Proliant DL380 G4 No Post, no Video...but already successfully booted 2-3 times...

[Jbruneaux]

Hi,

I've a HP Proliant DL380 G4 Server which do not pass POST. It stops on a 0x61 error (Port 85 error code) during post. This code seems related to the 'Expansion boards' and the troubleshooting manual tells us to replace the PCI Riser board. I did replaced it but it still doesn't solve the problem...

I've been able to successfully boot the server 2 or 3 times and it worked fine (booted from a live CD to make tests). Unfortunately, after having played a bit around with disks and other things, it now won't boot.

I can't really tell what made it boot as I was trying to make it work for hours, so I might have hotplugged HD or anything else, and voila...it booted.

I run out of idea on what's the exact problem. During the boot sequence, I can see the 'port 85' leds going through the 0x60 code, then a few other codes and then is stuck to the 0x61h (just before that, it do the 0x51h code). I've already removed all components, cleaned everything but it doesn't help. I'm pretty sure the 0x60h is a kind of presence check and 0x61h is an initialisation step, but maybe I'm wrong).

I can't connect to iLo (don't know why, I can't configure it because of no video, and it doesn't seems to show up on the dhcp....even if the lights are OK on the ethernet port).

Does someone have any more precise meaning of the port 85 leds (especially the 0x61h code).

Regards, Jerome

[PGTRI]

hi,

It seems that you need to replace the systemboard. Please check, if the server is still under warranty and open a ticket by HP.

Please be sure , that the riser cage is correctly installed, because the server cannot boot w/o it.

regards,

[Jbruneaux]

Hi,

The server is no more under warranty so I was trying to repair it by myself. I was just wondering why the server has been able to boot 2 or 3 times (consecutive boots with power cord removed, not just reboot).

As i'm working in electronics, I try to find what the real problem (at first, I followed the guide and changed the riser board, but it wasn't the faulty part because it didn't changed anything).

I was guessing if anyone at HP has a more detailled description of the bios post codes because the manual is quite basic (all 6xh codes means changing the board or the system board, as well as 5xh codes and others...not really usefull except if you have some spares to try).

Regards, Jerome

[rguha]

Hi,

Try clearing the NVRAM and check:

Steps to clear NVRAM:

-Power off the server.
-Locate the maintenace switch on the system board
-Switch the S6 switch to ON position. By default it will be OFF.
-Power on the server. Wait for a min.
-Power off server.
-Toggle the S6 switch back to off position and restart server.

To locate the server's System Maintenance Switch, use the server's Maintenance and Service Guide. Page 96

http://h20000.www2.hp.com/bc/docs/support/SupportManual/c00687567/c00687567.pdf

Also, try to boot in minimum config and check:

1. Shutdown the server and remove all the PCI cards manually.

2. Remove all the memory modules, except for one pair of DIMM (Dual Inline Memory Module).

3. If the server has two processors installed, please remove the second processor.

4. Reseat the existing processor and Processor Power Module.

NOTE: Please reseat the processor carefully as it may bent the pins in the system board.

5. Remove the power connectors from the hard drives, tape drive (if any), floppy drive and CD ROM drive.

6. Boot the server and check if it powers on. Try to remove and reseat the power supply unit as well.

7. If the server powers on, reinstall all the components, one by one, until the faulty one is recognized.

8. If the server works fine, then it is a reseating issue with the components.

Note: Do not uninstall any configured Smart Array controllers.

[Jbruneaux]

Hi,

Thanks for these steps but unfortunately I've already tried them.

I've removed everything from the server except the minimal requirement and still not booting.

I'm running out of idea and don't wan't to buy lots of parts just to find that it was only one of them which was faulty. Plus, as it already booted a few times, I'm sure that everything can work, but there must be a little buggy thing...

If an engineer or technician of HP could have some debug info (a hidden switch to make bios output some messages over a serial, special tricks to override some steps, ....), it would really help me find out the problem.

Jerome

[Jbruneaux]

Hi,

This weekend, I managed to get the server to boot again. It seems to be power-supply related as It worked until I removed the power plug. I could make it reboot a lot of times, made the firmware updates from HP Firmware CD, installed a linux OS on it. The only problem was that iLO was crashed (didn't even showed up at boot whereas it did the first time I had the server booted). The iLO leds where showing and error code (something like 0xB3h and something else alternativeley displayed on leds).

After having all the OS set up, I tried to remove the power to allow iLO to reset and now it doesn't boot anymore.

I tell it seems to be the power supply because when the system is plugged but not powered, I can ear a clicking noise (like an electric arc click you can get in some motors brushes). I've opened up the power supply and tried to locate the clicking noise (might come from a relay in the middle of the power supply) but would be too complicated to replace.

I need to check the power supply with an oscilloscope but it seems the clicking noise (well, its origin) makes some spikes on the power supply (from the measure I could do with a voltmeter on the +12V.

I wish I had an other power supply to swap-test it...If the spikes are present, I may buy a spare power supply. I can't believe the motherboard is faulty, it would have made the system crashing during the installation or bootup of the OS (almost all peripherals are probed / tested during the boot).

[untitle05]

Thans you  a lot. You save me.