DL360 G5 turns itself off and PPM light is amber

[Simon Haslam]

I have a DL360 G5 with 2 E5405s, 32GB RAM, 1 PSU and 6 SFF SAS disks - its warranty expired last August.

A week ago the server powered off - the iLO log just said "Server power removed.". I didn't think to check the slide-away Systems Insight Display and so powered it back on - it was fine for about 5 hours then shut down. I took the PSU out and it was very hot and let it cool down.

Anyway, I have had the server powered on a couple of times since and when it just turned itself off I see on the SI Display that the PPM indicator is amber and the system fault indicator is flashing red. Note the Power Supply 1 failure light is not on.

Looking at the DL360 G5 Maintenance and Service Guide page 78 it looks like the PPM is integrated on the motherboard (unlike the DL380 G5 - see this DL380 G5 discussion ).

Note I have been checking iLO temps and when it fails the biggest difference is that the "I/O Board Zone" was showing 50C or 47C, whereas normally it is around 31C (with ~18C ambient). All the other temperatures, including Power Supply Zone, were around 30. All disks were working normally though it has been complaining about battery charge low on the P400i.

When I remove and reapply the power to the server the PPM indicator goes off.

Does anyone know the most likely failure here? I could buy a new HP replacement power supply (which is what I first assumed was wrong) if it has a good chance of fixing the problem (e.g. if it was actually the power supply which somehow triggered PPM error) but I don't want to waste my money if it's a fault on the motherboard (and so unlikely to be economically repairable).

Simon

[Goutham_Sabala]

G5 servers had quite a issues with power supplies , I would suggest to check with a spare psu and a psu backplane, board may not be the cause, Im suggesting psu as you had mentioned it is getting hot and once reseated servers works for few hours , to me power backplane and psu looks like the most likey cause

[Simon Haslam]

Thanks for your thoughts Goutham. The funny thing is the power supply was only hot the first time it shut down unexpectedly - the last time I didn't remove it but it didn't feel hot through the top of the (opened) charris. It's a shame there's nothing very helpful in the iLO log.

If it's the power backplane at fault is it worth trying the PS in the other slot (or does that not work - slot 1 always has to be populated)?

Unfortunately I don't have spares for either so would have to buy them (about $150 each for a PS or a system board, haven't looked for power backplane) - whilst I don't mind doing that if it fixes the problem I'd rather not have to buy both, especially if not confident it will work! (I'll replace the server once the E5s are a bit more affordable).

Anyway, I've got the server running again and will wait for the next failure and report back if the PPM or PS light is on etc.

[Simon Haslam]

Well, curiously the server has run without any problems since my last post. I left it for a few days, then used the Proliant Support Pack to update the BIOS, iLO firmware etc (hadn't done that for over a year). The iLO is currently showing I/O Board Zone at 43C and Power Supply Zone at 28C for an Ambient Zone reading of 17C - although the I/O Zone temperature is higher than I posted earlier the server has been running continuously again and these values seem fairly steady.

I don't imagine the problem has fixed itself but maybe it will last a few more months until it is replaced.

[Goutham_Sabala]

Im glad the server is under normal operation without having to replace any parts, it is best not to restart or change the hardware unless until necessary :) . Best of Luck