Abstract
Managing distributed high-performance storage systems is complex and, although sharing common ground with traditional network and systems management, presents unique storage-related issues. Integration technologies and frameworks exist to help manage distributed network and system environments. Industry-driven consortia provide open forums where vendors and users cooperate to leverage solutions. But these new approaches to open management fall short addressing the needs of scalable, distributed storage. We discuss the motivation and requirements for storage system management (SSM) capabilities and describe how SSM manages distributed servers and storage resource objects in the High-Performance Storage System (HPSS), a new storage facility for data-intensive applications and large-scale computing. Modem storage systems, such as HPSS, require many SSM capabilities, including server and resource configuration control, performance monitoring, quality of service, flexible policies, file migration, file repacking, accounting, and quotas. We present results of initial HPSS SSM development, including design decisions and implementation trade-offs. We conclude with plans for follow-on work and provide storage-related recommendations for vendors and standards groups seeking enterprise-wide management solutions.
Highlights
Management applicationsand tools focused management (SSM) capabilitiesand describe how system management (SSM) on networks and components on networks, such as hubs, manages distributed servers and storage resource objects in bridges and routers
Modern storage systems, such as High Performance Storage System (HPSS), subsequent transition to more open heterogeneous require many SSM capabilities,including server and environmentsled to distributedapplicationsconsistingof resource configurationcontrol, performance monitoring, multiple cooperating processes running on several nodes
A well-constructed storage system should meet the following SSM concerns: Hierarchical storage management Fault detection and data recovery Resource configuration control Secure management operations Accounting and usage limits Quality of service flexibility Balanced internal instrumentation Proactive/seE-managed components. We describe how these requirementswere add~ssedin a modern, highperformance, distributedstoragesystem, in tbe presence or absenceof suitablemanagement technologiesfor storage
Summary
Reference models and frameworks specificmanagement roles and rules, but rather provides an extensible structurein which to implement and enforce. SSM requires an ability to define policies that ensure HSM operationsmeet specific site requirements Another issue for storage systems, especially those that are highly distributed is flexible m e r and resource configuration control. A well-constructed storage system should meet the following SSM concerns: Hierarchical storage management Fault detection and data recovery Resource configuration control Secure management operations Accounting and usage limits Quality of service flexibility Balanced internal instrumentation Proactive/seE-managed components. Data Server include starting, stopping, and modifying states of servers, configuration of Encina files, importing and exporting physical media, Creation of new virtual volumes, control of storage devices and jobs, viewing and updating managed objects, and delogging. As a predecessor to HPSS, NSL-UniTree shares some HPSS high-level SSM design Both systems use a general Health and Status display and both separate the Data Server and System Manager functions. The flow of information through HPSS when starting a server is briefly described below
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
