e i g h t - c u b e d . c o m 

Problems solved

It's so simple when you look back in hindsight. A patch for the memory channel driver was just released, and it appears to be a perfect fit for the problem we are experiencing.

Why hadn't we run across this problem before? Just recently, we repartitioned the GS160s to make more CPUs available. The two instances that have experienced the hang are the two instances with the additional CPUs. The theory is that adding the new CPUs changed the timing through the driver code enough to expose the problem.

Typical.