Asuswrt-Merlin 378.56 is available

[john9527]

-rw-rw-rw-    1 admin    root          6397 Jul 31 19:00 upnpc_xml.log
-rw-rw-rw-    1 admin    root         24244 Oct 29 14:48 syslog.log
-rw-rw-rw-    1 admin    root      48655985 Oct 29 14:51 mDNSNetMonitor.log

Thanks. I appreciate it.

A 48M log file....I think that qualifies as 'obvious'

 

[RMerlin]

A 48M log file....I think that qualifies as 'obvious'

Syslog flood, /tmp logfile filling... Sounds like whoever was working on mDNSNetMonitor forgot to remove a lot of debug stuff before the 9135/9177 release.

 

[RMerlin]

This is the Asus model to a T, "Thanks for the money"
They used to be a great company & no longer give a toss about things like releasing a UK model with the allowed channels.

But hey, that's what a hex editor and a bootloaders for.

Asus actually added a a territory code selector a few months ago.

Bootloader of the newest Broadcom devices is encrypted, so forget about hex editing in the future.

 

[tracker.ca]

I am experiencing the same issue described here:

http://www.snbforums.com/threads/ac68-merlin-378-56-beta2-openvpn-cant-start.27774/#post-212783

That was during beta, but I am having it with the final .56 version. AC56U

OpenVPN Server cannot start (TUN, Static Key, Routed Point2Point VPN)

Everything seems to be OK while using 54_2

Excerpts from the Syslog:

54_2 (Working)

Oct 29 12:52:10 openvpn[673]: /usr/sbin/ip link set dev tun22 up mtu 1500
Oct 29 12:52:10 kernel: ADDRCONF(NETDEV_CHANGE): tun22: link becomes ready
Oct 29 12:52:10 openvpn[673]: /usr/sbin/ip addr add dev tun22 10.0.1.1/4 broadcast 255.255.255.253
Oct 29 12:52:10 openvpn[673]: /usr/sbin/ip route add 10.254.3.0/24 via 10.0.1.1


56 (Not Working)

Oct 29 12:43:22 openvpn[662]: /sbin/ifconfig tun22 10.0.1.1 netmask 10.0.1.2 mtu 1500 broadcast 255.255.255.253
Oct 29 12:43:22 kernel: ADDRCONF(NETDEV_CHANGE): tun22: link becomes ready
Oct 29 12:43:22 openvpn[662]: Linux ifconfig failed: external program exited with error status: 1
Oct 29 12:43:22 openvpn[662]: Exiting due to fatal error

My uneducated guess is that something changed in the parsing of the parameters and IFCONFIG is missing the NETMASK

(edited for clarification)

 

[user1999]

My RT-AC68U's NetworkMap page isn't showing properly after upgrading to 56. There's no left navigation and the top part of the page is missing. It seems the page is cropped somehow. Also there's no scrollbar and most of the page is blank. If I navigate to the other pages by typing the url directly they work fine. The problem only happens to the Network Map page, which is the front page. This is quite annoying.

Anyone got the same problem? Is there a fix?

 

[CiscoX]

My RT-AC68U's NetworkMap page isn't showing properly after upgrading to 56. There's no left navigation and the top part of the page is missing. It seems the page is cropped somehow. Also there's no scrollbar and most of the page is blank. If I navigate to the other pages by typing the url directly they work fine. The problem only happens to the Network Map page, which is the front page. This is quite annoying.

View attachment 4711

Anyone got the same problem? Is there a fix?

Have you tried another browser?
And did you do a factory reset after upgrading the firmware?

 

[RMerlin]

My RT-AC68U's NetworkMap page isn't showing properly after upgrading to 56. There's no left navigation and the top part of the page is missing. It seems the page is cropped somehow. Also there's no scrollbar and most of the page is blank. If I navigate to the other pages by typing the url directly they work fine. The problem only happens to the Network Map page, which is the front page. This is quite annoying.

Anyone got the same problem? Is there a fix?

Open your web browser error console and look for a Javascript error. My guess is you have a network client with an invalid name that's breaking the Javascript code - please post the content of the line with the error.

 

[CiscoX]

For me it is working fine on my RT-AC87U,
View attachment 4709

It is working now. I had to change the control channel from Auto to an static number. For the 5Ghz band.

 

[CiscoX]

Open your web browser console and see if there's any Javascript error. Repost the content of the line with the error (if any).

Hi RMerlin,

It seems to work now. I was just playing around with the settings for 5Ghz band, and changed the Control Channel from AUTO to an static number.
After that the Site Survey is working. I don't know why.

This is the only javascript error i get:
21:57:57.473 unreachable code after return statement1 help.js:43:0

 

[Haggster]

Flashed the new firmware now and lovin it so far.

But I'm posting because of the following: The Adaptive QoS seems to have switched itself off after modifying one of the bandwidth limiting values. I was performing a speedtest, adjusted a value, applied the adjustments, performed another speedtest - and suddenly the values went way beyond what I had defined. So I shut QoS off again, applied, on again - same result. Adjusting the values again, applying - same result.

I definitely know that adjusting it on the fly was possible with the .55 firmware without issues. Now, I know that the DPI engine is closed source and that Merlin cannot modify it, but maybe the issue is in some connection to any of the open source components?

The log shows this:

Oct 29 20:27:43 rc_service: httpd 439:notify_rc restart_qos;restart_firewall
Oct 29 20:27:45 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:27:45 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:27:45 haggster: sh /opt/S50downloadmaster.1 firewall-start
Oct 29 20:27:46 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:27:46 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:27:47 haggster: sh /opt/S50downloadmaster.1 firewall-start
Oct 29 20:29:01 rc_service: httpd 439:notify_rc restart_qos;restart_firewall
Oct 29 20:29:02 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:29:02 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:29:03 haggster: sh /opt/S50downloadmaster.1 firewall-start
Oct 29 20:29:03 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:29:04 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:29:04 haggster: sh /opt/S50downloadmaster.1 firewall-start
Oct 29 20:30:35 rc_service: httpd 439:notify_rc restart_qos;restart_firewall
Oct 29 20:30:36 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:30:37 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:30:37 haggster: sh /opt/S50downloadmaster.1 firewall-start
Oct 29 20:30:37 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:30:38 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:30:38 haggster: sh /opt/S50downloadmaster.1 firewall-start
Oct 29 20:30:48 rc_service: httpd 439:notify_rc restart_qos;restart_firewall
Oct 29 20:30:49 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:30:50 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:30:50 haggster: sh /opt/S50downloadmaster.1 firewall-start
Oct 29 20:30:51 start_nat_rules: apply the nat_rules(/tmp/nat_rules_ppp0_eth0)!
Oct 29 20:30:51 haggster: sh /opt/S50asuslighttpd.1 firewall-start
Oct 29 20:30:51 haggster: sh /opt/S50downloadmaster.1 firewall-start

I've tried searching the thread for similar reports, but since "QoS" is a too short search term to be accepted, I was then searching for "Adaptive QoS" instead, so sorry in case this has been mentioned already...since it's not listed as a known issue and I didn't find anything regarding this I thought I'd mention it.

Edit: I've performed a reboot now and now it's clear: The upstream QoS limit isn't working. The downstream one limiting exactly like I have set it, but the upstream value is being ignored. Any ideas on how to fix this (if possible at all)?

I'm using an RT-AC87 by the way.


EDIT2: If I select e.g. "Media Streaming" as the Adaptive QoS profile it works; I was using the "Customize" option before (which was basically the only option in the last firmware). Does anyone else have this problem, that the upstream QoS limit is not working when using Customized under Adaptive QoS?

 

[RMerlin]

Oct 29 12:52:10 openvpn[673]: /usr/sbin/ip addr add dev tun22 10.0.1.1/4 broadcast 255.255.255.253

That is an incredibly large netmask you have set there - one that isn't valid. The 10.0.0.0 subnet shouldn't be larger than a /8.

 

[tracker.ca]

That is an incredibly large netmask you have set there - one that isn't valid. The 10.0.0.0 subnet shouldn't be larger than a /8.

Hi RMerlin, thanks for looking into it.
That Subnet is actually on the firmware that works (54_2) and it is somehow calculated automatically (no place on the GUI to specify it).

 

[RogerSC]

I might tie it to an nvram setting, so advanced users will be free to manually enable it. But I won't be exposing it to the webui, as then it would bring more questions.

Tying to an nvram control setting is easy, since the buffer flush is just a single line of code. Similar to how I've left SMB 2.0 support controllable through nvram.

Someone might want to start a wiki page someday with those undocumented settings...

What good are undocumented settings if they aren't documented somewhere *smile*?

 

[fr33z0n3r]

HTTPS on the router runs on 8443 port, not on HTTPS (443) port, thats due to 443 port being reserved to Cloud service.

8443

No, I changed that to port 443, and just verified that AiCloud is still on a different high port number. And disabled.

Any other ideas?


UPDATE - Just confirmed it worked in 55. The code on that page at line 397 appears to be incorrect - unless this is an expected change.

if (HTTPS_support && (document.form.http_enable[0].selected != true) && !validator.range(document.form.https_lanport, 1024, 65535) && !tmo_support)
return false;

 

[CiscoX]

Hi RMerlin,

It seems to work now. I was just playing around with the settings for 5Ghz band, and changed the Control Channel from AUTO to an static number.
After that the Site Survey is working. I don't know why.

This is the only javascript error i get:
21:57:57.473 unreachable code after return statement1 help.js:43:0

@RMerlin

I found out. If you don't have any clients connected to the 5Ghz band. Then the "Site Survey" stay stock on the "Searching for available networks, please wait... "
And when i connect my phone to the 5Ghz and run the "Site Survey" then is shows "Visible Networks". Stange?

 

[user1999]

Open your web browser error console and look for a Javascript error. My guess is you have a network client with an invalid name that's breaking the Javascript code - please post the content of the line with the error.

Yes you're right. Here are the errors:

 

[Ronv42]

Hey Merlin, I did the update to 378.56 release on my AC87R from 387.55 and I noticed that there is a new "global" icon in the network map that shows if a host has access to the internet or not. It did show that my printers and cameras have no access to the internet as expected because I put them into parental controls / time scheduling not to have access. Then I noticed for the information you can edit now from the node in network map that you can block access there, change the status of time schedule, and set a static name in DHCP.

So I tried a little change and removed one device from scheduled parental controls and went back to it in network map and changed the device there to block internet access. After saving the change the named DHCP list wiped the name for that device from the list but it was still blocked from the internet as expect It did keep the mac and IP addressed binded though. But the network map for the device still shows my assigned name.

I know you have had issues with this network map editing before. Is this a bug? I think I should just stay with my original way of restricting a device intenet access via parental controls verses the network maps. What is your thoughts? I am attaching a screen shot of what I changed in the network map for the device.

 

[RMerlin]

No, I changed that to port 443, and just verified that AiCloud is still on a different high port number. And disabled.

Any other ideas?


UPDATE - Just confirmed it worked in 55. The code on that page at line 397 appears to be incorrect - unless this is an expected change.

if (HTTPS_support && (document.form.http_enable[0].selected != true) && !validator.range(document.form.https_lanport, 1024, 65535) && !tmo_support)
return false;

Asus chose to limit ports to unprivileged ones for some reason. I could look into removing that restriction if it doesn't create any other problem.

The reason it "worked" with 378.55 was because of a bug in previous versions that wasn't properly checking the port.

 

[RMerlin]

Yes you're right. Here are the errors:
View attachment 4712

Click on client_function.js and repost the content of that line.

 

[RMerlin]

Hey Merlin, I did the update to 378.56 release on my AC87R from 387.55 and I noticed that there is a new "global" icon in the network map that shows if a host has access to the internet or not. It did show that my printers and cameras have no access to the internet as expected because I put them into parental controls / time scheduling not to have access. Then I noticed for the information you can edit now from the node in network map that you can block access there, change the status of time schedule, and set a static name in DHCP.

So I tried a little change and removed one device from scheduled parental controls and went back to it in network map and changed the device there to block internet access. After saving the change the named DHCP list wiped the name for that device from the list but it was still blocked from the internet as expect It did keep the mac and IP addressed binded though. But the network map for the device still shows my assigned name.

I know you have had issues with this network map editing before. Is this a bug? I think I should just stay with my original way of restricting a device intenet access via parental controls verses the network maps. What is your thoughts? I am attaching a screen shot of what I changed in the network map for the device.

Asus is making changes to that code with every new firmware release, so I'm waiting for it to stabilize before I spend any time looking at it.