topic Re: DL385G5 start-up problem after reboot in ProLiant Servers (ML,DL,SL)

DL385G5 start-up problem after reboot

AutomatiseringBR — Mon, 20 Dec 2010 09:31:38 GMT

We have multiple DL385G5 servers.
They all run Esx or Esxi.

One of them (last friday) displayed a memory error on the frontpanel.
- 1. iLO check: nothing visible (strange)
- 2. shutdown of all VM's eand initiated a reboot

Then the problems began. After the initiated reboot the screen kept beeing blank. You could hear the fans spinning up to max but no more.
No POST, no display whatsoever. Just a blank screen.

I did several cold boots via iLO: nothing.
I tried resets via iLO: nothing.
I disconnected both power supplies for 5 mins and... still nothing.
The i again disconnected power and dialed our vendor for support. 15 mins later i would give it a final try and.... YES the server completed it's POST and booted.

Now, after a remote reboot the server will hang before the POST screen. It looks like it needs a lot of no-power time before it will boot properly again.

I did an extensively search on the internet but found little to nothing of relevance.

Has anyone ever had this issue?
So yes, what is the/a solution?

Re: DL385G5 start-up problem after reboot

Johan Guldmyr — Mon, 20 Dec 2010 09:47:59 GMT

Hey,

strange stuff!

Firmwares are they up to date? I don't know the dl385g5 firmwares by heart, maybe there is something that helps with this?

If it freezes before POST, maybe you do have a DIMM or CPU issue?

If you get it to boot, maybe you can run some diagnostics on it?

Are there no error leds lit? On system board maybe?

Re: DL385G5 start-up problem after reboot

Suman_1978 — Mon, 20 Dec 2010 09:56:15 GMT

Not seen this kind of issue.

If the server is booting fine after removing and connecting the power supply, I would imagine its something to do with Power supply or Power supply backplane.

Try swapping power supply from a similar server and test.

Re: DL385G5 start-up problem after reboot

AutomatiseringBR — Mon, 20 Dec 2010 10:08:24 GMT

Bios version is: A09 05/29/2008; backup system ROM: 05/29/2008.

I know there are newer versions. The revision history states no (documented) improvements for this problem.
And all our DL385G5's now are identical (Esx-cluster)
We like to keep it that way, at least for now.

After the 15 mins of no-power, the memory error on the frontpanel was no longer lit.

Testing is a bit difficult. I can't affort systems to go down. We have a mayor upgrade plannend early next year (Esx 3.5 > 4.0).
Then all the firmwares will be upgraded.

Re: DL385G5 start-up problem after reboot

Johan Guldmyr — Mon, 20 Dec 2010 10:53:06 GMT

Understand.

It is a bit difficult to test memory issues inside an OS, best done outside from a boot cd.

But maybe somebody else will read this who has had the same issue.

As it happens before post it shouldn't be caused by an extra-card like HBA/controller.
But might be worth checking out the release notes for later firmwares there as well.