【正文】 ty. To address these challenges, this dissertation presents a twolevel data management system in which file system virtualization provides applicationtailored gridwide data access, and servicebased middleware enables autonomic management of the data provisioning. In particular, this system has made the following contributions: It provides ondemand, crossdomain data access transparently for unmodified applications and O/Ss based on userlevel virtualization of widely available O/Slevel distributed file systems (DFSs). It supports applicationtailored enhancements designed for gridstyle environments on several important aspects of remote data access, including performance, consistency, security, and reliability. It employs middleware services to achieve ?exible and interoperable management of gridscale data provisioning, which is capable of controlling the lifecycles and configurations of dynamic data sessions based on application needs. It develops autonomic functions to automatically optimize the data management according to highlevel objectives, in order to reduce the plexity of managing data sessions and adapt them promptly to changing environments. Finally, the proposed system has been demonstrated, with thorough experimental evaluation, that it is efective and can significantly outperform conventional DFSbased approaches in gridstyle environments。為了進(jìn)一步降低管理數(shù)據(jù)會(huì)話的人為干預(yù)，使他們迅速適應(yīng)不斷變化的環(huán)境，植物神經(jīng)功能內(nèi)置于數(shù)據(jù)管理服務(wù)，使他們自動(dòng)監(jiān)測(cè)，分析能力 ， 優(yōu)化電網(wǎng)范圍內(nèi)的數(shù)據(jù)會(huì)分散的實(shí)體，合作共同努力實(shí)現(xiàn)所需的數(shù)據(jù)配置和資源使 用的目標(biāo)。為了解決這個(gè)問題，用戶級(jí)別的 DFS支持自定義，則建議應(yīng)用定制的 GVFS中數(shù)據(jù)會(huì)話。相反，數(shù)據(jù)網(wǎng)格管理需要專門處理這些獨(dú)特的問題。 ?規(guī)模：大量資源可以在網(wǎng)格式的環(huán)境中匯總 。 ?活力：在網(wǎng)格式的環(huán)境中部署的系統(tǒng)具有高度的活力。 在網(wǎng)格式環(huán)境中的 DFS為基礎(chǔ)的辦法，它也被 成功在生產(chǎn)電網(wǎng)系統(tǒng)部署了數(shù)年，支持科學(xué)工具和許多學(xué)科的用戶。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。 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(qǐng)所需的定制。s needs. Therefore, the answer to the second question is to use the applicationtailored enhancements enabled by GVFS to provide gridwide data sessions with applicationdesired performance, consistency, security, and reliability. ServiceBased Autonomic Data Management The third question is, how to manage data provisioning in a gridscale system with dynamically changing environments? Based on the GVFS approach, data sessions can be started on demand and independently customized for applications. However, in a largescale system, the management of many dynamic data sessions is another challenging task due to its plexity. Data sessions need to be dynamically established and destroyed based on the lifecycles of applications and the locations of their instantiations and data storage. Customization of data sessions also implies the consideration of various relevant factors and tuning of many parameters, in accordance with the desired behaviors and the surrounding environments. Dynamically changing application workload and resource availability further require continuous monitoring of data sessions and timely adaptation of their configurations. These requirements are often beyond the capability of endusers and even system administrators. Yet the goals of users or administrators are rather simple and explicit. For example, from an applic