Improving Application Launch Performance on Solid State Drives

Abstract

Application launch performance is of great importance to system platform developers and vendors as it greatly affects the degree of users' satisfaction. The single most effective way to improve application launch performance is to replace a hard disk drive (HDD) with a solid state drive (SSD), which has recently become affordable and popular. A natural question is then whether or not to replace the traditional HDD-aware application launchers with a new SSD-aware optimizer. We address this question by analyzing the inefficiency of the HDD-aware application launchers on SSDs and then proposing a new SSD-aware application prefetching scheme, called the Fast Application STarter (FAST). The key idea of FAST is to overlap the computation (CPU) time with the SSD access (I/O) time during an application launch. FAST is composed of a set of user-level components and system debugging tools provided by Linux OS (operating system). Hence, FAST can be easily deployed in any recent Linux versions without kernel recompilation. We implement FAST on a desktop PC with an SSD running Linux 2.6.32 OS and evaluate it by launching a set of widely-used applications, demonstrating an average of 28% reduction of application launch time as compared to PC without a prefetcher.

Reducing the User-perceived Latency of Browsers with NVRAM

“…In [7], they proposed a prefetching scheme that overlaps the computation time and I/O access time during an application launch. Actually, their work provided many hints for us to make a research.…”

“…1(a). It is a cold start scenario, which is the first launch of an application since the system boot-up [7].…”

HyCache: A User-Level Caching Middleware for Distributed File Systems

“…Also for extending virtual memory with Storage Class Memory (SCM), SCMFS [27] concentrates more on the management of a single SCM device. FAST [28] proposed a caching system to pre-fetch data in order to quicken the application launch. [29] considers SSD as a read-only buffer and migrate those random-writes to HDD.…”

Application Prefetcher Design Using both I/O Reordering and I/O Interleaving

“…Once an application accesses a set of data blocks in a particular order during a launch, it is highly likely to access almost the same set of data blocks in almost the same order in the next launches [1]- [3]. Such an I/O sequence is called an application launch sequence in recent work [2]. The deterministic launch behavior of applications allows application prefetchers to predict with high accuracy what data block will be requested next.…”

“…FAST [2]-originally proposed for SSDs-is an alternative approach by overlapping CPU computation with I/O accesses. This sort of "I/O interleaving" technique has been studied for HDDs as well [4].…”