no, only via POE. I’ll redo following your steps on the weekend to see if that fixes this occasional reboot loop issue.
Just to confirm I did these steps too
dtparam=pciex1
dtparam=pciex1_gen=3
Enable auto detection PCIe and booting from NVME.
sudo rpi-eeprom-config --edit
PCIE_PROBE=1
BOOT_ORDER=0xf416
but not these one as per Geekpi wiki
sudo fdisk /dev/nvme0n1
and then input:
> p
> n
> p
> 1
> Enter
> Enter
> w
> q
* Format the partition
<pre>
sudo mkfs.ext4 /dev/nmve0n1p1
Create mounting point and mounting the partion to the mounting point.
mkdir ~/mydata
sudo mount -t ext4 /dev/nvme0n1p1 /home/pi/mydata -v
Automount configuration, edit /etc/fstab file and adding following parameters:
sudo vim.tiny /etc/fstab
adding:
/dev/nvme0n1p1 /home/pi/mydata ext4 defaults,noatime 0 0
While you did neither of these… which also means you’re not running PCI Gen3.
Hi,
I have been running this setup since June, and it, well, sort of works. It is running well enough, at least to a degree that I didn’t put any energy into trying to fix my issues.
So what are the issues? Well, in a random fashion, the Pi just loses the disk. I haven’t found any pattern or managed to see anything in the logs, but sometimes everything works as a charm for a month, and sometimes it just loses the disk after 12 hours.
The easy fix is a power cycle and then everything is fine… For a while
My main problem now is that when the winter is coming, I can’t have this behavior anymore as some of my radiators are controlled by simple thermostats.
Does anyone else have similar issues?
Any ideas would be very appreciated!!
Using geekworm X1001 with a crucial P3 500G gen 3.
Maybe I will have to create some external watchdog or go back to SD Card…
Edit: And yes, I’m using the recommended standard pi5 power supply.
Well, as long as I stay on HAOS 13.2, it works now. HAOS 14.0 makes it extremely unstable. Four out of five times, it doesn’t even manage to boot up HA and throws disk errors all over the place.
I’ll have to stay on 13.2 for now.
EDIT:
Sorry for spamming this thread but hopefully it might help someone else with similar issues.
After rolling back to 13.2 I’m now back to where I was BEFORE changing the dtparam settings in config.txt. Pi5 now drops the NVMe after 12-24 hours again. Checking logs now again shows i/o errors.
I wonder if it’s Wi-Fi/BT interference. I’ve read more than once about people having RF-related issues on Pi5 with NVMe hats.
One example: https://forums.raspberrypi.com/viewtopic.php?t=368054&sid=0baae0cf153eecce78a74671403f5c46#p2208203
Hey just want to say thank you for posting a link to that foil thing guide. I was having the same issue of Pi5 randomly freezing and the foil thing fixed it for me on my GeeekPi P33 M.2 NVME M-Key PoE+ with Kingston Nvme drive. Soo yeah… if you guys are having issues with lockups try wrapping your ribbon cable in foil
I had been running HA in my Synology as a virtual instance. It ran well but recently moved out to a Pi5 in an Argon ONE V3 M.2 NVME case. It has been dropping out daily for a couple days. Power cycling it brought it back fine but is annoying. I was using an official RPi 27-watt power supply. I swapped out the power supply for an Anker 67-watt (over kill, but it is what I had) and so far the Pi has been running stable. It looks go so far