Hardware Errors? CPU or ECC-RAM?

[unJack]

Hi Guys,

 

after setting up my new hardware with Unraid the last days, I'm pretty happy with everything!

But today i discovered by accident that there are a lot of hardware errors reported in the syslog (btw, can those be reported via Notifications?).

 

I'm concerned as I plan to store data on unRAID which must not be corrupted. Can somebody derive a possible root cause from those log messages? Is it a faulty CPU, MB or the ECC-RAM?  

 

 

Diagnostics: unraid-diagnostics-20210707-0906.zip

 

Hardware:

- be quiet! Pure Power 11 400W (80Plus Gold)

- AORUS B550 AORUS ELITE V2, latest BIOS

- AMD Ryzen™ 5 3600 (6c/12t)

- MSI GeForce GT 710 1GD3H LP

- 32GB ECC RAM (2x Mushkin DIMM 16 GB DDR4-2133 ECC)

 

- M.2 NVMe 1TB (Apacer AS2280P4 1 TB, PCIe 3.0 x4, NVMe 1.3, M.2 2280)

- M.2 NVMe 1TB (Apacer AS2280P4 1 TB, PCIe 3.0 x4, NVMe 1.3, M.2 2280)

- Seagate BarraCuda 4 TB ST4000DM004 (SATA 6 Gb/s, 3,5")

- Seagate BarraCuda 4 TB ST4000DM004 (SATA 6 Gb/s, 3,5")

- GIGABYTE SSD 240 GB (SATA 6 Gb/s, 2,5")

 

I could buy another 32GB (same 2x16 modules) on short notice, to test things out (and to keep if all 4 modules are fine).

[JorgeB]

1 hour ago, unJack said:

I could buy another 32GB (same 2x16 modules) on short notice, to test things out (and to keep if all 4 modules are fine).

Looks like a RAM problem, you can also test the current DIMMs by using one at a time.

[unJack]

And again you were right. In the end I replaced one module after watching 

grep "[0-9]" /sys/devices/system/edac/mc/mc*/csrow*/ch*_ce_count

closely, and that fixed it.

Thanks Jorge!