I have a rock pi 4 with a POE hat and the heatsink that goes underneath. I just plugged it into my kubernetes cluster and I noticed that services running on it seem to be gone after 20 or 30 mins. Thats when I noticed that the rock pi 4 board reboots by itself. Any idea about this ? Should I send it back ?
What about hiring someone to reason / debug the problem? I don’t (have) power via POE Hat but otherwise boards are not rebooting for no reason or act strange. In most cases software can fix things … unless design is failed beyond repair.
I ran an extreme stress testing session with Rockpi and … no problems:
I have a rock pi 4 with a POE hat and the heatsink that goes underneath.
Which poe hat are you using? Is your poe switch 802.3at or 802.3af? For 802.3af, 13W is not enough for ROCK Pi 4.
From my POE switch specification:
IEEE 802.3at Power over Ethernet (PoE+)
Provides up to 30 W per port, which allows support of the latest PoE+ capable devices such
as IP phones, wireless Access Points, and security cameras, as well as any IEEE 802.3af
compliant end device; reduces the cost of additional electrical cabling and circuits that would
otherwise be necessary in IP phone and WLAN deployments (Models: JH019A, JH328A,
I also updated the OS with armbian and a linux kernel 5. Seems to be running smoothly now. I ll let you know how it goes.