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

[cheeseandbeans]

Has anyone managed to get BT sport working with a BT Youview box? I've been searching the net for hours but can't find the right setting to get it to work. I have an Ac66u. It was working on the last release of Merlin but not on the latest release of this. Tearing my hair out with it.

Sent from my WAS-LX1A using Tapatalk

 

[john9527]

Has anyone managed to get BT sport working with a BT Youview box? I've been searching the net for hours but can't find the right setting to get it to work. I have an Ac66u. It was working on the last release of Merlin but not on the latest release of this. Tearing my hair out with it.

Sent from my WAS-LX1A using Tapatalk

Don't know if it will help, but a quick search turned up this post.
https://www.snbforums.com/threads/ac87u-problems-with-bt-television.22319/#post-224037

 

[ciobster]

RT-N66U running latest 33E7 crashed with memory leak, complete logs attached:

Aug 6 21:08:19 dropbear[13344]: Exit before auth: Exited normally
Aug 6 21:56:42 kernel: OOMkill: task 388 (usbled) got 14 points
Aug 6 21:56:42 kernel: init invoked oom-killer: gfp_mask=0xd0, order=0, oomkilladj=0
Aug 6 21:56:42 kernel: Call Trace:
Aug 6 21:56:42 kernel: [<8000f41c>] dump_stack+0x8/0x34
Aug 6 21:56:42 kernel: [<80057508>] out_of_memory+0x1fc/0x264
Aug 6 21:56:42 kernel: [<8005937c>] __alloc_pages+0x318/0x360
Aug 6 21:56:42 kernel: [<800593ec>] __get_free_pages+0x28/0x4c
Aug 6 21:56:42 kernel: [<800c073c>] proc_info_read+0x58/0x174
Aug 6 21:56:42 kernel: [<8007d864>] vfs_read+0xbc/0x164
Aug 6 21:56:42 kernel: [<8007de2c>] sys_read+0x50/0xbc
Aug 6 21:56:42 kernel: [<800111b0>] stack_done+0x20/0x44
Aug 6 21:56:42 kernel: Mem-info:
Aug 6 21:56:42 kernel: Normal per-cpu:
Aug 6 21:56:42 kernel: CPU 0: Hot: hi: 186, btch: 31 usd: 61 Cold: hi: 62, btch: 15 usd: 59
Aug 6 21:56:42 kernel: HighMem per-cpu:
Aug 6 21:56:42 kernel: CPU 0: Hot: hi: 186, btch: 31 usd: 165 Cold: hi: 62, btch: 15 usd: 8
Aug 6 21:56:42 kernel: Active:2159 inactive:1928 dirty:0 writeback:0 unstable:0
Aug 6 21:56:42 kernel: free:27690 slab:25929 mapped:905 pagetables:114 bounce:0
Aug 6 21:56:42 kernel: Normal free:2816kB min:2884kB low:3604kB high:4324kB active:12kB inactive:0kB present:520192kB pages_scanned:60 all_unreclaimable? yes
Aug 6 21:56:42 kernel: lowmem_reserve[]: 0 14224
Aug 6 21:56:42 kernel: HighMem free:107944kB min:512kB low:3032kB high:5556kB active:8624kB inactive:7712kB present:1820672kB pages_scanned:0 all_unreclaimable? no
Aug 6 21:56:42 kernel: lowmem_reserve[]: 0 0
Aug 6 21:56:42 kernel: Normal: 0*4kB 42*8kB 5*16kB 3*32kB 0*64kB 0*128kB 1*256kB 0*512kB 0*1024kB 1*2048kB 0*4096kB = 2816kB
Aug 6 21:56:42 kernel: HighMem: 444*4kB 241*8kB 169*16kB 105*32kB 44*64kB 15*128kB 3*256kB 1*512kB 2*1024kB 0*2048kB 22*4096kB = 107944kB
Aug 6 21:56:42 kernel: Swap cache: add 0, delete 0, find 0/0, race 0+0
Aug 6 21:56:42 kernel: Free swap = 0kB
Aug 6 21:56:42 kernel: Total swap = 0kB
Aug 6 21:56:42 kernel: Free swap: 0kB
Aug 6 21:56:42 kernel: 589823 pages of RAM
Aug 6 21:56:42 kernel: 458751 pages of HIGHMEM
Aug 6 21:56:42 kernel: 529905 reserved pages
Aug 6 21:56:42 kernel: 3034 pages shared
Aug 6 21:56:42 kernel: 0 pages swap cached


I am also seeing this behaviour a lot in the logs, i guess someone is abusing the webserver?

Aug 5 21:59:27 watchdog: restart httpd - SSL, error detected or process not responding (28)
Aug 5 21:59:27 rc_service: watchdog 360:notify_rc stop_httpd
Aug 5 21:59:27 rc_service: watchdog 360:notify_rc start_httpd
Aug 5 21:59:27 httpd: start httpd
Aug 5 21:59:27 httpd: start httpd - SSL
Aug 5 22:02:13 watchdog: restart httpd - SSL, error detected or process not responding (28)
Aug 5 22:02:13 rc_service: watchdog 360:notify_rc stop_httpd
Aug 5 22:02:13 rc_service: watchdog 360:notify_rc start_httpd
Aug 5 22:02:13 httpd: start httpd
Aug 5 22:02:13 httpd: start httpd - SSL
Aug 5 22:04:24 watchdog: restart httpd - SSL, error detected or process not responding (28)
Aug 5 22:04:24 rc_service: watchdog 360:notify_rc stop_httpd
Aug 5 22:04:24 rc_service: watchdog 360:notify_rc start_httpd
Aug 5 22:04:24 rc_service: waiting "stop_httpd" via watchdog ...
Aug 5 22:04:25 httpd: start httpd
Aug 5 22:04:25 httpd: start httpd - SSL
Aug 5 22:05:09 watchdog: restart httpd, error detected or process not responding (28)
Aug 5 22:05:09 rc_service: watchdog 360:notify_rc stop_httpd
Aug 5 22:05:09 rc_service: watchdog 360:notify_rc start_httpd
Aug 5 22:05:09 rc_service: waiting "stop_httpd" via watchdog ...
Aug 5 22:05:10 httpd: start httpd
Aug 5 22:05:10 httpd: start httpd - SSL
Aug 5 22:05:10 watchdog: restart httpd - SSL, error detected or process not responding (7)
Aug 5 22:05:10 rc_service: watchdog 360:notify_rc stop_httpd
Aug 5 22:05:10 rc_service: watchdog 360:notify_rc start_httpd
Aug 5 22:05:10 rc_service: waiting "stop_httpd" via watchdog ...
Aug 5 22:05:12 httpd: start httpd
Aug 5 22:05:12 httpd: start httpd - SSL

 

[john9527]

RT-N66U running latest 33E7 crashed with memory leak, complete logs attached:

Think this is the first time I've seen an OOM on the fork. I'm definitely not versed in reading the dump, but from your log attachment it may be pointing at dnsmasq. Are you running any scripts that interface with dnsmasq or trying to load a huge hosts file for example?

I am also seeing this behavior a lot in the logs, i guess someone is abusing the webserver?

This is usually due to one of two things....
- you are trying to connect from two clients simultaneously over HTTPS (this can sometimes cause a deadlock in httpd and this detects it and restarts httpd). It could be someone trying to hack your system. Do you have HTTPS WAN access enabled?
- your HTTPS certificate is corrupt or not installed correctly

 

[ciobster]

Think this is the first time I've seen an OOM on the fork. I'm definitely not versed in reading the dump, but from your log attachment it may be pointing at dnsmasq. Are you running any scripts that interface with dnsmasq or trying to load a huge hosts file for example?

This is usually due to one of two things....
- you are trying to connect from two clients simultaneously over HTTPS (this can sometimes cause a deadlock in httpd and this detects it and restarts httpd). It could be someone trying to hack your system. Do you have HTTPS WAN access enabled?
- your HTTPS certificate is corrupt or not installed correctly

It's just an easy setup, WAN is PPPoE and i have 3 outdoor APs and 1 surveillance camera wired to the 4 Port Switch with DHCP reservations, there are no custom scripts. At the time the router crashed, there was not much going on, at best there were 5-6 wireless clients surfing the internet, nothing that could have overloaded the router. Even the port scan neighbors left:

Aug 6 18:19:21 dropbear[12577]: Exit before auth: Disconnect received
Aug 6 18:57:05 dropbear[12745]: Child connection from 112.85.42.102:37061
Aug 6 18:57:18 dropbear[12745]: Exit before auth: Disconnect received
Aug 6 19:02:03 ntp: start NTP update
Aug 6 19:02:33 ntp: NTP update successful after 2 attempt(s)
Aug 6 20:02:28 ntp: start NTP update
Aug 6 20:02:33 ntp: NTP update successful after 2 attempt(s)
Aug 6 21:02:33 ntp: start NTP update
Aug 6 21:03:03 ntp: NTP update successful after 2 attempt(s)
Aug 6 21:08:19 dropbear[13344]: Child connection from 123.20.185.199:7451
Aug 6 21:08:19 dropbear[13344]: Exit before auth: Exited normally
Aug 6 21:56:42 kernel: OOMkill: task 388 (usbled) got 14 points

Yes, i have enabled SSH and Web access from WAN side for remote management purposes. Usually i do it from SSH but while on route i tend to connect to the GUI from my phone, feels easier. I will consider disabling the Web access in case i see it abused a lot.

Does one know an easy method to block the port scan abusers wihtout putting serious stress on the router? I mean it's an old device and i really like to keep the traffic flowing without bottlenecking it with intensive firewall rules...

 

[john9527]

Yes, i have enabled SSH and Web access from WAN side for remote management purposes. Usually i do it from SSH but while on route i tend to connect to the GUI from my phone, feels easier. I will consider disabling the Web access in case i see it abused a lot.

Does one know an easy method to block the port scan abusers wihtout putting serious stress on the router? I mean it's an old device and i really like to keep the traffic flowing without bottlenecking it with intensive firewall rules...

Much better to set up an OpenVPN server on the router and access it that way (on a non-standard OVPN port even better).

 

[erkr]

Hi John
Updated two rtn66u devices in use as AP a few days ago to 33e7
First impression is it runs rock solid. Didn't observe any memory leak (free memory is stable), no strange log entries either. Thanks for maintaining this for fork!!!!
Eric

 

[Builder71]

Hi John
Updated two rtn66u devices in use as AP a few days ago to 33e7
First impression is it runs rock solid. Didn't observe any memory leak (free memory is stable), no strange log entries either. Thanks for maintaining this for fork!!!!
Eric

Same here, after 7 days of uptime on 33E7.
I use the RT-N66U in: "Wireless router mode (Default)"
No problems so far. Thx!

 

[Brenneke]

What is the difference between 33L7 (Jul-02) and 33E7 (Jul-01) for N16? Which should I use for this router?

 

[Grisu]

short: E (enhanced) builds are "regular" builds, L (legacy) have old more powerful wireless driver but not all updates (missing CRACK) and not applicable for all (newer) router releases.

'E' Builds are the default releases and the AC68U firmware supports all revs and variations of the AC68 with a new SDK
'L' Builds are based on the older AC68U SDK of the original fork and the AC68U firmware only support revs A1,A1,B1

https://onedrive.live.com/?authkey=!AAkJBWoo-OiyKJo&cid=09332FC159A7E129&id=9332FC159A7E129!2788&parId=9332FC159A7E129!2779&o=OneUp

 

[Brenneke]

short: E (enhanced) builds are "regular" builds, L (legacy) have old more powerful wireless driver but not all updates (missing CRACK) and not applicable for all (newer) router releases.

'E' Builds are the default releases and the AC68U firmware supports all revs and variations of the AC68 with a new SDK
'L' Builds are based on the older AC68U SDK of the original fork and the AC68U firmware only support revs A1,A1,B1

https://onedrive.live.com/?authkey=!AAkJBWoo-OiyKJo&cid=09332FC159A7E129&id=9332FC159A7E129!2788&parId=9332FC159A7E129!2779&o=OneUp

Which of the two would you go with for an N16 set up as a repeater?
Thank you.

 

[john9527]

What is the difference between 33L7 (Jul-02) and 33E7 (Jul-01) for N16? Which should I use for this router?

For the N16 there is no change in the wireless between the two builds since Asus never released the KRACK update for the N16 wireless drivers. There may be a couple of shared MIPS components like CTF that may be updated in the 'E' builds. I'd use the same recommendation as for the other. models....start with the 'E build.....if you have any problems then go to the 'L'. (You should be able to go between them without a factory reset required)

 

[spicypixel]

Hey guys, I've tried the latest ASUS release, latest merlin and now latest release of this LTS fork, and yet it's seemingly a very experience on all.

Is the media bridge function broken? After a certain amount of time the router is no longer contactable via IP but you still have WAN access from client devices, later you lose IP - I assume as the DHCP allocation expires and you're left with a 169.254.x.x IP.

Is this to be expected? I've been unable to find any firmware where the media bridge function works correctly on my AC66U and it's quite frustrating as when it is working it's very fast and much better at keeping a stable connection than the client machines trying to connect via their own wifi.

 

[joe a]

Will this work on the new Asus RT-n66u_c1?

 

[ColinTaylor]

Will this work on the new Asus RT-n66u_c1?

I don't know, but the wikidevi entry for it says the hardware is exactly the same as the RT-AC66U_B1 (apart from the lack of AC support). So it's completely different to the original RT-N66U.

Do you have one? What is the FCC ID on it? What's to output of cat /proc/cpuinfo

 

[john9527]

Will this work on the new Asus RT-n66u_c1?

I'm also guessing it may be a single core version of the processor....the /proc/cpuinfo output will tell.

 

[ColinTaylor]

I'm also guessing it may be a single core version of the processor....the /proc/cpuinfo output will tell.

Déjà vu https://www.snbforums.com/threads/rt-n66u-c1.46672/

 

[ColinTaylor]

QoS (again!)

After the discussion we had here I was looking again at the QoS effects on DNS. I noticed something that hadn't occurred to me before, or if it had I'd long since forgotten.

Traffic that leaves to router and goes to the internet is classified as expected. So that's things like DNS or HTTP requests etc.

Traffic going to the LAN, so that's things like replies to said requests, is also classified as expected.

Traffic from the internet that is destined for the router, so that's things like DNS replies to dnsmasq, are not categorised at all.

That make sense because the interfaces monitored are eth0 (in my case) for egress and br0 for ingress.

Thoughts?

EDIT: Thinking about this some more (and making my head hurt), this is a bit of a red herring. The incoming traffic is marked correctly because of the "CONNMARK --restore-mark". But that's actually irrelevant because the traffic has already arrived at its destination (the router) in the case of dnsmasq. QoS can only effect traffic as it is sent to an interface, either the WAN interface (eth0) or the LAN interface (bro).

So, "as your were". And let's never speak of this again.

 

[maurer]

I had some issue in one of my networks meaning the Repeater (tp-link tl-wr842ndv1 running latest openwrt 18.06) was losing its connection every minute or so.
After countless hours of debug (including replacing the repeater) I've concluded that the issue was the main gw/AP RT-N66u running latest 33E7.
The quick fix was to revert to 33L7 (via mtd-write) which fixed my wifi - long story short if you want better range and stability (and using AP only - as repeater is vulnerable to KRACK) just use the L version.
A big thanks to John for still supporting the Legacy drivers ! (although I voted some time ago to quit maintenance of these drivers - luckily others voted to keep it )

 

[macsmister]

Sorry if this has been answered before... but what is the benefit of using this build (Asuswrt-Merlin 374.43 LTS releases V33E7) over the official Asus build (Version 3.0.0.4.382.50702)? I have a RT-N66-U btw, which now is on Merlin-WRT 380.70. Since Merlin-WRT is no longer supported for my router, I need to switch to either this fork, or go back to the official Asus firmware.