Advising cache for lower cache

Abstract

In a virtualized environment, such as a cloud computing environment, guest and host operating systems run simultaneously. Both of the operating systems have page caches for disk accesses. In such an environment, the host operating system cache does not work effectively because of negative temporal locality of access and duplicated storing in both the caches. In this paper, we propose a method for increasing the hit ratio of the host operating system page cache. The method increases locality of reference in host operating system cache by actively controlling cache replacement in guest operating system. The method makes certain data tend to be dropped in the guest operating system cache and frequently accessed in the host operating system cache. As a nodule, the specified data are actively stored in the guest operating system cache and accesses to these data are not transmitted to the host operating system cache. This implies that diversity of data in the host operating system cache decreases. This causes increase of locality of reference and hit ratio in the host operating system. After the proposal, we introduce our experimental implementation using a system call for giving advice about file accesses. We then present performance evaluation using filesystem benchmark in guest operating system and demonstrate that the method significantly improves I/O performance in guest operating system in situations wherein cache has large impact on performance.