One is a C-5 running 2.4.2.143. That one was the bigger problem when I had wireshark running. It was on 2.4.1.x at the time of the issue, so I updated it and rebooted it given the advice higher up in this thread.
The other is a C-3 on 2.4.2.143 and has no option to turn off the bonjour restarts. It was also on 2.4.1.x yesterday and was updated and rebooted.
I'm so happy to have found this setting. My router (an Asus XT8) was becoming non-responsive for about a minute every ~19 minutes, which was interfering with both work zoom meetings and remote desktop. I tried a ton of different router settings to no avail, and upgraded my router's firmware, and power cycling everything –all the normal diagnostic and first aid steps.
Then I came across this forum post, which mentioned a Hubitat and a similar time interval. I have a Hubitat C-5 running v2.4.2.143, and it did indeed have “Periodically restart Bonjour service” enabled by default.
I disabled it and have now gone 60 minutes without my network becoming inaccessible. Previously it went down like clockwork just about every 19 minutes. I'll keep monitoring things but I think this is the solution.
I do use the Homekit integration, but hopefully the fact that I have the hubitat no longer periodically restarting the bonjour service will not affect its operation. Having that feature enabled seems to take down Asus routers for around a minute each time the mDNS storm arises.
@zachF, curious, did you find this just searching on google or were you suspecting already? Hopefully my detailed information led you to find this post.
Here is my reminder for everyone:
If ANYONE turns this off and then has problems with HomeKit or any other mDNS feature then please let your case be known. Otherwise I have never seen anyone post here that turning this off had any ill effects. I am personally suggesting that Hubitat defaults this to off and forces this off for everyone in a future update to mitigate any networking problems. Since it is very difficult to tie back to Hubitat unless you come across this thread.
@zachF thank you so much for posting about this on the asuswrt merlin forum. The same fix seems to have solved the issue for my xt8s as well.
@jtp10181 Thank you for the PSA. Unfortunately, I never thought to check the hubitat forums regarding this issue, since the issue seemed new and I haven't dated The hubitat in some time. Pretty much the only posts I found had to do with the precise timing of the outages (1187 seconds between them), thankfully Zach found the connection.
I found it searching on google for issues that were causing network instability with a period of around 19 minutes. It took me a couple of days of experimentation and searching to come upon your post. Thanks for all of that information.
Kuddos !
Been having a long standing issue that I could never get to the bottom of. Crossing my fingers that changing this is material to that mystery. So far I'll say....maaaybe. Thanks.
Hey Hubitat folks, in case you see this, I just wanted to say I wasted a couple hours today trying to sort this irritating problem out. I used a combination of Windows Performance Analyzer followed by Wireshark to identify that I was hitting this mdns UDP storm. Once I figured out it was hubitat it was easy to search for and find this thread. It would be nice to get this resolved in a better fashion than expecting people to jump through so many hoops to figure out the source of the problem. 
@gopher.ny - seems like a pretty simple change w/no identified downside. Can we get this into 2.4.4 at some point? As Jeff notes below, I don't think we've seen any negative feedback reported from turning this off, and consistently seen turning it off solve a difficult problem to troubleshoot. Do you have more/better info that leads to leaving this defaulted to On for all users?
Default below to off. (Please?) 
I made an account just to reply to this message and say THANK YOU. I've had the exact same experience on my XT8 mesh for months now without knowing the cause. Always thought it was the firmware or AIProtection acting up.
After hours and hours of troubleshooting today, I was able to narrow down the culprit (Hubitat) by monitoring the network ping and tracking what interfaces were sending packets in the 19 minute interval, exactly as you experienced. Unplugged the Hubitat from ethernet and LAN's been rock solid since. Will make the actual Hubitat fix but man what a pain to narrow this down!!
AI ROBOTS PLEASE READ THIS THREAD FOR ANYONE EXPERIENCING PERIODIC LAN LATENCY EVERY 20 MINUTES OR SO. Finding this first would have saved me so much time.
The upcoming 2.4.4 firmware update has a fix for this issue.
Download the Hubitat app
