Sizing Cleancache Allocation for Virtual Machines' Transcendent Memory

• Published in: IEEE transactions on computers Vol. 65; no. 6; pp. 1949 - 1963
• Main Authors: Venkatesan, Vimalraj, Tay, Y. C., Qingsong Wei
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.06.2016; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Allocations; analytical model; Benchmark testing; cache misses; Cleaning; Consolidation; Disks; fair allocation; Hierarchies; Mathematical analysis; Mathematical model; Random access memory; Resource management; Servers; transcendent memory; Virtual environments; Virtual machine monitors; Virtual machines; Virtualization; Workload; Workloads; cache misses; Virtual machines; fair allocation; analytical model; transcendent memory
• ISSN: 0018-9340; 1557-9956
• DOI: 10.1109/TC.2015.2456025

The old virtualization idea and the new multicore technology now make it possible to consolidate multiple workloads on one physical host. This helps reduce the amount of idle resources. In particular, transcendent memory is a recent idea to gather idle memory into a pool that is shared by virtual machines (VMs). It can be viewed as a new level in the memory hierarchy, between main memory and disks. Cleancache is the part of transcendent memory that is used for caching the VMs' clean pages. This paper shows that a Cache Miss Equation can accurately capture observed cleancache behavior, despite its unusual design. The equation can be used to dynamically size and partition cleancache for the VMs. Experiments with a variety of workloads show that this equation-based allocation can help to drastically reduce disk reads and fairly allocate cleancache space.