I have a whitebox ESXi server running 3.5.0 build 163429. The system had been running perfectly stable (ie. over 2 months) then we desided to add another 4 GB of RAM to bring it to 8 GB total. Well only about 7-10 days after adding the memory it crashed. So we removed it back to 4 RMA'ed the memeory thinking well it must be had. But the new stuff in and sure enough in 7 -10 days CRASH. All of it is the same exactly Kingston memory (Kingston KVR667D2D8F5/2G DDR2-667 2GB ECC). I just cannot explain why its fine with 2x2GB sticks of that but becomes unstable with 4x2gb sticks. This whole chassy has been replaced so I dont think its the motherboard. Thanks.
This issue can be caused by the motherboard/memory combination. On one dell model below a bios update the system will crash if all the dimms are filled on the motherboard.
The issue is not caused by bad memory, it is just a motherboard memory timing issue.
The PSoD that you posted suggests that your problem is in PCPU 3, since you didn't specify your specific model ( or more specifically how many CPUs and how many cores per CPU ) I can't tell you which physical CPU that is. It looks like the TLB failed to acknowledge a request sent to it. I think you're barking up the wrong tree looking at the RAM.
The CPU is an Intel Quad Core Xeon E5420 2.5GHz. So bascily its either a memory timing issue or processor? I guess I still find it odd as it run fine with just 2x2 GB sticks but not with 2x4. Then again we have had other PSOD related to that.
