DL360p Gen8 will post but not boot + not reacting to keyboard

[y2keable]

Hi all

One of my HDDs has reportedly failed and since then, my server won't boot. It might just be a coincidence.

I've attaached a screen shot taken at the point the server hangs, it's reight between finishing POST and about to boot.

During POST, there's no response from the keyboard. I've mostly been using the HTML5 Integrated Remote Console but there's no response from a physical USB keyboard either. It may be that the USB ports aren't active as I can't get a numlock light to appear on the keyboard,

This problem exists whether ot not the HDDs are inserted and just in case it's a fault with the controller, I've tried booting with it unplugged.

I've tried disconnecting the power and the board battery but I'm pretty sure that only resets the RTC.

I've tried swapping ROM and I've tried resetting iLO

Nothing in the log pops out to me and no faults showing

I'm beet, any help would be appreciated

[TVVJ]

Hello,

Please clear the server's NVRAM and check if there is any change. You may refer to section "System board components" on page 12 of HPE ProLiant DL360p Gen8 Server User Guide for steps to identify the system maintenance switch. Refer to section "System maintenance switch" on page 14 of the user guide for steps to clear CMOS and NVRAM.

If not, then reduce the server to base configuration and check if it completes POST successfully.

Regards,

[BPSingh]

Greetings!

From the screenshot, I can see logical drive and multiple physical drive failures. 
Ideally, the drive failures should not cause the current issue that the server is exhibiting. 
You can clear the NVRAM and check for the issue. 
Also do check the server in minimum hardware configuration to further narrow down if the above doesn't resolve the issue. 
 

[y2keable]

Thanks Chaps

I started the server with switch 6 "on" and after POST, I got a message that said something along the lines of "Maintenance switch is on, you must power-off the server and reset switch"

I did that and it made no difference, server still hung after POST and didn't respond to any key presses.

With regards to booting with minimal hardware configuration, I've removed both drives and the SATA chache card, removed the NIC and although I still haeve both CPUs in their sockets, I only have a single stick of 8GB RAM installed.

Do I need to be more minimal?

[BPSingh]

Greetings!

Also please remove riser card if any. Test with one CPU at a time in CPU slot 1 with the single DIMM that you have. 

Same issue in minimum configuration could be system board related most of the times.
 

[y2keable]

@BPSingh 

Thanks but wenth through all of that and absolutely zero change.

If hardware (like a bad CPU), shouldn't I get an error in the log? What's frustrating me most is out of all the complexity and data logging, absolutely zero errors in the log >.<

[BPSingh]

Greetings!

If you have tested with both the CPUs with one CPU at a time then it should not be a CPU issue. 
You can try booting the server from the redundant ROM by turning on (S1, S5, and S6) on the maintenance switch 
if the server is still exhibiting the same issue after performing all the discussed steps, then I would suggest you log a support case for further investigation. 

[y2keable]

@BPSingh 

Yeah, a ROM swap was one of the first thing I tried..

Is it possible that I've just chaged something in the iLO interface like the server name or domain? Is it possible that the sytem clock is wrong? (system thinks it's Apr 2020)

I'm relly pulling my hair out on this how because I had a hypervisor installed on it and was running VMs, I just powered it down one day to transfer some RAM from it to another identical server which needed it most and when I started it up again, it just refused to boot and stopped responding to the keyboard.

I baught my servers as used (referbished), will HP still help me if I raise a  support case? Considering I don't pay for insight agent and the iLO4 key is just the same one came supplied with the servers, I'm guessing it ws purchased by the company who's owned them since new,.

[BPSingh]

Greetings!

iLO settings or system clock won't prevent the system from booting.

I believe you have already performed the basic troubleshooting.

Please contact HPE support to know the support policy for this product.

[y2keable · Accepted solution]

Ok, so just wanted to provide an update on this. I've solved it.

It seems rather silly, really but I'm rather new to servers and it wasn't obvious at the time but now it is.

I only had one stick of RAM installed and so one of the CPUs had no RAM.

I took out all the RAM but one stick as I wanted to pack as much RAM into my #1 server as possible. As this server was going to be my backup server,I fugured I only need a minimal amount of RAM in order to configure it. Of course, I was wrong. I installed a second stick and it booted no problem.

Confusing though, no errors reported in ILO and it POSTed just fine. Surely, if both CPUs need RAM then POST would have failed or there'd be errors in the log or just errors in general. Nothing and because of that, I don't feel like a complete idiot

[Sunitha_Mod]

Hello @y2keable,

That's great!

We are glad to know that you were able to find the solution and we appreciate you for keeping us updated.