HP DL160 G6. Blinking red light how to debug.

[webuser1200]

Hi,

I have 10 HP dl160 g6 1u servers. On one of them one of my processes dies after about a week and there is a red blinking health light on this server. This has happened routinely over the last few months.

I shut down the machine for 5 days and the red blinking light has gone, but I'm guessing will come back again.

Any idea on how I can debug what the red blinking health light (when it comes on) signifies?

Regards,

WU

 

[Patriot]

log into ILO and see what is wrong... there is a health log there...

 

[webuser1200]

I bought these over on ebay. Do you have to have an ilo license to log in. Any instructions on how to use ilo if you don't have a license?

 

[Diavuno]

MY DL G5 does the same thing, If you figure it out let me know.

Otherwise I just built a X8 system to replace it.

 

[Patriot]

I bought these over on ebay. Do you have to have an ilo license to log in. Any instructions on how to use ilo if you don't have a license?

No, just for after OS boot kvm.
The dns name, user and password should be on a sticker somewhere on the chassis as well as a tab on the front of the server that you can pull out.

Also, on the 10pin dip switch on the mobo, you can flip 1 to on to disable ILO security in case the password on the tab doesn't work.

 

[pricklypunter]

These are usually reliable as the day is long, but if your iLO doesn't point you in the right direction, check all the usual suspects.

Here's my usual quick check list for "ghosts in the machine":

Re-seat your processors, clean them and their heatsinks and apply fresh compound.
Re-seat all your memory.
Re-seat all your cards and cables, including any risers.
Give the whole thing a good clean out, removing any dust, paying particular attention to fans and cableforms and make sure no cableforms or anything else is restricting your airflow.

It could be an actual failing component at the root of the trouble, but until you rule out all the daft things it could be, you'll not know for sure if it was bad memory or just a dimm not quite seated properly etc

 

[Diavuno]

These are usually reliable as the day is long, but if your iLO doesn't point you in the right direction, check all the usual suspects.

Here's my usual quick check list for "ghosts in the machine":

Re-seat your processors, clean them and their heatsinks and apply fresh compound.
Re-seat all your memory.
Re-seat all your cards and cables, including any risers.
Give the whole thing a good clean out, removing any dust, paying particular attention to fans and cableforms and make sure no cableforms or anything else is restricting your airflow.

It could be an actual failing component at the root of the trouble, but until you rule out all the daft things it could be, you'll not know for sure if it was bad memory or just a dimm not quite seated properly etc

When my DL started to crash it was resolved by reseating ram.
eventually I pulled some of it out.

 

[alex1002]

I think it's bad ram. I noticed one of my hp servers at a client was doing this. It was actually ram. I pulled it out and blew out dust and seems fine now.

 

[pricklypunter]

Yup, chip creep was the main issue that I have come across several times, well that and overheating because of dust build up