【正文】
difications are required to their source code, libraries, or binaries. In addition, the proposed approach is based on userlevel virtualization techniques, which requires no changes to existing O/Ss and can be conveniently deployed on grid resources. Furthermore, userlevel enhancements designed for gridstyle environments are built upon the virtualization layer to enable data provisioning with applicationdesired characteristics. In short, the proposed GVFS approach answers the first question by providing transparent gridwide data access for unmodified applications and O/Ss through the userlevel DFS virtualization. ApplicationTailored Grid Data Provisioning The second question is, how to provide data with applicationtailored optimizations? Typical O/Ss are designed to support generalpurpose applications, but it is often the case that one size does not fit all. Applications have diverse characteristics and requirements, in terms of, for example, data access patterns, acceptable caching and consistency policies, security concerns, and fault tolerance requirements. To provide the desired performance, security, and reliability to a grid application, data provisioning needs to be optimized according to the application39。 FILE SYSTEM VIRTUALIZATION AND SERVICE FOR GRID DATA MANAGEMENT 1. INTRODUCTION Computations are being increasingly larger scale, in terms of both size and geographical and administration distribution. Examples include scientific grids [1] which harness resources among several institutions for coordinated problem solving, and enterprise information systems that aggregate eforts from multiple sites for collaborative development. Common in these systems is that applications and data are distributed on resources across administrative boundaries and widearea works. Such environments can be referred as the gridstyle environments, which have the following distinctive characteristics: Heterogeneity: There exist a wide variety of applications and resources in a gridstyle environment. The resources typically have diferent hardware configurations (., CPU speed and architecture, memory size, disk bandwidth and capacity) and software setups (., operating systems and libraries)。這種自主管理是適用于幾個(gè)重要的方面,包括數(shù)據(jù)會(huì)話緩存配置,數(shù)據(jù)復(fù)制和會(huì)話重定向。這些服務(wù)還提供可互操作的接口允許與其他網(wǎng)格中間件服務(wù)和自動(dòng)化的直接相互作用數(shù)據(jù)處決供應(yīng)任務(wù)。 本論文提出的數(shù)據(jù)管理服務(wù)來管理每個(gè)應(yīng)用程序 GVFS中會(huì)話,執(zhí)行獨(dú)立會(huì)話之間的隔離,并為每個(gè)會(huì)話申請所需的定制。例如,從一個(gè)應(yīng)用程序用戶的角度來看,它是理想的作業(yè)執(zhí)行快捷,安全,可靠,從一個(gè)資源提供者的角度來看,這是預(yù)期資源利用是健康和盈利能力。這些要求往往超出了最終用戶和系統(tǒng)的能力,甚至管理員。 會(huì)議定制的數(shù)據(jù)也意味著按照所需的行為和周圍的環(huán)境各相關(guān)因素的 考慮許多參數(shù)和調(diào)整。然而,在大型系統(tǒng)中,許多動(dòng)態(tài)數(shù)據(jù)會(huì)話管理是另一個(gè)具有挑戰(zhàn)性的任務(wù),因?yàn)樗鼜?fù)雜性。 因此,對第二個(gè)問題的答案 是使用應(yīng)用程序定制啟用增強(qiáng)了 GVFS中提供應(yīng)用程序所需的電網(wǎng)范圍內(nèi)的數(shù)據(jù)會(huì)議性能,一致性,安全性和可靠性。特別是,增強(qiáng)設(shè)計(jì)網(wǎng)格式的環(huán)境是虛擬化層時(shí)提供 GVFS中,這包括可定制的磁盤高速緩存的高性能數(shù)據(jù)和多線程接入,應(yīng)用程序所需的數(shù)據(jù)一致性協(xié)議的一 致性效率,強(qiáng)并網(wǎng)兼容的安全保障網(wǎng)范圍內(nèi)的數(shù)據(jù)訪問和可靠性協(xié)議支持應(yīng)用程序透明的故障檢測和恢復(fù)。工具包為基礎(chǔ)的解決方案通常為用戶提供強(qiáng)大的 API 方案遠(yuǎn)程數(shù)據(jù)訪問與期望的行為,但很少有熟練的程序員利用這些 API 的有效利用。因?yàn)橐粋€(gè)應(yīng)用程序定制的優(yōu)化(例如,有進(jìn)取心的預(yù)取文件內(nèi)容),可能會(huì)導(dǎo)致性能(例如,其他幾個(gè)稀疏文件退化,數(shù)據(jù)庫),應(yīng)用量身定做的功能通常沒有實(shí)現(xiàn)通用 O / S的內(nèi)核。應(yīng)用程序具有不同的特點(diǎn)和要求,例如,數(shù)據(jù)訪問模式,可以接受的緩存和術(shù)語一致性的政策,安全問題和容錯(cuò)的要求。 總之,這種方法提供建議 GVFS中回答了第一個(gè)問題 ,即 透明網(wǎng)寬未修改的應(yīng)用和 O /不銹鋼通過數(shù)據(jù)訪問用戶級別的 DFS的虛擬化。 /不銹鋼并可以方便地部署在網(wǎng)格資源。由于眾所周知的 DFS的界面是由 GVFS中保存并提交給應(yīng)用程序,沒有修改要求他