Got a HP EliteDesk 800 G4 Mini as my first homelab, running Proxmox VE, with one VM to run my services.
However I’m getting random hard resets every 1-2 days, causing my services to go offline, and having to manually restart the VM.
No kernel panic, OOM, or I/O errors. Just showing “crash” when I run last reboot .
Specs:
Power brick is 90W OEM HP (19.5V 4.62A).
-----------------------------------------------------
I’m about to run memtest overnight to rule out RAM.
Has anyone run 64GB in this model long-term and seen similar instability?
Is 90W borderline once you’re running 64GB + 2x NVMe + ZFS + VMs?
Anything else I should be checking before I replace the power adapter?
Wondering if anyone else has issues running these Minis as hypervisors.
However I’m getting random hard resets every 1-2 days, causing my services to go offline, and having to manually restart the VM.
No kernel panic, OOM, or I/O errors. Just showing “crash” when I run last reboot .
Specs:
- HP EliteDesk 800 G4 Mini
- i7-8700T
- 64GB RAM (2x32GB Samsung DDR4 2666 SODIMM, non-ECC)
- NVMe 1: SK Hynix PC611 256GB (OS)
- NVMe 2: Samsung 990 PRO 1TB (firmware 5B2QJXD7)
- ZFS on root
- 90W OEM HP power brick
- Proxmox VE (Debian trixie base)
- Debian VM running:
- WireGuard
- Gitea (Docker + Postgres)
- Joplin Server
- Light homelab services, nothing crazy load-wise
- No OOM events
- No kernel panic logs
- No MCE / hardware error logs
- NVMe SMART clean (0 media errors, no critical warnings)
- Temps normal
- ZFS ARC tiny (~250MB)
- unsafe_shutdowns incrementing on NVMe (suggesting abrupt power loss(?))
Power brick is 90W OEM HP (19.5V 4.62A).
-----------------------------------------------------
I’m about to run memtest overnight to rule out RAM.
Has anyone run 64GB in this model long-term and seen similar instability?
Is 90W borderline once you’re running 64GB + 2x NVMe + ZFS + VMs?
Anything else I should be checking before I replace the power adapter?
Wondering if anyone else has issues running these Minis as hypervisors.