ML350 G5 does not start

[Francois Lacharite]

We have a ML350 G5 on which the health LED lights up RED. The system does not boot up, it does not display any information, it will only power up and then constantly reboot on its own.

The mainboard LEDs are all normal with the exception of the OVER_TEMP LED which goes Amber when the server shuts down to reboot.

We have replaced the mainboard, the CPU and tried a different power supply, and none of these have had any effect on the problem. Without any more information from the LEDs and no access to the system since it does not boot, I'm stuck.

[Mynor Aguilar]

Have you replace the CPU along with it's heat sink? a while ago I had an issue with a ML350 G4 where it reboots by itselfs and sometimes we have a overtemp message at boot (health red led also). We replaced System Board, and CPU, however the problem was on a faulty heat sink. we ask for a replacement and everything was fine (I also had the OVER_TEMP LED) on the main board. However, as you know it was a G4 instead a G5

[James ~ Happy Dude]

Seems like a minor "Thermal grease" issue. Did it wear out ?
how about the PPMs ?

[Francois Lacharite]

It seems unlikely to me that it would be something as simple as a thermal paste issue or heatsink issue because the system never boots up. It doesn't matter if we leave the system off for hours or place it in a cool room, it never displays anything on screen and reboots every 10 sec.

When the problem first occured, we ordered a mainboard. While replacing it, I had to remove the processor and heasink, and reapply thermal paste on it. Same thing happens.

After replacing the mainboard, we ordered a CPU. The heatsink was not replaced, but once again the thermal paste was reapplied. No change.

Keep in mind every time I retested, we unplugged all cards, hard drives and removed the memory. I tested each RAM module individually and in pairs. I also tested the devices by plugging them each one at a time.

No matter what I've tried, the result is always the same. If it truly was a heat issue, I'd at least expect the delay before the server reboots to be affected by the ambiant temperature or the fact it's been offline for 12 hours. To me this seems like a hardware failure but we've already eliminated all the possible sources I could think of: mainboard, cpu, hard drives and cards.

[Francois Lacharite]

The problem has been resolved.

The source of the error was in the Drive Bay. Although I had removed all the HD, the power cable was still plugged into the bay and was causing the over_temp error to reboot the system.