Vintage Proliant 6000 meets earthquake

[p0rtmanteau]

Hi

I have a 6000 Proliant that has been doing its job faultlessly for the last 6 years, as a media server in an entertainment venue.

3 months ago it was in my building in Christchurch New Zealand which was damaged in a major earthquake that you may have read of in the news

Finally yesterday I was able to retrieve the server from my building, in the hope of retrieving years of media files from it

There was no physical damage to the server, but it will not POST for me :(

It has quad cpus and a management display

At turn-on all indicator leds seem to check out OK: all voltages, system interlocks etc seem to be correct, but one of the 2 redundant psus has no leds at all. It is possible that this psu failure could be historic. I do not have easy access to a spare PSU (all my spare parts including terminator boards etc are still trapped in my building which is inside civil defense cordons)

The 2 3100ES SCSI cards go through what appears to be a normal LED sequence, and then hang with 1 led only on the lower row of LEDS

IMD shows firmware version and no more

All HDD Leds flash at turn on and then stay off

There is no VGA signal

There are 4 amber LEDS on the processor board ... normal or not, I can't remember?

Two things occur to me

1: Flat battery from 3 months of no use?

2: Maybe because of the quad cpus the system requires 2 working PSUs, and will not POST with just one?

Could I be checking anything else?

Any help/advice much appreciated

I had waited all this time to access my media, and now find that the effort to retrieve the 6000 from a 1/2 collapsed building might all have been in vain. I hope not

P.S. This thread has been moved from ProLiant Servers (ML,DL,SL) to ProLiant Servers - Netservers - HP Forums Moderator

[Rob Buxton]

You're right the 6000 is pretty old.
Drop me an e-mail at rob dot buxton at wcc dot govt dot nz.
I may have a couple of ideas for you.

[p0rtmanteau]

done

[p0rtmanteau]

I just pulled the dead PSU apart, and it has taken a big hit on the primary ac input 10A fuse, which would be consistent with another building falling on top of my switchboard, and a city in general that was flying into a million pieces.

I'm leaning more towards the theory that the Proliant with a full load of CPUs does in fact need at least two PSUs?

Could anyone confirm that theory before I waste too much time tracking down a second good one please?

If I can just coax this old girl into life for long enough to salavage the content of 18 drives then I will give her an honourable burial!

[Johan Guldmyr]

Found the maintenance and service guide here:

http://h30499.www3.hp.com/t5/ProLiant-Servers-ML-DL-SL/Proliant-6000-Pentium-Pro-Manuals/m-p/3535353#M41339

There is a POST error "1612-Primary power supply failure".

Wouldn't you be able to access the Integrated Remote Console when the server is not powered on? Apparently you can access it both via modem and network :)

[p0rtmanteau]

Hi. Thanks for the response, but I am increasingly convinced that either processor board or processors are damaged

I think the clue to that conclusion lies in the 4 orange leds that are showing on the processor board. I can't find any mention of exactly what they mean, but they are labelled "processor init." so I'm guessing they should change state in some way as initialisation takes place, but in fact they show amber immediately at power-on and do not alter at all

I am moving on to try and salvage data direct from the mirrored RAID array, from individual discs. Hopefully in the next few days I will have some success with that

*fingers crossed*

[Johan Guldmyr]

might be worth a shot to try to run it with less CPUs

good luck

[PastaMan]

Hey,

Realise this post is cold now, but did you get it running to recover your data?

I just retired a proliant 6000 and if you still need parts could probably help out.

We're based in Melbourne.

Cheers