Improving I/O performance using soft-QoS-based dynamic storage cache partitioning

Abstract

Abstract-Resources are often shared to improve resource utilization and reduce costs. However, not all resources exhibit good performance when shared among multiple applications. The work presented here focuses on effectively managing a shared storage cache. To provide differentiated services to applications exercising a storage cache, we propose a novel scheme that uses curve fitting to dynamically partition the storage cache. Our scheme quickly adapts to application execution, showing increasing accuracy over time. It satisfies application QoS if it is possible to do so, maximizes the individual hit rates of the applications utilizing the cache, and consequently increases the overall storage cache hit rate. Through extensive trace-driven simulation, we show that our storage cache partitioning strategy not only effectively insulates multiple applications from one another but also provides QoS guarantees to applications over a long period of execution time. Using our partitioning strategy, we were able to increase the individual storage cache hit rates of the applications by 67% and 53% over the no-partitioning and equalpartitioning schemes, respectively. Additionally, we improved the overall cache hit rates of the entire storage system by 11% and 12.9% over the no-partitioning and equal-partitioning schemes, respectively, while meeting the QoS goals all the time.

On Urgency of I/O Operations

“…[13], [9] partition the shared file system cache between multiple processes by detecting the access patterns. [20] proposes an online learning technique to manage shared storage cache in order to provide soft-QoS guarantees to applications. [10], [16] and [3] use feedback control theory techniques to provide long-term hit rate assurances to competing classes of applications.…”

Design and implementation of caching services in the cloud

“…The variants of the cache partitioning problem that we discussed arise in various other contexts such as hardware caches [21], web caches [11], storage caches [27][28][29][30], and in database cache management [20,31]. These papers are focused on fair service differentiation and QoS goals [11,20,28], maximizing the total cache hit rate [21,31], or a combination of the two as in SC2 [27,29]. Ko et al [28] and Storm et al [20] discuss control theoretic approaches to stabilizing cache partitioning over time, which we intend to incorporate into a future version of SC2.…”

Improving Caches in Consolidated Environments

“…First, memory ballooning [Wal02b] can impact the guest file system buffer cache quickly, but it is difficult to exactly control the size of the buffer cache. Second, memory partitioning has been used in order to provide softQoS controls for workloads [PSPK09,PGSK09]. Third, I/O throttling [GAW09] allows for rapid control of the VM I/O performance in a cluster.…”