I am running 6.5.3 and experienced a hard crash yesterday, and needed a hard reboot - as the array had locked up I was not able to obtain a diagnostics report.

 

I have run a correcting parity check, which found 925 sync errors. However I am concerned about the hardware faults listed in the syslog which I run after the check was completed.

 

Please find attached diagnostics report and syslog

backup-diagnostics-20180917-0851.zip

backup-syslog-20180917-0844.zip

Likely a bad DIMM, errors that appear on the syslog were correct by ECC, before there was probably an uncorrectable error and when this happens with ECC RAM the system halts to prevent corruption, you need to identify the bad DIMM, e,g, remove one at the time and check syslog for errors after a few minutes or hours.

Thanks for your reply, the problem is that I have a pair of 4GB ECC RAM installed, and I can't run on one RAM module

 

I am going to try and run Memtest for a good few hours and see if it throws up any errors.

Edited September 17, 20187 yr by clevoir
Spelling Mistake

24 minutes ago, clevoir said:

and I can't run on one RAM module

Why not?

 

24 minutes ago, clevoir said:

I am going to try and run Memtest for a good few hours and see if it throws up any errors

Memtest won't throw any errors for ECC correctable errors, and like mentioned an uncorrectable error will halt the system.

Thanks for your reply, I was under the impression that RAM had to be installed in pairs of modules.

 

But if Memtest is not going to throw up any faults, I'll have a go at running each module at a time and look for errors as you have suggested.

 

Thanks