Beta Asuswrt-Merlin 3004.388.6_x test builds (dnsmasq 2.90)

[TomMe]

Updated my AX86U about 12 hours ago. Used some with and without DNSSEC. No problems have been noticed.

 

[aex.perez]

The router decided to time travel early this morning, I was done with SSH a little after midnight, then it popped back at 3:25AM to right where it left off

Feb 20 00:23:30 Router dropbear[30172]: Exit (Master) from <192.168.1.99:51068>: Error reading: Connection reset by peer
Feb 20 03:25:25 Router rc_service: service 19702:notify_rc restart_dnsmasq
Feb 20 03:25:25 Router custom_script: Running /jffs/scripts/service-event (args: restart dnsmasq)
Feb 19 16:08:12 Router dnsmasq[4022]: started, version 2.90 cachesize 1500
Feb 19 16:08:12 Router dnsmasq[4022]: compile time options: IPv6 GNU-getopt no-RTC no-DBus no-UBus no-i18n no-IDN DHCP DHCPv6 no-Lua TFTP no-conntrack ipset no-nftset no-auth cryptohash DNSSEC no-ID loop-detect no-inotify no-dumpfile
Feb 19 16:08:12 Router dnsmasq[4022]: DNSSEC validation enabled
Feb 19 16:08:12 Router dnsmasq[4022]: configured with trust anchor for <root> keytag 20326
Feb 19 16:08:12 Router dnsmasq[4022]: warning: interface br2 does not currently exist
Feb 20 03:25:25 Router dnsmasq[4022]: overflow: 18 log entries lost
Feb 20 03:25:25 Router custom_config: Appending content of /jffs/configs/dnsmasq.conf.add.
Feb 20 03:25:25 Router stubby[19795]: Read config from file /etc/stubby/stubby.yml

Only dnsmasq had this, everything else, every other script was logging results fine...

Other than this little anomoly, every once in a random while a web page fails to load completely, particularly those that contain multiple elements that also need DNS resolution or include ads and stuff. A quick refresh cleans it back up and the requested page loads properly. Due to the randomness and the fact that it only happens for a few minutes at a time I've not been able to capture it to insolate this as an issue with this version of dnsmasq. I've removed the browser extenision Adblocker (wasn't a problem with 388.* before dnsmasq 2.90) and it didn't help. I ve also tried various browers. I've got no Adblocker scripts on the router either. I haven't reverted to 388.6 to see if I can replicate this but as I said this didn't occur with any 388.*, I suppose I should revert if only to rule out other factors.

Not sure where to look, let alone how to collect dnsmasq metrics when this occurs.
Any ideas so I don't have to revert to test this?

 

[John Fitzgerald]

every once in a random while a web page fails to load completely, particularly those that contain multiple elements that also need DNS resolution or include ads and stuff. A quick refresh cleans it back up and the requested page loads properly. Due to the randomness and the fact that it only happens for a few minutes at a time

Same and some slowness when first navigating a site (inconsistent), (dnssec = on)

 

[ColinTaylor]

The router decided to time travel early this morning, I was done with SSH a little after midnight, then it popped back at 3:25AM to right where it left off

Your post is impossible to read with certain forum styles. Can you not post using psychedelic colours. Thanks.

 

[John Fitzgerald]

Your post is impossible to read with certain forum styles. Can you not post using psychedelic colours. Thanks.
View attachment 56638

You are not a fan of dark mode?

 

[aex.perez]

Your post is impossible to read with certain forum styles. Can you not post using psychedelic colours. Thanks.
View attachment 56638

Guessing italics is ok?

 

[visortgw]

Guessing italics is ok?

Yes. Unfortunately, colors display differently based upon light vs dark background mode.

 

[aex.perez]

Same and some slowness when first navigating a site (inconsistent), (dnssec = on)

I can't say its slowness for me, DNSSEC is on though but the pages don't load complete, but now that I see this I'm going to try Firefox set to use DNSSEC and use different providers to see if it happens with that as well...

 

[dave14305]

Yes. Unfortunately, colors display differently based upon light vs dark background mode.

Still looks bad. Invisible ink…

 

[aex.perez]

Still looks bad. Invisible ink…

View attachment 56639

try again,

 

[bennor]

The color tag removed for readability. Expand quote to see their full post.

The router decided to time travel early this morning, I was done with SSH a little after midnight, then it popped back at 3:25AM to right where it left off

Feb 20 00:23:30 Router dropbear[30172]: Exit (Master) from <192.168.1.99:51068>: Error reading: Connection reset by peer
Feb 20 03:25:25 Router rc_service: service 19702:notify_rc restart_dnsmasq
Feb 20 03:25:25 Router custom_script: Running /jffs/scripts/service-event (args: restart dnsmasq)
Feb 19 16:08:12 Router dnsmasq[4022]: started, version 2.90 cachesize 1500
Feb 19 16:08:12 Router dnsmasq[4022]: compile time options: IPv6 GNU-getopt no-RTC no-DBus no-UBus no-i18n no-IDN DHCP DHCPv6 no-Lua TFTP no-conntrack ipset no-nftset no-auth cryptohash DNSSEC no-ID loop-detect no-inotify no-dumpfile
Feb 19 16:08:12 Router dnsmasq[4022]: DNSSEC validation enabled
Feb 19 16:08:12 Router dnsmasq[4022]: configured with trust anchor for <root> keytag 20326
Feb 19 16:08:12 Router dnsmasq[4022]: warning: interface br2 does not currently exist
Feb 20 03:25:25 Router dnsmasq[4022]: overflow: 18 log entries lost
Feb 20 03:25:25 Router custom_config: Appending content of /jffs/configs/dnsmasq.conf.add.
Feb 20 03:25:25 Router stubby[19795]: Read config from file /etc/stubby/stubby.yml

Only dnsmasq had this, everything else, every other script was logging results fine...

Other than this little anomoly, every once in a random while a web page fails to load completely, particularly those that contain multiple elements that also need DNS resolution or include ads and stuff. A quick refresh cleans it back up and the requested page loads properly. Due to the randomness and the fact that it only happens for a few minutes at a time I've not been able to capture it to insolate this as an issue with this version of dnsmasq. I've removed the browser extenision Adblocker (wasn't a problem with 388.* before dnsmasq 2.90) and it didn't help. I ve also tried various browers. I've got no Adblocker scripts on the router either. I haven't reverted to 388.6 to see if I can replicate this but as I said this didn't occur with any 388.*, I suppose I should revert if only to rule out other factors.

Not sure where to look, let alone how to collect dnsmasq metrics when this occurs.
Any ideas so I don't have to revert to test this?

 

[visortgw]

try again,

Another option is to use CODE tags for readability of logs:

Log data

 

[Wisiwyg]

For those running PiHole, new version of FTL incorporating these changes is up. FTL 5.25.1

 

[bbunge]

Went back to Merlin a couple of days ago. Loaded up the test version this afternoon and all seems well. Dig shows the AD flag for DNSSEC.

 

[DJones]

I feel like I’m late to the show for this function, but I just noticed Traffic Classification now works with Cake! Used to say it wasn’t compatible I believe on earlier Merlin versions.

 

[learning_curve]

I don't... normally run Beta versions of Firmware, but seeing as I make full use of DNSSEC (pic) I thought it relevant to try the Beta 2 release and post my own findings.
In my case, zero problems, zero issues appearing in the logs, zero collateral damage elsewhere (as a result of running this Beta2 release).
It's all good for me & the setup that I use.

 

[Damit1]

2 days uptime.

Didnt noticed any issue/ slowness.
I'd suggest to maintain a list of "slow-loading" websites so we can all try.

Although, I believe it is related already to the DNS query speeds of the DNS server, and not with the firmware/patch.
Im using CloudFlare and Google, DoT and DoH enabled.

 

[bbunge]

2 days uptime.

Didnt noticed any issue/ slowness.
I'd suggest to maintain a list of "slow-loading" websites so we can all try.

Although, I believe it is related already to the DNS query speeds of the DNS server, and not with the firmware/patch.
Im using CloudFlare and Google, DoT and DoH enabled.

If your browsers have DoH enabled then you are by-passing the router DNS.

 

[DJones]

Side topic: Only thing DoT is missing is ECH/ESNI, unfortunately from Cloudflare/Firefox it’s looking like support is going to push toward DoH browser support instead. https://blog.cloudflare.com/announcing-encrypted-client-hello

 

[aex.perez]

2 days uptime.

Didnt noticed any issue/ slowness.
I'd suggest to maintain a list of "slow-loading" websites so we can all try.

Although, I believe it is related already to the DNS query speeds of the DNS server, and not with the firmware/patch.
Im using CloudFlare and Google, DoT and DoH enabled.

Various devices, different configs, not webite specific
All using Cloudflare
Chrome and others using DOT
Firefox using DOH

Any website, happens at random and only lasts for a few minutes at a time, but only with Chrome (and others) DOT on the PC, iPhone, iPand, Tablets, Laptop (Router DNS Cloudflare).
For example, on the Reddit news feed, click on any article any one redirects you to the news source by getting a redirect from out.reddit.com. When DNS fails no redirecttion from out.reddit.com, you get a blank page and that's it. Though after a few minutes it works as it should again. Clicking on a news story the redirection from out.reddit.com then works. With other sites, they will partially load requiring a browser refresh to load completely (happened with snbforums.com several times when it randomly started failing, until it started working again), but only with Chrome, Edge, Safari and so on.

Firefox / Cloudflare DOH, bypassing router DNS has not failed yet, or side by side when the non DOH is failing or shown this behavior.

Firefox set to use DOH - Chrome / Edge / Safari (Any Non DOH browser)

When it fails, at least from a perception perspective it seems that some process is not getting the resources it needs as it feels as if websites a loading slower, maybe running TOP/HTOP during and after might shed some clues as it lasts for a few minute, because it doesn't show up in the log .

I have noticed that the random failure happens when this shows up in the log "rc_service: service 15014:notify_rc restart_dnsmasq" though I can't say 100% for sure, because too much has been going on when it happens on to correlate to this.

Also planning to use DNS Director and laptop (static IP) to DNS - 8.8.8.8 to bypass both DOT, DOH and Cloudflare as well to see if it happens that way.
Need more to go on....