Oh yeah that’s a good idea too. Sure any one client device will be limited to 1g but your NAS could use a super cheap multi-port ethernet card to get 2 or 4g bonded link speeds so it can serve multiple devices at full speed.

Yup. Same age, same design, same failures… and array rebuilds are super intense workloads that often force a lot of random reads and run the drive at 100% load for many hours.

2.5 is still really new in the networking space and nobody has hit economies of scale yet. I very much also want to build out my home LAN to be entirely 2.5g compatible since 1g is limiting for my NAS use case (video storage), 10g is overkill and not supported by my client devices, and I only need 16/24 ports. but good God the hardware just isn’t reasonable yet.

You pretty much have to bite the bullet if you really want 2.5 right now. What might honestly be worthwhile is finding a used enterprise 1g switch with the number of ports you need, and will still be “enough”, as those can be had for only a couple hundred dollars. Sit on that for 2-3 years until the 2.5g and 5g hardware market starts to fill out and you can decide how badly you need 2.5g then

Really. Anything branded from Samsung or Crucial(Micron) is going to be fine. They are the top producers of NAND, produce high quality products, and stand behind warranties. But you are gonna pay out the nose for the privilege of enterprise grade hardware.

You might just be buying lower quality consumer SSD’s though, since even they should be able to handle a surprising amount of abuse.

If I had a dollar for every time rebuilding a RAID array after one failed drive caused a second drive failure in the array in less than 24 hours… I’d probably buy groceries for a week.

…I thought you just said you’re running OwnCloud?

FYI, read this.

I’ve switched almost exclusively to using RustDesk these days for anything outside of my LAN, and have had zero problems with it running on barebones machines, zero monitor connected. It’s open source!

If you’re wanting to use the windows built in RDP, there apparently is a DLL wrapper that enables the RDP server on Home installations. See github. I haven’t had to use it though because my machines are all, well, “legal” copies of 10 or 11 pro, so i cant speak to whether it works or not.

Is there trim in the crease or it is a sharp corner? If there is trim and its not painted clean over, you can cheat and use paperclips. Fold the long leg out straight, so its horizontal, and the short U curl down at a 90, so it forms a hook. If you stick the horizontal wire in behind the trim itll usually wedge in without damaging the trim or paint visibly and hold enough for a cat5e.

If no trim, the sticky hangars are probably the best. Be careful not to buy cheap ones that might fall down, leave residue, or stick too hard and rip the paint off when you go to remove it.

The ARM cpu will still beat out an i3/i5 in idle and load power consumption, and the mini PC also does not have GPIO for people who care about that kind of stuff.

There’s no such thing as overkill, only extra overhead to do more things with. Hell, if you found yourself with a ton of excess resources and good cooling, you could run a distributed computing project like BOINC on some of the spare cores and help out some scientists.

You wouldn’t see much of a bump in CPU performance, 6cores to 8 cores with a 200mhz clock speed improvement isn’t ground breaking.

Going to 8gb of memory will give caching benefits.

But… That’s all well and good. However. What I found the most beneficial on a OPi 5, and the entire reason I bought it over other boards, is the onboard NVME m.2 slot. Yes, the orange pi 5 can support 2230 and 2242 M.2 NVME drives at PCIe3.0x1 speeds, and it makes a WORLD of difference in performance. Like you would not even believe how fast compiling and installing software becomes when it’s not bottlenecked by the ~500 iops an SD card can struggle through. SD cards are ungodly slow, and OS level writes tend to kill them every few months (they’re not designed to handle that kind of work). Even the cheapest aliexpress M.2 drives, which I bought a 512gb KingSpec one for like $16, blow SD cards out of the water, and will last for YEARS with a typical pi’s workload compared to the few-months of an SD card. Plus they’re big enough to even do a bit of file hosting on.

The original Pi B had a single core 700mhz ARMv6 processor and 512mb of memory. It’s fine for embedded projects using GPIO or a mini LCD screen, but that’s about it. You’d be lucky to even decode 720p video on it as a streaming box.

It might work neat as a monitoring device to keep tabs on the rest of your homelab machines and display a status output or something.

As far as security goes, 1 open port and 1000 open ports contain basically the same level of risk exposure to the internet- a hole will always be a hole. If you need to open a few more for rustdesk, just do it, and be diligent about your firewall rules.

I’ve used VNC inside my LAN before with decent results, and it doesn’t break graphics drivers like windows RDP does. However, not FOSS.

You’re welcome to call their tech support and ask about using custom routing hardware, at least. It’s possible they already have an approved list of SFP’s that they would be willing to work with, just don’t hold your breath since they probably want (and need) to have management control over the entire router endpoint, not just the module.

2GB of memory is fine for openWRT. Routing is surprisingly light tbh, consider that most all home/SOHO routers run integrated SoC’s with <256MB of memory.

routing speed is more dependent on CPU +cache speed

i’d eat my boot if a residential ISP let you run your own SFP fiber module. they have to pretty tighttly control those things to keep signal levels right and have wavelengths in the right spots. plus they’ll need to upstream reconfigure it somewhat frequently as the local network changes and if it’s not their hardware, they’ll get mad.

I highly doubt homeassistant will ever abuse the pi4 CPU enough to make active cooling necessary. As long as you have a heatsink of some form on the SoC you’ll likely be perfectly fine. Obviously monitor it for core temps once installed.

Remember the Pi uses the entire PCB copper ground plane in the board as a heatsink too.

I dont know of any specific docker/web app, but its very simple command line job in linux actually if you just need to record what “normal” routing paths are for reference

Set up cron job to run a script once a day. in said bash script, get the current date using date "+%Y-%m-%d", store in a variable, then redirect the output of traceroute into a text file by doing traceroute > {variable}.txt Then you’ll have a day by day snapshot of your typical routing stack saved to your hard disk.

It won’t actively monitor it. But it’ll save a record. You can increase the frequency to as much as you want (hourly?) if you want too.

Well, whatever you end up using for documentation, print it out and actively maintain an up to date paper hard copy in a 3-ring binder somewhere. That way when all your shit falls over and you have a nonfunctional LAN you can still remember how everything was set up. Don’t ask me how I know…

Orange pi 5 can boot from SD and then use a NVME M.2-2242 SSD as storage. Or boot directly from the NVME if you set up their spi flash bootloader. Good memory and cpu options but they’re pricey for what you get. Still, an actual m.2 ssd will last much MUCH longer than any SD card when running an os and moving program data around…

They had a period of about 6 years starting around the Intel Ivy Bridge era where they actually made some really nice stuff. People started recognizing them as quality, but it’s like they put that effort in, saw results, and then went back into coast mode.