STORAGE AREA NETWORKS AND THE HIGH PERFORMANCE STORAGE SYSTEM

Diploma

Abstract

The High Performance Storage System (HPSS) is a mature Hierarchical Storage Management (HSM) system that was developed around a network-centered architecture, with client access to storage provided through third-party controls. Because of this design, HPSS is able to leverage today's Storage Area Network (SAN) infrastructures to provide cost effective, large-scale storage systems and high performance global file access for clients. Key attributes of SAN file systems are found in HPSS today, and more complete SAN file system capabilities are being added. This paper traces the HPSS storage network architecture from the original implementation using HIPPI and IPI-3 technology, through today’s local area network (LAN) capabilities, and to SAN file system capabilities now in development. At each stage, HPSS capabilities are compared with capabilities generally accepted today as characteristic of storage area networks and SAN file systems.

 

Introduction

Storage Area Network (SAN) technology has a bright future as measured by its growing market acceptance. Web information source allSAN.com [10] reports that:

 

Within the mainframe arena, SANs already represent upwards of 25% of data center traffic. Outside of the mainframe area, SANs are expected to account for 25% of external disk storage and approximately 50% of multi-user tape storage by 2003

 

We believe that SAN technology will only reach its full potential when it can be used to provide secure sharing of data between heterogeneous client systems. To realize this potential requires appropriate storage system software and hardware architectures. One use for such a capability is a SAN-based global file system. A generic host-based file system provides capabilities such as a naming mechanism, data location management, and access control. A global file system extends this capability to multiple independent operating systems by using specialized protocols, locking mechanisms, security mechanisms, and servers to provide device access. A SAN-based global file system is distinguished from other global file systems by the characteristic that client computers access storage devices directly, without moving data through a storage server. 

 

The High Performance Storage System design and implementation are focused on hierarchical and archival storage services and therefore are not intended for use as a general-purpose file system. HPSS is nevertheless a file system, and specifically, a global file system. While any client applications (such as a physics code) can access HPSS devices with normal Unix-like calls to the HPSS client API library, in normal operation these applications are data transfer applications that transfer data between HPSS files and the local file system. HPSS has a network-centered architecture that separates data movement and control functions and offers a secure, global file space with characteristics normally associated with both LAN-based and SAN-based architectures. 

SAN Terminology

Several definitions of a Storage Area Network exist as related to common, shared repositories of data. The Storage Networking Industry Association (SNIA) online dictionary offers the following definition of Storage Area Network:

 

1.      A network whose primary purpose is the transfer of data between computer systems and storage elements and among storage elements. Abbreviated SAN. SAN consists of a communication infrastructure, which provides physical connections, and a management layer, which organizes the connections, storage elements, and computer systems so that data transfer is secure and robust. The term SAN is usually (but not necessarily) identified with block I/O services rather than file access services.

2.      A storage system consisting of storage elements, storage devices, computer systems, and/or appliances, plus all control software, communicating over a network.