This is more "home networking" than "homelab," but I imagine the people here might be familiar with what in talking about.

I'm trying to understand the logic behind ISPs offering asymmetrical connections. From a usage standpoint, the vast majority of traffic goes to the end-user instead of from the end-user. From a technical standpoint, though, it seems like it would be more difficult and more expensive to offer an asymmetrical connection.

While consumers may be connected via fiber, cable, DSL, etc, I assume that the ISP has a number of fiber links to "the internet." Those links are almost surely some symmetrical standard (maybe 40 or 100Gb). So if they assume that they can support 1000 users at a certain download speed, what is the advantage of limiting the upload? If their incoming trunks can support 1000 users at 100Mb download, shouldn't it also support 1000 users at 100Mb upload since the trunks themselves are symmetrical?

Limiting the upload speed to a different rate than download seems like it would just add a layer of complexity. I don't see a financial benefit either; if their links are already saturated for download, reducing upload speed doesn't help them add additional users. Upload bandwidth doesn't magically turn into download bandwidth.

Obviously there's some reason for this, but I can't think of one.

I generally try to stay informed on current events. With the exception of what gets posted here, I normally get my news from CNN. I tend to lean left politically, but not always.

The problem I always run into is that every news site I read, regardless of where they stand on the political spectrum, is always filled with pointless bullshit. Specifically, sports, celebrity news, and product placement. "Some shitty pop singer is dating some shitty actor" or "These are our recommendations for the best mass-produced garbage-quality fast fashion from Temu" or "Some overpaid dickhead threw a ball faster than some other overpaid dickhead."

What I'd love to find is a news source that's just news that matters. No celebrity gossip, sports, opinion pieces, etc. Just real events that have an impact on some part of the world. Legislation, natural events, economic changes, wars, political changes, that kind of thing.

Does this exist, or is all journalism just entertainment?

A few months ago, I upgraded all my network switches. I have a 16-port SFP+ switch and a 1GB switch (LAGG to the SPF+ with two DACs). These work perfectly, and I'm really happy with the setup so far.

My main switch ties into a remote switch in another building over a 10Gb fiber line, and this switch ties into another switch of the same model (on a different floor) over a Cat6e cable. These switches are absolute garbage: https://www.amazon.com/gp/product/B084MH9P8Q

I should have known better than to buy a cheap off-brand switch, but I had hoped that Zyxel was a decent enough brand that I'd be okay. Well, you get what you pay for, and that's $360 down the toilett. I constantly have dropped connections, generally resulting in any attached devices completely losing network connectivity, or if I'm lucky, dropping down to dial-up speeds (I'm not exaggerating). The only way to fix it is to pull the power cable to the switch. Even under virtually no load, the switch gets so hot that it's painful to touch. Judging from the fact that my connection is far more stable when the switch is sitting directly in front of an air conditioner, that tells me just about all I need to know.

I'm trying to find a pair of replacement switches, but I'm really striking out. I have two ancient Dell PowerConnect switches that are rock solid, but they're massive, they sound like jet engines, and they use a huge amount of power. Since these are remote from my homelab and live in occupied areas, they just won't work. All I need is a switch that has:

• At least 2 SFP+ ports (or 1 SFP+ port for fiber and a 10Gb copper port)
• At least 4 1Gb ports (or SFP ports; I have a pile of old 1GB SFP adapters)
• Management/VLAN capability Everything I find online is either Chinese white-label junk or is much larger than what I need. A 16-port SFP+ switch would work, but I'd never use most of the ports, and I'd be wasting a lot of money on overkill hardware. As an example, one of these switches is in my home office; it exists solely so I have a connection between my server rack, two PCs, and a single WAP. I am never going to need another LAN connection in my home office; any hardware is going to go in the server rack, but I do need 10GB connectivity on at least one of those PCs.

Does anyone have a suggestion for a small reliable switch that has a few SFP+ ports, is made by a reputable brand, and isn't a fire hazard?

I have been using the BlueIris NVR integration (from HACS) for quite some time, and it works great for triggering BI from HA. I've trying to do the opposite now: Fire off automations in HA whenever BI detects motion on one of my cameras.

I've never used MQTT before, so I'm learning as I go, but I think I have most of my setup configured properly. I've installed Mosquitto and the MQTT integration in HA. I've configured BI to connect to HA, and running "Test" in the "Edit MQTT Server" menu in BI shows a good connection and no errors. I've set my cameras to post an MQTT event when the alert is triggered (and I've verified that the alerts are in fact being triggered).

Nothing happens in HA, though. The "Motion" sensor for my camera in HA stays at "Clear." In fact, the history shows no change at all, ever.

I have the events in BI set up as follows: On Alert: MQTT Topic - BlueIris/&CAM/Status and Payload - { "type": "&TYPE", "trigger": "ON" } On Reset: Exactly the same, but change ON to OFF.

I've tried change the MQTT autodiscovery header in HA from "homeassistant" to "BlueIris," and it made no difference. The Mosquitto logs show a login from HA, so I feel like I'm close, but I'm not sure where else to look.

Edit: I installed MQTT explorer, and I've verified that the messages are making it to Mosquitto, and they appear to be correctly formatted.

UPDATE: I set the MQTT integration to listen to the MQTT messages coming from BI, and sure enough, they were coming through just fine. For some reason, the BI integration just wasn't seeing them. Digging through the system logs, I saw some errors "creating a binary sensor" coming from the BI integration. The only thing I can think is that because I didn't have MQTT set up when I first installed the BI integration, something went wrong with the config (although I had already rebooted the system several times). I re-downloaded the BI integration and re-installed it, and now everything works perfectly.

This isn't strictly "homelab" related, but I'm not sure if there's a better community to post it.

I'm curious what kind of real-world speeds everyone is getting over their wireless network. I was testing tonight, and I'm getting a max of 250Mbit down/up on my laptop. I have 4 Unifi APs, each set to 802.11ac/80Mhz, and my laptop supports 2x2 MIMO. Testing on my phone (Galaxy S23) gives basically the exact same result.

The radio spectrum around me is ideal for WiFi; on 5Ghz, there is no AP in close enough range for me to detect. With an 80Mhz channel width, I can space all 4 of my APs so that there's no interference (using a non-DFS channel for testing, btw).

Am I wasting my time trying to chase higher speeds with my current setup? What kind of speeds are you getting on your WiFi network?

I have been programming in C++ for a very long time, and like a lot of us, I have an established workflow that hasn't really changed much over time. With the exception of bare-metal programming for embedded systems, though, I have been developing for Windows that entire time. With the recent "enshittification" of Windows 11, I'm starting to realize that it's going to be time to make the switch to Linux in the very near future. I've become very accustomed to (spoiled by?) Visual Studio, though, and I'm wondering about the Linux equivalent of features I probably take for granted.

• Debugging: In VS, I can set breakpoints, step through my code line-by-line, pause and inspect the contents of variable on-the-fly, switch between threads, etc. My understanding of Linux programming is that it's mostly done in a code editor, then compiled on the command line. How exactly do you debug code when your build process is separate from your code editor? Having to compile my code, run it until I find a bug, then open it up in a debugger and start it all over sounds extremely inefficient.
• Build System: I'm aware that cmake exists, and I've used it a bit, but I don't like it. VS lets me just drop a .h and .cpp file into the solution explorer and I'm good-to-go. Is there really no graphical alternative for Linux?

It seems like Linux development is very modular; each piece of the development process exists in its own application, many of which are command-line only. Part of what I like about VS is that it ties this all together into a nice package and allows interoperability between the functions. I can create a new header or source file, add some code, build it, run it, and debug it, all within the same IDE.

This might come across as a rant against Linux programming, but I don't intend it to. I guess what I'm really looking for is suggestions on how to make the transition from a Visual Studio user to a Linux programmer. How can I transition to Linux and still maintain an efficient workflow?

As a note, I am not new to Linux; I have used it extensively. However, the only programming I've done on Linux is bash scripting.

I've noticed recently that my network speed isn't what I would expect from a 10Gb network. For reference, I have a Proxmox server and a TrueNAS server, both connected to my primary switch with DAC. I've tested the speed by transferring files from the NAS with SMB and by using OpenSpeedTest running on a VM in Proxmox.

So far, this is what my testing has shown:

• Using a Windows PC connected directly to my primary switch with CAT6: OpenSpeedTest shows around 2.5-3Gb to Proxmox, which is much slower than I'd expect. Transferring a file from my NAS hits a max of around 700-800MB (bytes, not bits), which is about what I'd expect given hard drive speed and overhead.
• Using a Windows VM on Proxmox: OpenSpeedTest shows around 1.5-2Gb, which is much slower than I would expect. I'm using VirtIO network drivers, so I should realistically only be limited by CPU; it's all running internally in Proxmox. Transferring a file from my NAS hits a max of around 200-300MB, which is still unacceptably slow, even given the HDD bottleneck and SMB overhead.

The summary I get from this is:

• The slowest transfer rate is between two VMs on my Proxmox server. This should be the fastest transfer rate.
• Transferring from a VM to a bare-metal PC is significantly slower than expected, but better than between VMs.
• Transferring from my NAS to a VM is faster than between two VMs, but still slower than it should be.
• Transferring from my NAS to a bare-metal PC gives me the speeds I would expect.

Ultimately, this shows that the bottleneck is Proxmox. The more VMs involved in the transfer, the slower it gets. I'm not really sure where to look next, though. Is there a setting in Proxmox I should be looking at? My server is old (two Xeon 2650v2); is it just too slow to pass the data across the Linux network bridge at an acceptable rate? CPU usage on the VMs themselves doesn't get past 60% or so, but maybe Proxmox itself is CPU-bound?

The bulk of my network traffic is coming in-and-out of the VMs on Proxmox, so it's important that I figure this out. Any suggestions for testing or for a fix are very much appreciated.

In c++17, std::any was added to t he standard library. Boost had their own version of "any" for quite some time before that.

I've been trying to think of a case where std::any is the best solution, and I honestly can't think of one. std::any can hold a variable of any type at runtime, which seems incredibly useful until you consider that at some point, you will need to actually use the data in std::any. This is accomplished by calling std::any_cast with a template argument that corresponds to the correct type held in the std::any object.

That means that although std::any can hold a type of any object, the list of valid objects must be known at the point that the variable is any_cast out of the std::any object. While the list of types that can be assigned to the object is unlimited, the list of types that can be extracted from the object is still finite.

That being said, why not just use a std::variant that can hold all the possible types that could be any_cast out of the object? Set a type alias for the std::variant, and there is no more boilerplate code than you would have otherwise. As an added benefit, you ensure type safety.

I'm looking for a portable air conditioner (the kind with 1 or 2 hoses that go to outside air). The problem I'm running into is that every single one I find has some kind of "smart" controller built in. The ones with no WiFi connectivity still have buttons to start/stop the AC, meaning that a simple Zigbee outlet switch won't work. I could switch the AC off, but it would require a button-press to switch it back on. The ones with WiFi connectivity all require "cloud" access; my IoT devices all connect to a VLAN with no internet access, and I plan to keep it that way.

I suppose I could hack a relay in place of the "start" button, but I'd really rather just have something I can plug in and use.

I can't use a window AC; the room has no windows. I'll need to route intake/exhaust through the wall. So far, I can't find any "portable" AC that will work for me.

What I'm looking for is a portable AC that either:

• Connects to WiFi and integrates with HA locally.
• Has no connectivity but uses "dumb" controls so I can switch it with a Zigbee outlet switch.

Any ideas?

Yesterday, Brian Dorsey was executed for a crime he committed in 2006. By all accounts, during his time in prison, he became remorseful for his actions and was a "model prisoner," to the point that multiple corrections officers backed his petition for clemency.

https://www.cnn.com/2024/04/09/us/brian-dorsey-missouri-execution-tuesday/index.html

In general, the media is painting him as the victim of a justice system that fails to recognize rehabilitation. I find this idea disgusting. Brian Dorsey, in a drug-induced stupor, murdered the people who gave him shelter. He brutally ended the life of a woman and her husband, and (allegedly) sexually assaulted her corpse. There is an argument that he had ineffective legal representation, but that doesn't negate the fact that he is guilty.

While I do believe that he could have been released or had his sentence converted to life in prison, and he could have potentially been a model citizen, this would have been a perversion of justice. Actions that someone takes after committing a barbaric act do not undo the damage that was done. Those two individuals are still dead, and he needed to face the ramifications for his actions.

Rehabilitation should not be an option for someone who committed crimes as depraved as he did. Quite frankly, a lethal injection was far less than what he deserved, given the horror he inflicted on others. If the punishment should fit the crime, then he was given far more leniency than was warranted.

I just set up a local instance of Invidious. I created an account, exported my YouTube subscriptions, and imported them into Invidious. The first time I tried, it imported 5 subscriptions out of 50 or so. The second time I tried, it imported 9.

Thinking there might be a problem with the import function, I decided to manually add each subscription. Every time I click "Subscribe," the button will switch to "Unsubscribe," then immediately switch back to "Subscribe." If I look at my subscriptions, it was never added.

My first thought was a problem with the PostgreSQL database, but that wouldn't explain why some subscriptions work when I import them.

I tried rebooting the container, and it made no difference. I'm running Invidious in a Ubuntu 22.04 LXC container in Proxmox. I installed it manually (not with Docker). It has 100GB of HDD space, 4 CPU cores, and 8GB of memory.

What the hell is going on?

As is stands, parents are able to claim their children as dependents on their tax returns, which lowers their overall tax liability and in effect means that the parents either pay less in taxes or receive a higher return at the end of each year.

Until they reach the age at which they can work, children are a drain on society. They receive public schooling and receive the same benefit from public services that adults do, yet they contribute nothing in return. At the point that they reach maturity and are gainfully employed and paying taxes, they become a functioning member of society.

If a parent decides to have a child, they are making a conscious decision to produce another human being. They could choose to get a sterilization surgery, use birth control, or abort the pregnancy (assuming they don't live in a backwards state that's banned it). Yet even if they decide to have 15 children, the rest of society has to foot the bill for their poor decisions until the child reaches adulthood.

By increasing taxes on parents instead of reducing them, you not only incentivize safe sex and abortion, but you shift the burden of raising a child solely to the individuals who are responsible for the fact that that child exists.

I am a strong advocate for social programs: Single-payer healthcare, welfare programs, low-income housing, etc, but for adults who in turn contribute what they can. A child should only be supported by the individuals who created it.

The majority of my homelab consists of two servers: A Proxmox hypervisor and a TrueNAS file server. The bulk of my LAN traffic is between these two servers. At the moment, both servers are on my "main" VLAN. I have separate VLANs for guests and IoT devices, but everything else lives on VLAN2.

I have been considering the idea of creating another VLAN for storage, but I'm debating if there is any benefit to this. My NAS still needs to be accessible to non-VLAN-aware devices (my desktop PC, for instance), so from a security standpoint, there's not much benefit; it wouldn't be isolated. Both servers have a 10Gb DAC back to the switch, so bandwidth isn't really a factor; even if it was, my switch is still only going to switch packets between the two servers; it's not like it's flooding the rest of my network.

Having a VLAN for storage seems like it's the "best practice," but since both servers still need to be accessible outside the VLAN, the only benefit I can see is limiting broadcast traffic, and as far as I know (correct me if I'm wrong), SMB/NFS/iSCSI are all unicast.

I've been running HA for a while, and it's been working well; I haven't had to change much in a few months. That being said, it's fun to tinker with it, and I'm curious to hear what kind of automations the rest of the community is using. What automations are you most proud of? What are your favorite? What kind of interesting automations have you written?

My personal favorite is an automation that displays the current "apparent" temperature on a Hue bulb. It takes an average of the temperature, humidity, and luminance around my property and uses the average to compute an "apparent" (feels like) temperature. Then it applies a cosine function to the apparent temperature (to approximate how people feel temperature change), uses the resulting value to calculate a level between blue and red in CIELAB (a perceptually uniform color space), converts the results to RGB, and sets the color value of the hue bulb. The result is a bulb that changes color so that the change in color (as perceived by the eye) mirrors how the temperature "feels" outside. Ultimately what that means is that we can look at a small lamp with the hue bulb and say "It feels cold outside; we should put on a coat." It's probably overkill, but it was a fun programming exercise. We've started saying things like "It's really blue today, I don't feel like going out."

I'd really enjoy reading what kind of interesting automations everyone else has written.

I have a decent amount of video footage that I'd like to share with friends and family. My first thought was Youtube, but this is all home videos that I really don't want to share publicly.

A large portion of my video footage is 4k/60, so I'm ideally looking for a solution where I can send somebody a link, and it gives a "similar to Youtube" experience when they click on the link. And by "similar to Youtube," I mean that the player automatically adjusts the video bitrate and resolution based on their internet speed. Trying to explain to extended family how to lower the bitrate if the video starts buffering isn't really an option. It needs to "just work" as soon as the link is clicked; some of the individuals I'd like to share video with are very much not technically inclined.

I'd like to host it on my homelab, but my internet connection only has a 4Mbit upload, which is orders of magnitude lower than my video bitrate, so I'm assuming I would need to either use a 3rd-party video hosting service or set up a VPS with my hosting software of choice.

Any suggestions? I prefer open-source self-hosted software, but I'm willing to pay for convenience.

I've been going through my system logs and working on resolving the miscellaneous errors; mostly it's just due to poorly-written automations where the automation would be called while it was still running. Easy fix.

What I can't seem to fix is a constant stream of "Unknown cluster command" errors on cluster 0xef00 coming from ZHA. I've discovered that the 0xef00 cluster is a manufacturer-specific cluster. All of my errors come from Tuya mmWave sensors; apparently Tuya uses this cluster for inter-device communication.

All of my devices work, but this error is polluting the logs to a large degree; right now it's showing over 100k instances of this error. Is there a way to have HA just ignore this cluster completely? It's not causing any issues with functionality, but I would rather the logs just show actual errors so I can more easily identify problems and fix them.

I've read that Z2MQTT doesn't have this issue, but swapping from ZHA to Z2MQTT is not an option for me, especially for what amounts to a logging issue. It would require re-pairing almost 100 Zigbee devices and modifying nearly every single automation.

I upgraded the head unit in my car recently. The head unit itself runs Android, and it supports Android Auto. So far, I've been using Android Auto via bluetooth, and it works great. I have no complaints.

I started using Android Auto just because it seemed logical, but I'm not understanding exactly what the benefits are. Since the head unit runs Android, couldn't I just install the apps I need on the head unit itself and just tether my phone for internet access? It also supports a 5G connection, so if I installed a SIM card, I don't think I'd need my phone at all. To be honest, I'm leading toward that; it just seems easier and a lot more straightforward.

I have no complaints about Android Auto, I just don't really see what it brings to the table other than a layer of abstraction over the head unit's native interface. It might be worth mentioning that the only thing I do in my car is streaming music and navigation.

What features am I missing? Surely there is a compelling reason for Android Auto to exist.

I'm using a variety of PIR motion sensors and mmWave presence sensors; most work fairly well, with a few exceptions. At this point, I have all the lights in my house automated, but with one exception: the master bedroom. I'd like to automate my bedroom lights so that they turn on when someone enters the bedroom unless someone else is already in bed sleeping. So far, none of the sensors I've used are precise or reliable enough to do this. I've thought about using the status of our phones (charging/not charging), but my girlfriend doesn't always plug in her phone when she's asleep. Scheduling won't work, since we both sleep at random times when we're off work, sometimes during the day. Maybe a pressure sensor under the mattress?

Aquara makes a device that's advertised as being able to detect multiple people as well as sleeping people. This would be perfect if it worked, but Aquara devices seem to be the ones that always cause me the most problems.

Any suggestions?

I've found that when I'm logged into Youtube, the algorithm works fairly well to suggest videos that are at least somewhat related to my interests. I'm specifically curious about how the "default" algorithm works when not logged in.

If I open a private window and look at the front page, it almost feels like the algorithm is doing its best to show me the opposite of what I want to see. Obviously this isn't true, but I just don't get how it chooses the videos it shows. As an example, I almost always get:

• Right-wing news clips (I'm not a Republican)
• Sports (I don't watch or play sports)
• Gaming streams (I've not once watched a gaming stream)
• Christian content (I am not a Christian)
• Gen-Z and Gen-Y entertainment (I'm almost 40)

I feel like some of this is geographic. My router load-balances between two internet connections, and I can sometimes tell if it's using my "local" connection or my satellite connection (with an endpoint in another state) based on what videos it shows. Regardless, though, the content I see isn't really appropriate for the demographic in either location. Out of curiosity, I tried it with a VPN using an endpoint in Canada (I'm in the USA); the front page was mostly really crappy reality TV content (think stuff on TLC).

If I was programming the algorithm, I'd want to have it show content that applies to a broad audience, but that really doesn't seem to be the case. While I don't intend on ever using Youtube without being logged in (and having a ton of browser plugins active to improve the experience), from a technical standpoint, I'm very curious how it chooses what to put on the front page.

I have a rainwater collection system that feeds into several water tanks. I'd like to be able to monitor the tank levels for a variety of reasons, first and foremost because of the fact that we get almost no rain in the summer, and I have run the tanks to empty before.

In the past, I have used Milone eTape connected to an Atmel microcontroller to monitor water levels for a few different applications. I'd like to use this again if possible.

The eTape essentially acts as a potentiometer in a circuit. I've done some research on various ways to feed this into HomeAssistant, but I'm not really sure what's best. My water tanks are just behind the wall from my HA server, so I could connect directly with USB, serial, or ethernet. WiFi or Zigbee or also options; nearly every other device I have connected to HA is Zigbee.

What I'm ideally looking for is a device that can take a voltage level between 0-5VDC, or a 5k potentiometer, and feed the result into HA. Building something isn't out of the question (as long as I can program it in C or C++), but an off-the-shelf solution that integrates with HA would be ideal.