Stuck at post code 15

[Mike77]

Hi all,

Since I've moved my server, it's stuck at 'system initializing' with post code 15 ('Pre-memory North Bridge initialization is started').

The server used to run without issues. We moved the system to another room and after that it wouldn't properly boot anymore. Steps already taken:

- unplugging drives
- reseating all memory - still stuck at 15
- checking psu and cables - they work
- testing memory in other system - they work
- booting with memory in just one slot ( with different DIMMs and different slots) - still stuck at 15

Does anyone know what might be wrong and which steps In should take?

Thank you for your answer!

 

[elvisimprsntr]

I assume that means the server is on a new power source.
A quick check might be to move it back to original location.
1. Different UPS and/or different loads on same UPS?
2. Is there highly inductive or noisy loads on same circuit?

Another possibility is if any of the passive or active heat sinks were jostled during the move.

Is the ambient temperature in the new location higher?

 

[Mike77]

I assume that means the server is on a new power source.
A quick check might be to move it back to original location.
1. Different UPS and/or different loads on same UPS?
2. Is there highly inductive or noisy loads on same circuit?

Another possibility is if any of the passive or active heat sinks were jostled during the move.

Is the ambient temperature in the new location higher?

Thank you for your response. We moved it back, no improvement unfortunately.

We checked for components that came loose, but found none. Are there other issues that might occur with the heat sinks?

We moved the server within the same house, and both temperature and power supply/circuit are approximately the same.

 

[Ericloewe]

Is this the X10SL7-F?

 

[Mike77]

Is this the X10SL7-F?

Yes it is!

 

[Ericloewe]

Does the IPMI log show anything relevant?

 

[Mike77]

Does the IPMI log show anything relevant?

The IPMI log only shows fan1. The FRU reading is empty, but I read online that this is more often the case?!

Since my first massage we checked for bent pins (memory slots and cpu socket, Both back and front), reseated the CPU, did a CMOS reset and a reset by removing the battery. We took the motherboard from the chassis/enclosure and disconnected alle components not needed while trouble shooting. We also connected a different power supply. Nothing changed: still stuck at 15!

Perhaps relevant: a few weeks ago a heat sink came loose from the SATA controller. We glued it back on. The system functioned as it should after this incident.

Any ideas are very welcome!!! Thank you in advance!

 

[elvisimprsntr]

Perhaps relevant: a few weeks ago a heat sink came loose from the SATA controller. We glued it back on.

Glue? Hopefully you mean thermally conductive adhesive.

 

[Mike77]

Yes it was thermal glue.

Here's the link regarding the heatsink:

Help drives zfs pool disappeared

Hi, My entire zfs pool suddenly disappeared and I found a small gold coloured heat sink on the bottom insider my server. Ik have no Idea how to troubleshoot this problem. Importing the pool doesn't work anymore and listing the zpool gives me nothing. FreeNAS-9.10.2-U5 (561f0d7a1) Supermicro...

www.truenas.com

 

[Mike77]

I asked Supermicro. They say het MB is EoL and support limited. According to them the post code is related to memory, so I should try other memory.

I don't have memory from the Supermicro tested memory list available. I did try each DIMM individually in Dimm slot A: still code 15. I also tried the memory in a different system (MB asus p8z77-vlx, Processor i7 3700): all modules work in this system. I took the memory from this system (CORSAIR Vengeance LP 16GB (2 x 8GB) DDR3 SDRAM 1600 - CML16GX3M2A1600C10) and tried it in the failing server. I tried 4 corsair modules individually (in different slots) and combined. Still code 15. I even took memory (DDR3 SDRAM 2GB 1333MT/s 240-UDIMM- MT16JTF25664AZ-1G4G1) from a third computer and tried this. Again, code 15. No memory at all: code 15.

Could someone advise me: Should I buy new memory to test with (considering that this is a fairly old machine) or is it save to conclude this is not a problem with the memory?

I'm looking for a way to reinstall all firmware for the MB (without additional costs). And i was wondering: when checking the log, as Ericloewe suggested, I noticed there are no readings of the sensors. I wonder why. Anyone any ideas in these points?

 

[NugentS]

It does look like the motherboard at this point

 

[Ericloewe]

there are no readings of the sensors. I wonder why

Most sensors are offline when the system is not running. The boot process probably did not get to the part where it sets up the sensors, DRAM initialization is a very early step in the typical boot process, since working only within registers is a massive pain in the ass.

 

[Mike77]

Thank you for your answers! Supermicro sent me a new email, advising to try a bios recovery. I'm gonna try that by setting jumper jbr1 to 2-3. As the manual indicates, I'm copying the super.rom file from the Supermicro website to a USB flash drive. I'm using a Kingston 16 gB, FAT32 formatted, non bootable USB flash drive (the manual gave no specifics, so I'm guessing any flash drive/formatting will do). Unfortunately, there are no pins on jumper jbr1, so I guess I'll have to short it with a copper wire.

If this is not the right way to do it, please tell me!

I'll post the results when I've tried this! [EDIT] And it didn't work. I'm not even sure it did anything. Still no beeps and still stick at post code 15. I'm wondering how this should work, seeing that it needs a working USB port..

 

[jgreco]

The X10SL7-F should be new enough that you should be able to update the firmware via the IPMI/BMC interface.

 

[Ericloewe]

Yeah. It would need an unlock code, but it's something trivial like the first digits of a hex-encoded SHA1 of the IPMI MAC. A quick search will set you on the right path.

 

[jgreco]

Yeah. It would need an unlock code, but it's something trivial like the first digits of a hex-encoded SHA1 of the IPMI MAC. A quick search will set you on the right path.

Just to be clear, you're searching for "Supermicro Out-of-Band Software License Key (SFT-OOB-LIC)" which usually sells for thirty dollars, and not the blog article out there that talks about "reverse engineering Supermicro IPMI" and what the exact algorithm used to generate the key is.

 

[Redcoat]

There's a discussion here https://forums.servethehome.com/ind...0sl7-f-now-stuck-in-infinite-boot-loop.33549/ that might help you select a BIOS version,

 

[Mike77]

Thank you all! Supermicro also advises to use the SFT-OOB-LIC.

I've looked at the reverse engineering blog/ Sha1 encoding. I couldn't get it to work, so I guess I'm going to have to buy.

Before spending money on the key: I'm wondering why it's likely updating the bios will work, considering the early step the system is stuck at. If I understand correctly, cpu and memory aren't fully initialized yet. Is a separate/ redundant system used? I guess I'm missing some information...

 

[Mike77]

In case I'm still stuck at 15, after flashing the bios: is there anything left I can try (other than replacing the hard ware)?

 

[Mike77]

At the moment I've started testing all contacs in the memory slots with a multimeter...