Dears, appealing to this forum wisdom to help me recover my loved Debian. Please be nice with me on any mistakes below, I'm just an amateur Debian lover.
A little about the background of the problem. My home server (Supermicro X11SSM-F, Intel Xeon E3, HD SATA 2TB, 16GB, integrated ) has been serving us well since I bought it in 2015 (likely relevant, old hardware). I put Debian 8 at that time, and over the years upgraded it to Debian 10 which has been operational until a few weeks. Following some recommendations I read about, decided to get to 12 in two waves, first into 11 then from 11 to 12. The update to 11 didn't go well as I started to get some unexpected reboots. As the server has gone through a lot over these 9 years I decided it was time to give it some refresh so backed things up and went for a fresh install of 12. Install went fine until it finished, went for the first reboot and it didn't complete and the machine rebooted unexpectedly in the middle of the process. Starting it in rescue mode didn't help either, the boot process apparently went further on but before presenting the UI to the user again an unexpected reboot.
That is still where I am today.
Things tried already in chronological order.
+ Gave a gentle cleaning with an air spray and a brush, to remove dust and dead bugs from the inside.
+ Confirmed cables in place and memory modules properly inserted in the slots.
+ Run the live version of 12 directly from USB, same crash somewhere after the welcome to grub message
+ Run the live version of 12 directly from USB selecting the rescue option. It works. This is the only working option at the moment.
+ After learning about how to remove the quiet flag when booting, mount the HD to access the boot logs and how to read the journal I scanned through it not finding any clear hint of what is failing. It helped me though to identify the last line that was logged before the crash. Here's one from this morning:
Jan 13 12:39:14 lightH systemd-journald[340]: Time spent on flushing to /var/log/journal/e77c844966b145c1b1ca7c2fa7f8ab4a is 9.296ms for 928 entries
+ Motivated by some lines in the journal stating "[Firmware Bug]" I went to Supermicro to check for a BIOS update. There was one. Installed it careful not to lose the actual BIOS settings. No change, same problem, same place.
+ Possible hardware problem right? That's what I thought. Replaced HD as I have a spare new one, same problem.
Memory issue? run memtest86 for over a night, no errors detected. Still to be sure, removed both memory dimms and let the server run alternative with only one (8GB) and then the other. The problem repro in both cases in the same place.
+ Overheating? through the BMC I have access to the temperature sensors, all green.
+ Started to think about the power supply, but then decided to flash an old iso of Debian 10 and installed it just to test. It worked fine.
At that point I decided to stop considering a HW problem, and potentially might stay at Debian 10, but still would love to understand what is breaking my system if using Debian 12.
Other things I have tried out of desperation in a try-error fashion as read in the forums they helped other users:
+ nomodeset parameter
+ Secure boot, whatever this might be (?) it is disabled in the BIOS.
+ Confirmed SATA mode selection is AHCI (whatever this means)
+ Disable VT-d. This one I think I'll need it, remember enabling it when I started using VirtualBox, anyhow after disabled it, no difference.
There might be other items tried already that I'm forgetting at this moment, none have worked till now.
As mentioned at the beginning I'm just an amateur user so there might be tools or settings I havent read about that are behind the problem.
I'll appreciate any hint on what to check next or any further tools or setting that might be relevant.
Thanks in advance.
AN
A little about the background of the problem. My home server (Supermicro X11SSM-F, Intel Xeon E3, HD SATA 2TB, 16GB, integrated ) has been serving us well since I bought it in 2015 (likely relevant, old hardware). I put Debian 8 at that time, and over the years upgraded it to Debian 10 which has been operational until a few weeks. Following some recommendations I read about, decided to get to 12 in two waves, first into 11 then from 11 to 12. The update to 11 didn't go well as I started to get some unexpected reboots. As the server has gone through a lot over these 9 years I decided it was time to give it some refresh so backed things up and went for a fresh install of 12. Install went fine until it finished, went for the first reboot and it didn't complete and the machine rebooted unexpectedly in the middle of the process. Starting it in rescue mode didn't help either, the boot process apparently went further on but before presenting the UI to the user again an unexpected reboot.
That is still where I am today.
Things tried already in chronological order.
+ Gave a gentle cleaning with an air spray and a brush, to remove dust and dead bugs from the inside.
+ Confirmed cables in place and memory modules properly inserted in the slots.
+ Run the live version of 12 directly from USB, same crash somewhere after the welcome to grub message
+ Run the live version of 12 directly from USB selecting the rescue option. It works. This is the only working option at the moment.
+ After learning about how to remove the quiet flag when booting, mount the HD to access the boot logs and how to read the journal I scanned through it not finding any clear hint of what is failing. It helped me though to identify the last line that was logged before the crash. Here's one from this morning:
Jan 13 12:39:14 lightH systemd-journald[340]: Time spent on flushing to /var/log/journal/e77c844966b145c1b1ca7c2fa7f8ab4a is 9.296ms for 928 entries
+ Motivated by some lines in the journal stating "[Firmware Bug]" I went to Supermicro to check for a BIOS update. There was one. Installed it careful not to lose the actual BIOS settings. No change, same problem, same place.
+ Possible hardware problem right? That's what I thought. Replaced HD as I have a spare new one, same problem.
Memory issue? run memtest86 for over a night, no errors detected. Still to be sure, removed both memory dimms and let the server run alternative with only one (8GB) and then the other. The problem repro in both cases in the same place.
+ Overheating? through the BMC I have access to the temperature sensors, all green.
+ Started to think about the power supply, but then decided to flash an old iso of Debian 10 and installed it just to test. It worked fine.
At that point I decided to stop considering a HW problem, and potentially might stay at Debian 10, but still would love to understand what is breaking my system if using Debian 12.
Other things I have tried out of desperation in a try-error fashion as read in the forums they helped other users:
+ nomodeset parameter
+ Secure boot, whatever this might be (?) it is disabled in the BIOS.
+ Confirmed SATA mode selection is AHCI (whatever this means)
+ Disable VT-d. This one I think I'll need it, remember enabling it when I started using VirtualBox, anyhow after disabled it, no difference.
There might be other items tried already that I'm forgetting at this moment, none have worked till now.
As mentioned at the beginning I'm just an amateur user so there might be tools or settings I havent read about that are behind the problem.
I'll appreciate any hint on what to check next or any further tools or setting that might be relevant.
Thanks in advance.
AN
Statistics: Posted by GetMyDebianBack — 2025-01-13 17:04 — Replies 3 — Views 55