6

I have a problem with an HP ProLiant DL380 G7 that leaves me without much in the way of troubleshooting approaches.

This server will not boot. It was stored in a datacenter, and I went to retrieve it after I could not connect and the colo tech couldn't hard reboot it. It would basically flash all HDD and system health lights, and I unplugged the system after I could not access it or restart it by pressing and holding the power button. Since then, it does nothing and just sits with the Power LED solid orange, and no activity on the health lights. The fans don't spin, the power supplies don't light up, nothing happens.

I unplugged all of the drives and just tried to boot with the CPU and memory, but still nothing. The ethernet link lights do light up, though. The system has dual processors and two power supplies. It seems like a motherboard error but I can't tell.

One person mentioned the PSU back planes fail sometimes, and that makes sense, but plugging in only one power supply and checking every configuration possibility produces the same result.

The server is out of warranty and I need to find which part to replace. Is there anyway to continue troubleshooting it? The system has iLO installed but I don't know how to access it without being able to boot the machine to BIOS.

ewwhite
  • 194,921
  • 91
  • 434
  • 799
user233321
  • 61
  • 1
  • 1
  • 2

6 Answers6

8

Two things...

  • You can add a warranty to an HP server whose factory warranty lapsed. It's not very expensive to add another year of support to the machine you have now.

  • You can also initiate a per-incident support call/technician with HP and be liable for parts charges.

Depends on how desperate you are...

But here's a (lengthy) troubleshooting procedure to try first if you have access to the machine.

  • Remove the power supply units and swap them.

Test to see if the system will boot...

If that doesn't work:

  • Remove all power supplies from the chassis
  • Locate the System Maintenance Switch on the motherboard - It's a set of 10 DIP switches.
  • Turn switch #6 switch on.
  • Insert all power supply units.
  • Power on the server and allow it to idle for 3 minutes.
  • Power the server off.
  • Remove all power supplies.
  • Return DIP switch #6 to off (original) position.
  • Reinsert the power supplies.
  • Power the server on.

Test to see if the system will boot...

If that doesn't work:

  • Remove all power supplies from the chassis
  • Turn DIP switches #1, 5 and 6 switch on.
  • Insert all power supply units.
  • Power on the server and allow it to idle for 3 minutes.
  • Power the server off.
  • Remove all power supplies.
  • Return DIP switches #1, 5 and 6 to off (original) position.
  • Reinsert the power supplies.
  • Power the server on.

Test to see if the system will boot...

ewwhite
  • 194,921
  • 91
  • 434
  • 799
  • 2
    Many thanks, this has saved us after a power cut killed the server. (The last steps that is). Can someone explain what this actually did? – user2924019 Jun 03 '15 at 10:24
  • 1
    Oh, it's *magic*. – ewwhite Jun 03 '15 at 12:36
  • Changing the battery seems to have fixed it for us – mwfearnley Nov 24 '17 at 15:11
  • @mwfearnley Are you referring to the CMOS motherboard battery of the RAID cache battery? I assume you mean CMOS but wanted to verify as nothing seems to work for me beside taking out the CMOS battery and both PSUs, waiting a while, reconnecting it and then powering it on and redoing the iLo and CMOS configs for the options and such we have set. A few days later, the same thing happens again and of course a warranty from HP costs more than this company paid for the refurbished server a couple years back. – Pimp Juice IT Oct 11 '18 at 15:39
3

If all of the above did not help, like it was for me, unplug all PSUs, remove CMOS battery and check its voltage level. If it is below 2.9V when removed - trash it and get a new one.

After being hit by this issue we've did an extensive research using three servers available to us in a lab, four PSUs and one bad CMOS battery. To make a long story short, whenever power is unplugged, iLO logs an event and writes it to NVRAM. If your battery is no good, NVRAM gets corrupted and this orange led situation is one possible outcome, you may also see red health light with some nonexistent failures indicated, or even "fan solution not sufficient".

There's also an HPE advisory that does not mention the battery though, but shades some light on the root cause of this.

Peter Zhabin
  • 2,276
  • 8
  • 10
2

Same issue here with HP Proliant DL380.

Cracked case--removed Power Supply tested for continuity--test revealed circuit is open.

Replaced Power Supply from like (known working unit) and the server powered back up and returned to normal operational state.

If you have the AMBER light and the flashing green light on the back of the power supply--test it with a meter, but i am pretty sure the problem is a failed power supply.

New power supply on order. $40 fix.

Good luck!

Gregory A. Levine
  • 21
  • 1
0

Had the same issue on hp dl 160, solution was to remove one processor after power faults and it worked

Emmak
  • 1
0

In my case, the problem was the expansion board that host additional FC card. After we removed it, the server started without any problems. We are trying all the above and in the end it was a solution

Snakebl
  • 1
0

Seems to work always. The issue is related to some internal circuit that needs to be feeded before detects the PSUs

Put cabling and psu, wait 3 min. Unplug cabling and psu for few seconds. Plug 1 PSU outside and insert (small effort). The PSU will be ready to feed.

You get the power light in orange. Press power Plug 2nd PSU and cabling normally.

user508416
  • 1