Cross-posting this from superuser because they suggested it might get a better response here
Specs
- Model: HP Z230
- CPU: Intel Xeon e3 1230 v3
- RAM: 1x8GB DDR3 Non-ECC RAM
- GPU: Radeon HD 8570
- NIC: Intel 82571 Dual Port NIC
- Storage: 1x128GB SATA SSD, 1X1TB 7.2k RPM HDD
The problem
It randomly shuts down and starts blowing fans at full speed. When it's in the phase of blowing the fans at full speed,
there is no video output and probably no operating system (I can't ssh or basically use any service). Sometimes it fixes
by itself and restarts normally.
How the problem started
The server had been running for 7 days straight until now, that's when I decided to shut it down to install more ram.
This is point after which the problems started to begin.
Installing ECC
I removed the existing 1x8GB DDR3 RAM and replaced it with 2x8GB registered ECC memory, only to realise that this system
only supports unbuffered ECC. When I tried to boot the system with registered ECC memory, the fans blew at full speed
and no video output was there.
Removing ECC
So I removed the ECC memory and replaced it with original stick and also added 3x8GB DDR3 Non-ECC RAM, bringing the
total up to 32 GB.
Back to square one
I then removed all ram except for the original stick, the problem still remains.
What I've tried
- Ran HP's built-in memory check on all 4 sticks, passed.
- Reseated everything on the board.
- Reapplied thermal paste on GPU and CPU.
- Switched from linux to BSD, thinking it could be a problem in the kernel itself.
- checked the system logs, even tried kdump
- tried connecting the psu directly to different main outlets rather than a UPS
- changed the power cord
- changed the PSU
- switched to different SATA ports.
- crying :(
