☁️

For a complete distributed operating system, its network virtual memory cache strategy should satisfy: evaluate the size of all operands and operator definitions (function / method / behavior), cache the smaller ones to the host where the larger ones are in to perform operations; evaluate the allocatable amount of the hardware resources of the operators (or the performance of the host), perform operations on the host with a large allocatable amount (or high performance) preferentially; evaluate the traffic among operators and operands, move ones with higher traffic closer preferentially on the distance and occupy preferentially on the bandwidth. Due to the cached memory of the distributed virtual memory system is cached and stored back by multiple regional memories, the system must actively update the cached operands in other regional memories whenever an element is stored back. The distributed virtual memory system mainly integrates the reference space of the entire network to achieve the real Cloud realm.