This server has been running for weeks without issue, it's currently only using 1 of 2 CPUs as it is running in the office as a test bed (mainly due to the handle 12 bay storage/great for testing HDDs). The errors below seem to mainly be from AMD CPUs, it's only happened a single time and in the days since I got that error it hasn't occurrred.
Interestingly enough /proc/cpuinfo still shows all 4 cores of the CPU (Opteron 2373 Quadcore HE) and the functionality doesn't seem to be affected (nothing crashed etc..)
I'm not sure what the issue could be, I wonder if it's the power supply? It was getting a little whiny/noisy around the time this happened (the pitch it was making is also the beginning stages of power supply failure in my experience).
Message from syslogd@fs12home at Dec 6 18:05:50 ...
kernel:[3022800.704091] [Hardware Error]: CPU:0 MC4_STATUS[-|CE|MiscV|-|AddrV|CECC]: 0x9c054c84011c011b
Message from syslogd@fs12home at Dec 6 18:05:50 ...
kernel:[3022800.704127] [Hardware Error]: MC4_ADDR: 0x000000042d847a84
Message from syslogd@fs12home at Dec 6 18:05:50 ...
kernel:[3022800.704145] [Hardware Error]: Northbridge Error (node 0): L3 data cache ECC error.
Message from syslogd@fs12home at Dec 6 18:05:50 ...
kernel:[3022800.704169] [Hardware Error]: cache level: L3/GEN, tx: GEN, mem-tx: RD
hardware, cpu, mc, _status, ce, miscv, addrv, cecc, cache, gen, tx, mem, rd, northbridge, node, ecc, server, currently, cpus, mainly, hdds, errors, amd, hasn, occurrred, interestingly, proc, cpuinfo, cores, opteron, quadcore, functionality, doesn, etc, whiny, noisy, stages, syslogd, fs, dec, kernel, _addr,