This may be a bit off topic, but we have had quite a number of 6850s (say 2 or 3 per 100), single-core (declining), dual-core (declining), as well as quad-cores based, as they age, or should I say, as the RAM ages along with the 6850s, we have seen some odd issues. VMware ESX seems to abuse RAM. We tend to use our hardware for a minumum of 3 years and often use hardware the full 5 years or so, that typical end-of-life cycles support from Dell as well as other hardware vendors. Between 3 to 4 years of age, we see a inconsistent pattern of RAM issues. Servers that have worked without issue for 3 or more years just seem to start showing their age. Ignoring the expected hard disk failures, and once in a while a NIC or processor dies, or a cable goes bad, but RAM almost never until the last 3 years. In the last 3 years the frequency of RAM issues has dramatically increased. Be it a quality control issue given the major increase in density of DIMMs or materials quality issue, some how it happens. Most recently and frequently it has been with Dell OEM providers for RAM but I would not call it a Dell issue alone, since Dell relies on the same major RAM OEM providers as do HP, IBM, etc. What is a surprise is to have a 6850 that is just fine running ESX 3.5.x, rebuilt as ESX 4 (ESXi 4), and within hours, of initial reboot start throwing DIMM faults.
