论文部分内容阅读
内存空间紧张下频繁的外存交换严重影响程序的运行性能,在内存过载时提高内存利用率和减少外存交换问题上,现有的研究主要集中在内存压缩和去重两个方面,压缩只能删除页内重复数据,不能做到页与页之间的相同数据删除;将去重技术与系统虚拟内存结合,删除换出页中的重复页,仍不能有效避免外存交换区访问.提出一种内存过载下基于重删的内存优化策略.利用页框回收算法选出适合去重的内存页,过滤其中全零页以减少存储和去重计算开销,删除重复页以提高内存利用率,并通过存储唯一数据页于内存去重区,有效降低外存交换操作,同时提供可配置的去重内存空间占用上限,避免过度去重引入的颠簸问题.在IO密集型、计算密集型、及常规应用场景下的实验测试表明,该策略能有效删除重复内存页,可以极大程度地减少外存交换操作,同时提高程序运行性能最高可达54.013%.
Frequent memory space under the tension of the frequent exchange of memory seriously affect the operation of the program in the memory overload to improve memory utilization and reduce the memory swap problem, the existing research focuses on memory compression and de-duplication of both compression only Can delete duplicate data in the page, can not be the same data between pages and delete; Combining deduplication technology and system virtual memory, delete duplicate pages in the swap page, still can not effectively avoid the memory swap access. A memory-based memory optimization strategy based on deduplication.Use the frame recovery algorithm to select the memory pages suitable for de-duplication, filter all zero pages among them to reduce storage and de-duplication overhead, delete duplicate pages to improve memory utilization, And by storing unique data pages in the memory de-duplication area, effectively reduce the memory swap operation, while providing configurable memory footprint limit to avoid over-reloading introduced bumpy problems.In IO-intensive, computationally intensive, and Experimental tests in the conventional application scenarios show that this strategy can effectively remove duplicate memory pages and can greatly reduce the memory swap operation and improve the program running performance High up to 54.013%.