1. OCAU Merchandise is available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion in this thread.
    Dismiss Notice

Server died, any thoughts?

Discussion in 'Business & Enterprise Computing' started by Caffeine, Nov 4, 2025.

  1. Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    upload_2025-11-4_9-58-59.png

    upload_2025-11-4_9-56-0.png

    Woke up this morning to see my DL380 G9 dead, no lights or anything, no iLO.

    Removed and replaced all the power cables, it flashes all 4 lights and iLO is back, but shows the error message above.

    Replaced the button cell battery, (was 2.9V) with a new one at 3.2V, no change.

    Tried resetting the NVRAM using the system mainteance switches, still no luck.

    Anyone have any experience with this? :upset:
     
  2. fad

    fad Member

    Joined:
    Jun 26, 2001
    Messages:
    2,786
    Location:
    City, Canberra, Australia
    It from the fault codes looks like it popped a fuse
     
  3. heydonms

    heydonms Member

    Joined:
    Sep 15, 2008
    Messages:
    798
    Location:
    6112
    I don't think that's a fuse in the traditional 'melting wire' sense, it's an IC that monitors current, voltage, etc. and shuts things off if something goes out of range.

    Best starting point is probably too strip the machine back to the bare bones (no expansion cards, single PSU, single CPU, minimum ram sticks, disconnect the backplane, etc.) and see if it still complains
     
    JSmithDTV and Caffeine like this.
  4. OP
    OP
    Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    Yep, tried with no drives, 1 CPU (swapped both CPUs to slot 1 individually) and minimal RAM, still no improvement :(

    Wondering if the lightning last night in Sydney killed it, but it's behind a good rackmount eaton UPS so it shouldn't have.

    I do have 1 PCI card in there, an interface ot the two SSD boot SSDs, I'll see if I can get it to boot without that
     
  5. Sphinx

    Sphinx Member

    Joined:
    Sep 16, 2001
    Messages:
    11,764
    Location:
    Brisbane
    Sounds like an onboard System/Mainboard component failure, no ILO doesn't help either.
    I'm guessing she has a few years under her belt?
     
    JSmithDTV likes this.
  6. JSmithDTV

    JSmithDTV Member

    Joined:
    Jun 13, 2018
    Messages:
    13,922
    Location:
    Algol, Perseus
    Yeah EFUSE is a power regulation circuit... may have tripped due to a short circuit, failed component, overcurrent/voltage. Going to be difficult to isolate

    Any option to swap power supplies? Better inspect the board for visible damage...



    JSmith
     
  7. thecondor

    thecondor Member

    Joined:
    Jun 30, 2011
    Messages:
    2,928
    Tried turning it off and on again?
     
    SLIMaxPower likes this.
  8. OP
    OP
    Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    It has dual PSUs. I've tried each individually.

    All the reading seems to be related to power to the PCI socket, because mostly people run into this problem when trying to install a beefy GPU. I have none of that.

    I'm going to pull it out of the rack properly this afternoon and see if I can see any damage
     
    JSmithDTV likes this.
  9. heydonms

    heydonms Member

    Joined:
    Sep 15, 2008
    Messages:
    798
    Location:
    6112
    What about the backplane? There is quite a bit of power management stuff on those to handle hot swaps and such.

    That seems plausible. Surge suppressors can only do so much, and unless you have a double conversion UPS you're feeding whatever comes in on the mains wiring into your system until it has time to recognise a fault and respond to it.
     
    Caffeine and JSmithDTV like this.
  10. OP
    OP
    Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    With the backplane and PCI disconnected it will POST!

    Now to start plugging bit back in to hopefully isolate which one is causing it
     
  11. OP
    OP
    Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    Backplane it is it seems :thumbup:
     
  12. PabloEscobar

    PabloEscobar Member

    Joined:
    Jan 28, 2008
    Messages:
    14,717
    Lodge warranty fault, await replacement.
     
  13. OP
    OP
    Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    Well out of warranty unfortunately.

    Found that there are two sides to the backplane board, each with their own power connector.

    Side '2' with drives 7-12 works fine, but there is a dead short between GND and 12V on side '1' (drives 1-6)
     
  14. PabloEscobar

    PabloEscobar Member

    Joined:
    Jan 28, 2008
    Messages:
    14,717
    That's not very business or enterprisey of you :).
     
  15. Current

    Current Member

    Joined:
    Aug 10, 2021
    Messages:
    3,413
    probably stupid question but was there any magic smoke ? whether you can smell it now, or a cctv camera saw it ?
     
  16. OP
    OP
    Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    Nope, the efuse shut it all down before any magic smoke got released, which is great for not starting a fire, but makes diagnosis more difficult!
     
  17. OP
    OP
    Caffeine

    Caffeine Member

    Joined:
    Jul 1, 2003
    Messages:
    2,480
    Location:
    Sydney
    After a bit of fault finding and repair, I now have my server up and running again!

    I do have a dead hard drive now though, but I'm not sure if that's correlated.
     
    Phido likes this.

Share This Page

Advertisement: