I have built a new Truenas server with the following hardware:
ASRock Rack W680 WS
NEMIX RAM 64GB (2X32GB) DDR5 4800MHZ PC5-38400 2Rx8 1.1V 288-PIN ECC Unbuffered UDIMM
Intel i5-12600k
Solarflare SFC9020 (I’m waiting on an Intel X710-DA2 to arrive)
I Originally got it running then run memtest86 on it. I was getting ecc errors on one of the sticks. I got a replacement, and was able to run 4 passes with no issues. (Is it recommended to run more than 4 passes?)
I was running it for a few weeks alongside my old “production” server, testing out different scenarios with no issues.
So I decided to move my 5 HDDs from the prod server to the new one. It was running solid for a few days with no issues. In the past few days I have been seeing random reboots. I’m not seeing anything in the system logs (syslog/kernel.log), But I am seeing some errors in the BIOS Event logs. SMBios 0x01 and SMBios 0x02 errors, which seem to be Single bit ECC and MultiBit Ecc errors. the counts are 1 or 2 if that helps.
This is my first time working with ECC, so I’m trying to get a better idea o how it works, and want to validate some of the things that I have read.
Are the logs in the bios something that should crash the os? I thought that was the point of ECC? Are these indeed hardware errors, and I should be troubleshooting memory? I did read a thread on a similar motherboard and how the EDAC drivers were not in the current kernel, but it may just be for 13th gen chips, and its possible that they are already in there for my 12th gen chip?
Any help that someone could provide would be great, even some docs I can read to get a better understanding of ECC memory and how to troubleshoot it better.