Stubby-Installer-Asuswrt-Merlin

[Butterfly Bones]

I have the same yet I see this in my syslogs at the end of the reboot after Skynet is done starting:

Feb  9 12:02:10 (install_stubby.sh): 438 Ending Script Execution
Feb  9 12:02:10 (install_stubby.sh): 2541 Ending Script Execution

And the port forwarding log shows the instruction shows two more entries, so 4 in total. Looks like this:

Source    Proto    Port range    Redirect to    Local Port    Chain
ALL    UDP    53    192.168.14.1    53    PREROUTING
ALL    TCP    53    192.168.14.1    53    PREROUTING
ALL    UDP    53    192.168.14.1    53    PREROUTING
ALL    TCP    53    192.168.14.1    53    PREROUTING

There has to be something calling Stubby.sh twice somewhere in your scripts. I thing you need to comb through the all the jffs scripts to find it.

If you have a remote FTP client, try what I do - copy the jffs scripts directory to my computer and then do a file search for a string like "stubby" in all files in that directory. I use FileZilla on Linux to do that, not sure what exists for Windows... There is a "find" in my AC86U might be from Entware findutils, but the help output is way over my head, since regex is as foreign to me as scripting, I'm just a dolt with both.

 

[Butterfly Bones]

FWIW I think the "D" part of the redirect is not working or I would have only one entry.

Check the new install_stubby.sh, both the two -D and two -A iptables entries are at the very bottom in mine, and I see no dupes.

 

[skeal]

Check the new install_stubby.sh, both the two -D and two -A iptables entries are at the very bottom in mine, and I see no dupes.

Same here. I went through every script in /jffs cannot find a duplicate call for install_stubby.sh

 

[cmkelley]

You know there is a syslog-ng thread here, right?
https://www.snbforums.com/threads/configuring-syslog-ng-with-merlin-firmware.35095/
I tired for hours and days to get in running on my AC86U and finally found that the symlink was stopping copying the syslog output back into the GUI. That seems to be recently solved, thanks to elorimer and Adamm, changing Skynet script to work with the destructive sed command. Also I not there is a suggested syslog-ng method in your thread to deal with the dcd crashes. I'm going to revisit that, since I still have the configs I was working with a few months ago. And yes, wrapping ones head around syslog-ng IS a real mind f***! Ok, /offtopic

Yep, aware of the thread. I used a lot of that to get to where I am with syslog-ng. Also aware of the suggestion, might try it, but I don't see how it could work - the dcd message is a multi-line message.

 

[dave14305]

@Adamm I deleted the script called Redirect_DNS.sh and it's call from firewall-start or nat-start. The only thing changing forwarding chains are the alterations you made to the nat-start script. I have now rebooted 3 times same result.

You don't really need this nat-start script since you can create the same results using DNSFilter natively to Merlin. This latest feature is reinventing the wheel, in my opinion. But who cares what I think?

 

[skeal]

You don't really need this nat-start script since you can create the same results using DNSFilter natively to Merlin. This latest feature is reinventing the wheel, in my opinion. But who cares what I think?

I trust @Adamm will have this fixed in no time. @Adamm and @thelonelycoder are awesome at what they do!!!

 

[elorimer]

I'm fairly certain multi-line-prefix and multi-line-suffix are they key, but damned if I can find an understandable explanation.

I'm going to post over in the syslog-ng thread about this --that won't work.

 

[cmkelley]

I'm going to post over in the syslog-ng thread about this --that won't work.

I just posted a howto for syslog-ng there. It doesn't cover the dcd crash, but at least it gets syslog-ng working (I think - it needs testers).

 

[Joe Doe]

Not in Chrome yet according to these posts on GitHub.
https://github.com/chromium/badssl.com/issues/356
Chromium.org / bugs
https://bugs.chromium.org/p/chromium/issues/detail?id=908132

It also didn't work with Firefox 58 (Windows). Does it work for anybody else? If so, what browser are you using?

 

[Butterfly Bones]

It also didn't work with Firefox 58 (Windows). Does it work for anybody else? If so, what browser are you using?

Did you search this thread, there are quite a few posts about Encrypted SNI.
https://www.snbforums.com/threads/stubby-installer-asuswrt-merlin.49469/page-44#post-464709

 

[Joe Doe]

Did you search this thread, there are quite a few posts about Encrypted SNI.
https://www.snbforums.com/threads/stubby-installer-asuswrt-merlin.49469/page-44#post-464709

Thanks, seems to be currently limited to Firefox Nightly with specific settings. Thanks

 

[cmkelley]

Thanks, seems to be currently limited to Firefox Nightly with specific settings. Thanks

You don't need the nightly, but you do need to configure the specific settings, it doesn't do ESNI out of the box. I have Firefox 65 (Windows 64 bit) with the settings described and it works for me.

 

[skeal]

Ok the router restarted firewall-start. This ended up calling nat-start. the syslog shows this which is encouraging.

Feb  9 15:00:35 nat: apply nat rules (/tmp/nat_rules_eth0_eth0)
Feb  9 15:00:35 custom_script: Running /jffs/scripts/nat-start
Feb  9 15:00:35 (install_stubby.sh): 29303 Starting Script Execution
Feb  9 15:00:36 custom_script: Running /jffs/scripts/firewall-start (args: eth0)
Feb  9 15:00:36 adaptive QOS: Applying - Iptable Down Rules
Feb  9 15:00:36 Skynet: [%] Startup Initiated... ( skynetloc=/tmp/mnt/EXT4/skynet )
Feb  9 15:00:36 (install_stubby.sh): 29303 Ending Script Execution
Feb  9 15:00:36 adaptive QOS: Applying - Iptable Up   Rules (eth0)
Feb  9 15:00:36 adaptive QOS: TC Modification Delayed Start (5min)
Feb  9 15:00:58 Skynet: [#] 149499 IPs (+0) -- 1704 Ranges Banned (+0) || 0 Inbound -- 0 Outbound Connections Blocked! [start] [22s]

Notice that I see a "starting script execution" and a "ending script execution" log entry above. I don't see the "starting script execution" in my logs when I reboot. Only "ending script execution" logs twice.

 

[Butterfly Bones]

Ok the router restarted firewall-start. This ended up calling nat-start. the syslog shows this which is encouraging.

Feb  9 15:00:35 nat: apply nat rules (/tmp/nat_rules_eth0_eth0)
Feb  9 15:00:35 custom_script: Running /jffs/scripts/nat-start
Feb  9 15:00:35 (install_stubby.sh): 29303 Starting Script Execution
Feb  9 15:00:36 custom_script: Running /jffs/scripts/firewall-start (args: eth0)
Feb  9 15:00:36 adaptive QOS: Applying - Iptable Down Rules
Feb  9 15:00:36 Skynet: [%] Startup Initiated... ( skynetloc=/tmp/mnt/EXT4/skynet )
Feb  9 15:00:36 (install_stubby.sh): 29303 Ending Script Execution
Feb  9 15:00:36 adaptive QOS: Applying - Iptable Up   Rules (eth0)
Feb  9 15:00:36 adaptive QOS: TC Modification Delayed Start (5min)
Feb  9 15:00:58 Skynet: [#] 149499 IPs (+0) -- 1704 Ranges Banned (+0) || 0 Inbound -- 0 Outbound Connections Blocked! [start] [22s]

Notice that I see a "starting script execution" and a "ending script execution" log entry above. I don't see the "starting script execution" in my logs when I reboot. Only "ending script execution" logs twice.

Weird. The good thing about the above is the starting and stopping are the same pid.
When you see the two ending ones, check the pid of each to see what each process is, that should give you a better clue. (fingers crossed)

 

[skeal]

So if I restart the firewall or call "nat-start" the log shows a install_stubby.sh "start" and a "end" with the same pid. If I reboot I get two ends with different pids.

 

[Butterfly Bones]

So if I restart the firewall or call "nat-start" the log shows a install_stubby.sh "start" and a "end" with the same pid. If I reboot I get two ends with different pids.

What processes are the two different pids?

 

[skeal]

What processes are the two different pids?

I believe both say:

Install_stubby.sh end of script execution

 

[Butterfly Bones]

I believe both say:

Install_stubby.sh end of script execution

Looking at these two lines from your logging:

Feb  9 15:00:35 (install_stubby.sh): 29303 Starting Script Execution
Feb  9 15:00:36 (install_stubby.sh): 29303 Ending Script Execution

Use

ps w | grep 2903

to find that process. Since you have two, run it again with each pid to see what processes those call.

 

[skeal]

Looking at these two lines from your logging:

Feb  9 15:00:35 (install_stubby.sh): 29303 Starting Script Execution
Feb  9 15:00:36 (install_stubby.sh): 29303 Ending Script Execution

Use

ps w | grep 2903

to find that process. Since you have two, run it again with each pid to see what processes those call.

Check out this post, it's my results from a restart:https://www.snbforums.com/threads/stubby-installer-asuswrt-merlin.49469/page-47#post-465034

 

[Butterfly Bones]

Check out this post, it's my results from a restart:https://www.snbforums.com/threads/stubby-installer-asuswrt-merlin.49469/page-47#post-465034

I saw that, so run the above ps command using 438 and the ps command again using 541 to see what processes those number are. Use the current lines from syslog to get those two numbers as they change every time you run a script that restarts stubby.