM you soon.

PS. There is one more thing. Router's IPv6 DHCP server provides to the client DNS server address configured by tunnel even if it is not set in the IPv6 DNS Setting. I would expect client configured with DNS server=router's LAN IPv6 Address in this case. Thus the client behind the router is bypassing dnscrypt-proxy/opendns setup on the router while doing DNS requests. The test page returns unsuccessful report. It is unclear what's wrong in my setup: IPv6 tunnel's DNS server preference or dnscrypt-proxy bypass.
I guess I'll dig dnscrypt-proxy topic for this.

 

[polu]

Hello, on the Merlin fork 374.43_2-06j9527 version there is the following issue:

Wireless > General > 5GHz > Wireless Mode - The AC option is absent (please refer to the screenshot). Is it possible to fix it?

Thanks.

 

[john9527]

Hello, on the Merlin fork 374.43_2-06j9527 version there is the following issue:

Wireless > General > 5GHz > Wireless Mode - The AC option is absent (please refer to the screenshot). Is it possible to fix it?

Thanks.

Please read the release notes....this was done on purpose.

- With this release the Wireless Mode options are standardized across all supported routers with 3 choices....Auto, N Only and Legacy. If you are using AC mode clients, you should select 'Auto'.
In my testing the N+AC mode did not always have consistent results. These options should provide the most stable connections across all platforms.

 

[polu]

Please read the release notes....this was done on purpose.

- With this release the Wireless Mode options are standardized across all supported routers with 3 choices....Auto, N Only and Legacy. If you are using AC mode clients, you should select 'Auto'.
In my testing the N+AC mode did not always have consistent results. These options should provide the most stable connections across all platforms.

Thanks a lot, got it, sorry for not reading the release notes carefully.

 

[john9527]

Not a problem. When I re-read my original post, I realized it could come across as abrupt....this was not my intent. I'll be more careful when I post early in the morning...

 

[john9527]

Hi everybody!

First of all many thanks for Merlin, john9527, AndreyPopov and for all the other contributors of the great custom content you guys made for the Asus routers. Seriously, thanks a lot, you are awesome!
I'm a long time lurker here (actually I' reading the forum almost from the beginning), I just never made an account until now.



I'm using a 255.255.255.224 subnet mask since I have my n66u, I had it with my original fw, with Merlin's one and with your too, but never really got into any trouble with that. (Well minidlna doesn't really work for me with your fw because it rebuffers a lot, but I'm willing to accept that since the WIFI signal is much more stronger with your fork.)

Could you please tell me what kind of problems should I expect by not using 255.255.255.0 ?

Actually, first time I've read of someone tightening down the subnet It's been a while since I looked at that code, but my sense is you should be OK. The only think I can think of is that if you have somehow forced some static ip on your network outside of the 192.168.1.0 - 192.168.1.31 range, the networkmap may try and find it....then?

But to your other comment on the minidlna....one of the things I test before release on my AC68 is both wired and wireless streaming to a Roku 3 and haven't seen a problem. Is there anything unique about your setup that you can think of?

 

[john9527]

Not related to a specific FW, if I look into the System Log I see a few [undef] statements when OpenVPN is started.
For sure the incoming TCP connection port is defined.
For me, just the default port 1194.

No big deal, but maybe easy to fix?

Nov 27 21:48:41 openvpn[637]: Listening for incoming TCP connection on [undef]
Nov 27 21:48:41 openvpn[637]: TCPv4_SERVER link local (bound): [undef]
Nov 27 21:48:41 openvpn[637]: TCPv4_SERVER link remote: [undef]

Definitely not easy

I did some searching and in just about every log I looked at had the [undef] output in it. Best I can tell is that this is a common messaging code where the structures aren't completed until a client actually connects.