Alternatively cheapest fix is an m10 optane and nvme enclosure. You maybe need to enable usb quirks in the kernel depending on on enclosure. Costs about 20 bucks all in
That’ll keep the rasp online for essentially forever cause optane isn’t going to wear out
You can flash them to boot on an USB drive so any ssd, I have a RPI4 that had the sdcard slot ripped, flashed an sd card to instruct it to boot on usb, held it against the pins and now it boots without any sdcard.
Though I'd agree that if you got none the NUC is the better solution as if you add an ssd and power supply to the PI it becomes as if not more expensive for less power
Yup and I read TFA and it's not clear: "The death and rebirth of my home server" and, somehow, his Raspberry Pi is his server!? Running RAID, I'm Mich (immich), Jellyfin, etc. all from a Pi is amazing.
Don't get me wrong: I've got like seven Pis, 4 NUCs (but only one "recent" N100 one) and even though two of the Pis are up 24/7 (one runs the unbound DNS and the other, well, serves as a router/firewall to isolate my main PC from the rest of the machines) my server, as in the machine that has disks in RAID (ZFS), runs VMs, containers (including Immich and Jellyfin), etc. is a Xeon with ECC RAM.
I had no idea a Pi 4 was that beefy.
Now of course booting off a microSD card: yuk (I know, I know, there are read-only trickeries that can be achieved).
> The important stuff mostly lived in external HDDs and the configuration lived almost entirely in my NixOS configuration.
Yeah NixOS or what-have-you: if you have server, I'd recommend you have a plan B for when it fails. My little infra at home is mostly automated: it's all scripts provisioning the VMs, setting up cloudinit for the first boot, etc.
As long as I've got backups of my config/scripts and of my data, setting up a new server really doesn't take long (heck I did it not long ago, mostly to be sure the "death and rebirth" procedure of my own server was sound).
I personally prefer my 10+ years old Xeon with ECC as a server than my N100 NUC but then the Xeon consumes more energy... YMMV.
Classic power failure and raspberry pi with SD cards. When I last set up things for my parents, I simply imaged a bunch of SD cards and had the data stored on an external hard drive. I live in the US and they live in India and this is for their RaspBMC. With the SD cards they just swap a broken one with a new one and call it a day.
Been 12 years or so now. Went back to their home a few years ago and everything but the opensubtitles integration works. I was surprised. I could never recreate it now. Even has one of these flirc setups so the remote works.
India has unreliable power too so the SD card death was inevitable. I made a few more images and called it a day.
I'm not a sysadmin but I often hear about server failures due to hardware just wearing out. Is there an expected shelf life of the hardware? I would expect for sure on the storage devices. So if that were the case then why isn't there some sore of daemon or system software service that tracks the expected life of the hardware? I would expect that the software would start showing warnings of imminent useful life of devices to expire.
My primary home zfs server is from 2010, running 24x7 since then. It has a 4-way mirror zfs pool for data. Over the years I have replaced two of the four drives, so two are still original from 2010. Everything else on the machine is from 2010.
You can get unlucky and have components fail quickly, but usually if they last, they'll last for a long time.
But trying to run a server from an SD card like in the article, seems silly. False economy, just get a good SSD.
For drive health monitoring there is SMART, but IME while it does monitor drive usage and lifetime, it rarely predicts catastrophic failure since that happens suddenly. Both drives that failed in this system went from healthy (according to SMART) to unresponsive overnight.
There are plenty of sensors and systems that can "predict" failure and in some cases try and work round them.
Do most people actually use them? Pretty rarely. S.M.A.R.T is just trash anyway.
But you do need to properly design your systems to limit the amount of "use". No un-needed writes to disk etc. Proper ambient detection, no point trying to rush to shut down servers when the AC pops it, and you just so happen to walk into a server room.
Another useful advise would be to keep your boot drive 60% used at max and run TRIM after freeing a huge chunk of files to let the wear leveling algo do its job.
And every few years reinstall the OS to free up unwritten cells.
My Raspberry Pi 4 has been used to run both a Jitsi and Trackmania server over the years. I never considered using the SD card since it can boot and run from a USB 3 flash drive. Later on, I picked up an Argon One case to add a SATA SSD.
Thankfully, modern Rockchip SBCs tend to come with NVMe slots.
If i'm reading this correctly, can run Immich from a Raspberry Pi 4? That's pretty cool.
I bought all the stuff (new case, networking card and a Jonsbro N4 to repurpose some intel i7-9700 era processor and motherboard, but didn't get it setup yet, as I was starting to rethink the power consumption....
Also looking to use this rebuild as an opportunity to learn Nix.
Let me know if you all have some good article references, to point me at. I've been doing some research, but could always do more.
I may end up putting a low power NUC in the case, or something. I don't know...
I use a Waveshare hat for my Pi 5s. It does PoE and has a spot for an SSD. What I particularly like about this setup is the single wire running to each board, looks very clean.
> Not eager to ruin the evening by starting a debugging session
Use a clanker. Nothing has brought back so much joy into homelabbing for me as having LLMs handle the boring stuff.
I launch it in opencode on the system or elsewhere to ssh in (I think giving it KVM is an overkill but an option nonetheless) and tell it to fix things. Not just ask questions and generate configs. I have backups, let it rip. They have gotten shockingly good at it even when they write bash wrappers just to catch some logs.
> Swap is not generally about getting emergency memory, it's about making memory reclamation egalitarian and efficient. In fact, using it as "emergency memory" is generally actively harmful.
Lack of proper IO connector is the reason I abandoned Raspi. SD card crashes are nearly a guarantee. Total pita too since there’s no display during the BIOS.
34 comments
[ 4.7 ms ] story [ 52.2 ms ] threadThat’ll keep the rasp online for essentially forever cause optane isn’t going to wear out
I agree that its a good, long-lasting fix, but when was the last time you checked the pricing on those?
If you can even find any, they certainly wont be cheap nowadays!
Though I'd agree that if you got none the NUC is the better solution as if you add an ssd and power supply to the PI it becomes as if not more expensive for less power
Don't get me wrong: I've got like seven Pis, 4 NUCs (but only one "recent" N100 one) and even though two of the Pis are up 24/7 (one runs the unbound DNS and the other, well, serves as a router/firewall to isolate my main PC from the rest of the machines) my server, as in the machine that has disks in RAID (ZFS), runs VMs, containers (including Immich and Jellyfin), etc. is a Xeon with ECC RAM.
I had no idea a Pi 4 was that beefy.
Now of course booting off a microSD card: yuk (I know, I know, there are read-only trickeries that can be achieved).
> The important stuff mostly lived in external HDDs and the configuration lived almost entirely in my NixOS configuration.
Yeah NixOS or what-have-you: if you have server, I'd recommend you have a plan B for when it fails. My little infra at home is mostly automated: it's all scripts provisioning the VMs, setting up cloudinit for the first boot, etc.
As long as I've got backups of my config/scripts and of my data, setting up a new server really doesn't take long (heck I did it not long ago, mostly to be sure the "death and rebirth" procedure of my own server was sound).
I personally prefer my 10+ years old Xeon with ECC as a server than my N100 NUC but then the Xeon consumes more energy... YMMV.
Been 12 years or so now. Went back to their home a few years ago and everything but the opensubtitles integration works. I was surprised. I could never recreate it now. Even has one of these flirc setups so the remote works.
India has unreliable power too so the SD card death was inevitable. I made a few more images and called it a day.
You can get unlucky and have components fail quickly, but usually if they last, they'll last for a long time.
But trying to run a server from an SD card like in the article, seems silly. False economy, just get a good SSD.
For drive health monitoring there is SMART, but IME while it does monitor drive usage and lifetime, it rarely predicts catastrophic failure since that happens suddenly. Both drives that failed in this system went from healthy (according to SMART) to unresponsive overnight.
Do most people actually use them? Pretty rarely. S.M.A.R.T is just trash anyway.
But you do need to properly design your systems to limit the amount of "use". No un-needed writes to disk etc. Proper ambient detection, no point trying to rush to shut down servers when the AC pops it, and you just so happen to walk into a server room.
And every few years reinstall the OS to free up unwritten cells.
I now use a small SSD connected via USB3 and a slightly modified version of this : https://framps.github.io/raspiBackupDoc/introduction.html
...to take snapshots to my NAS periodically.
I keep a similar small SSD in the drawer ready for a restore and swap if the main system drive fails.
Thankfully, modern Rockchip SBCs tend to come with NVMe slots.
I bought all the stuff (new case, networking card and a Jonsbro N4 to repurpose some intel i7-9700 era processor and motherboard, but didn't get it setup yet, as I was starting to rethink the power consumption....
Also looking to use this rebuild as an opportunity to learn Nix.
Let me know if you all have some good article references, to point me at. I've been doing some research, but could always do more.
I may end up putting a low power NUC in the case, or something. I don't know...
Use a clanker. Nothing has brought back so much joy into homelabbing for me as having LLMs handle the boring stuff.
I launch it in opencode on the system or elsewhere to ssh in (I think giving it KVM is an overkill but an option nonetheless) and tell it to fix things. Not just ask questions and generate configs. I have backups, let it rip. They have gotten shockingly good at it even when they write bash wrappers just to catch some logs.
swap is used when you run out of ram. using a ramdisk for swap ... You see how that doesnt make sense?
This is not correct. Swap is not a RAM overflow.
> Swap is not generally about getting emergency memory, it's about making memory reclamation egalitarian and efficient. In fact, using it as "emergency memory" is generally actively harmful.
https://chrisdown.name/2018/01/02/in-defence-of-swap.html
You can compress the swap, so you can squezze a bit more by using CPU in that tiny amount of read you do.
Back in the day I ran s××× coin miners off flash drives without any apparent problems.
Teslas though... ;)
every other component is cheap enough but not Ram.
But this is the kind of experience that makes technical people fall in love with self-hosting, but also keeps the majority of the population away.
Most of the people don't want the hassle of managing a self-hosting box.
I think a middle ground option might where someone hosts it for you, but you can also move to your own hardware at any time might be best