Unsolved

1 Rookie

 • 

31 Posts

1511

May 4th, 2021 13:00

Poweredge R840, doesn't complete POST

R840 is out of warranty.

Server was rebooted and it didn't complete POST.  Power on, shows Service Tag, shows Booting message.  Power button goes off and then service tag and booting message repeat, to infinity.

Troubleshooting steps:

First thing was to disconnect both power cables and "flea" meaning press power button to clear things.  Also just left power cables unplugged for over a minute.

Pulled out one of the power supplies and tried exchanging power cables that were connected to a different power circuit.

No add in cards installed to begin with.

Removed top (daughterboard?) that contains two processors and their complement of memory.

On the motherboard, emoved all memory except for ones in slows A1 and B1.  No change in behaviour.

Swapped memory so that A1 and B1 have other memory installed that I got on the motherboard that were in different slots.  No change to how the server acts.

Disconnected cables from drive backplane/board: two SAS square cables and two cables at bottom.  I bet one of these is power, the other, I don't know.

Disconnected cable from back of DVD.  Still no change.

Interesting thing, I disconnected the CMOS battery and this time it didn't reboot over and over.  It complained about the battery being disconnected.  This time I had video but hung at the initial screen where it is initializing the life cycle controller.

Finally, I put things back together since all was good with the CMOS battery disconnected.

Following that thought, I tried the NVRAM reset jumper and reapplied power with battery reconnected.  Same, tired, story.

Last thing I did was put the daughterboard back in place (contains CPU's and RAM) C1, D1, etc.

This time, the power button goes on and off and repeats constantly.  When I disconnect the daughterboard it goes back to the original behaviour.  No pins are bent at the point where daughterboard fits into motherboard.

With removal of daughterboard, it goes back to how it was at the beginning of this thread.

I'm happy to leave out the daughterboard for now and I'll troubleshoot that as a secondary task.

Can someone here please let me know if there are further thoughts on how to troubleshoot?

Regards to everyone,

Brian

1 Rookie

 • 

31 Posts

May 4th, 2021 13:00

I also removed all memory to see if I'd get a beep code, but it doesn't seem to change anything.

4 Operator

 • 

3K Posts

May 4th, 2021 19:00

If you have access to iDRAC, Can you check "System Event Log", " Lifecycle Log" and Post Code to see whether it have any useful information on failure.
You can login to iDRAC GUI and see these information under "Maintenance" tab. Post Code will be available under "Troubleshooting" section.

1 Rookie

 • 

31 Posts

May 6th, 2021 13:00

I looked at the server again this morning.  My theory (based on what I was seeing when I initially looked at the server), is I could only get a partial boot up if battery was disconnected and one of the power supplies was pulled out from the chassis.  It worked this way one time and booted to the point of showing this on the screen:  Initializing Intel QuickPath Interconnect.  The cursor below is indicating where it was in the process, stopped under Init (first word) and would go no further.

At this point, I attempted to see what the IP address was for the DRAC and saw ...

I'm not surprised since I don't think the DRAC was ever set up.

Subsequent tries, the initial behaviour returned and no matter what I did (removed either power supply, battery in or out, removal of cpu/mem daughterboard) resulted in the same thing.  To recap, it would power on, show the service tag in the control panel and say it was Booting.  Then the light would go off of the power button and the process would repeat again.  Again I tried disconnecting power and pressed the power button for a few seconds.  No changes to how it is working.

 

0 events found

No Events found!

Top