HPE 9000 and HPE e3000 Servers
Showing results for 
Search instead for 
Did you mean: 

K580 HPMC - processor overheating

Geoff Gowers_1
Frequent Advisor

K580 HPMC - processor overheating

Mostly for info and comment:

We have been investigating a series of system crashes on a K580.

HPMC codes on chassis logs for respective processors:
0 5008 7d05 5408 5508
1 5108 7d05 7d03 5404
2 5208 7d05 5408 5508
3 5308 7d05 5408 5508
4 5608 5408 5508
5 5708 7d05 5408 5508

Our conclusion was that there were processor and processor memory bus checks happening on several cpu.Though initial 'what happened' report pointed at hpmc starting at cpu1, it was not clear to me that this was the culprit, or the only culprit.

On most crashes the system would restart successfully, but once, just cpu1 came up on its own.

Symptoms and codes very similar to a problem on this machine 8 months back where 2 cpus were replaced due to coolant leaks. Its ben good since then, and is in a cool environment, but has checked several times in the last 10 days.

On inspection, we found all 6 cpus are now leaking (cooling) oil to various degrees, and in all likelyhood randomly overheating and causing checks.

We are now looking for some good replacement cpus, and scheduling pd checks on our other K class.

The thread below also experiences this problem, and I am pretty sure I have seen this documented on an HP document when I was looking first time around.