There’s a set of special topics under homeassistant/ that devices also publish to that describe what each topic does and how HA should present it. HA will subscribe to everything under that root topic to discover all your MQTT devices.

Just updated and it looks like this one fixed the log spam:

json_loads was called from hacs, this is a deprecated function which will be removed in HA Core 2025.8. 
Use homeassistant.util.json.json_loads instead, please create a bug report at https://github.com/hacs/integration/issues

It’s a little weird they don’t have a download update button on the new HACS dashboard for an individual repository, now you have to go to Settings > Updates. I also wish I could hide new and available repositories and only show the ones I have installed (you can’t seem to select Pending Restart, Pending Update, and Downloaded at the same time.)

There’s two main ways of doing geo-based load balancing:

1. IP Any-casting - In this case, an IP address is “homed” in multiple spots and through the magic of IP routing, it arrives at the nearest location. This is exactly how 1.1.1.1 and 8.8.8.8 work. It works fine for stateless packets like DNS, however it has some risks for stateful traffic like HTTP.
2. DNS based load balancing. A server receives a request for “google.com”, looks at the IP of the DNS server and/or the EDNS Client IP in the DNS query packet and returns an IP that’s near. The problem is that when you’re doing Wireguard, it goes phone -> pi-hole (source IP is some internal IP) -> the next hop (e.g. 1.1.1.1 or 8.8.8.8), which sees the packet is coming from your home/pi-hole’s public IP. Thus it gets confused and thinks you’re in a different location than you really are. Neither of these hops really knows your true location of your phone/mobile device.

Of course, this doesn’t matter for companies that only have one data center.

Sorry, what do you mean route it directly? Maybe I didn’t clarify well enough.

My DNS is routed over the VPN but Internet traffic is routed directly. The problem is the load balancing is done based on where the DNS server is so say Google even though the traffic egresses directly to the internet bypassing the VPN it still goes to a Google DC near my home. Not all websites do this so its not always an issue.

Yes, but if you hit a company doing DNS based load balancing, DNS is going to return an IP that’s near to your DNS server which may not be near your device. That’s going to add to the latency.

I have Wireguard and I forward DNS and my internal traffic from my phone over the VPN to my pi-hole at home. All other traffic goes directly over the Internet, not the VPN. So that means only DNS encounters higher latency.

However, because a lot of companies do DNS based geo load balancing that means even if I’m on the east coast all my traffic gets sent to the West Coast because my DNS server is located there. That right there has the biggest impact on latency.

It’s tolerable on the same continent, but once I start getting into other continents then it gets a bit slow.

If you’re using it, Home Assistant natively supports Wake On Lan. This would only be able to handle the shutdown/sleep side of things.

Do the services themselves have any logs? Do you have a reverse proxy? Does it provide any logs? 503 means something received the request and tried to pass it on so something should have logs.

Totally. I used to contribute to Google maps quite a bit and got higher up in the Local Guides levels, but now I find myself contributing a lot to OSM. I feel a lot better about contributing to an open platform vs letting a company close up my changes.

I just haven’t made the switch to use it as a mobile client yet

I’ve been eagerly looking forward to the time when I can replay my Echo Dots with a self-hosted solution, but so far I haven’t found hardware that I really liked the look and style of.

Accidentally typo your password and get blocked. And if you’re tunneling over tor, you’ve blocked 127.0.0.1 which means now nobody can login.

Paperless does support defining a folder structure that you can use to organize documents within that paperless media volume however you should treat it as read only.

OP could use this as a way to keep their desired folder structure as much as possible, but it would have to be separate from the consumption folder.

Yeah I always use states when I can but the original post description made it sound like the integration was directly sending a notification. If it didn’t set a state (which would be weird) then you’d need an event.

https://www.home-assistant.io/docs/configuration/events/

In the dev tools > events tab you can subscribe and see what events are sent. There should be an event that is sent when it sends a notification. Create an automation that listens for that event

One of the problems with the cloud-polling integrations is that they will frequently poll the back-end APIs to get the current status of that device. A normal user might only open up the app once or twice a day and call the APIs, but these integrations will go 24/7 every 10s-5m. That can add up to a non-trivial amount of traffic. If there’s 100 users opening it up once a day, that’s not a lot of traffic, but 10 users polling every 1 minute is equivalent to 15k people doing something once a day.

I actually saw one of my integrations I used defaulted to updating every 10 seconds. I decreased that because I didn’t want to draw attention to it.

A business will look at their usage and ask why there’s more than expected traffic. They could be running their server on a potato. They could go back and support Matter, that costs money, requires skilled engineers, and cuts into profit margins.

While it sucks, that is something they could point to in a court about “economic harm”.

I don’t fully understand what you’re saying, but let’s break this down.

Since you say you get an NGINX page, what does your NGINX config look like? What exactly does the NGINX “login page” say? Is it an error or is it a directory listing or something else?

Then try something like:

Create Quanity unit of ml and a liter unit

In your product use: Unit stock: bottle or liter Unit purchase: bottle Consume: ml Price unit: ml

Set a product specific QU conversion of bottle to ml

Weirdly, the quick consume unit is based on the stock unit, not the consume unit. That seems like a bug.

The problem with Grocy is that going too fine grained means you’re unlikely to keep it up to date or it be accurate. I would not try to track your usage in ml. Just track it at the bottle level.

However you can still track the price per ml because grocy lets you independently set units. Just define a mapping between bottle and ml.

It just updated on my phone to the new icon. I tried to give it a chance but wow that looks not great. Something about the scale and lack of discerning features.

Will I still need to consider multicast DNS if my DNS server is on-prem (Pi-Hole + Unbound)

Multicast DNS is separate from DNS, so even if you have Pi-Hole, you’d still have devices using mDNS. It’s possible to route mDNS across separate IP networks seeing as how there’s mDNS relays across VLANs which would suggest Wireguard could support Multicast. Other things use Broadcast (e.g. WoL) which is a bit more challenging to forward across IP networks.

I’m not familiar with GRE so I couldn’t comment on whether it’s possible or not. I guess it all depends on how confident you are with your networking skills. If you get it working, you should definitely document it and share with others.

I didn’t quite do what you did, but I ran HA in a Kubernetes cluster which was logically a separate IP network. I had to setup the container with multiple network interfaces and specially craft the route table to forward broadcasts + multicast traffic to the correct network.