Wayne C. Hineman

ADSM: a multi-platform, scalable, backup and archive mass storage system

The ADSTAR Distributed Storage Manager (ADSM) is a client-server system whose goal is to provide the illusion of infinite storage for a range of applications. The first deployed application is backup and archive services. ADSM is a configurable system targeted to serve any client at user installations with heterogeneous architectures that offers the same level of data management for LANs, midrange and mainframe systems. ADSM moves the responsibilities of backup and archive media management to the server while maintaining the benefits of local workstation systems. In ADSM, incremental backups are all that a user needs to do to restore its data with maximum performance. The ADSM server has been deployed in five IBM and two non-IBM platforms, while the ADSM clients operate in more than twelve IBM and non-IBM platforms.

Archive storage system design for long-term storage of massive amounts of data

A dramatic shift is underway in how organizations use computer storage. This shift will have a profound impact on storage system design. The requirement for storage of traditional transactional data is being supplemented by the necessity to store information for long periods. In 2005, a total of 2,700 petabytes of storage was allocated worldwide for information that required long-term retention, and this amount is expected to grow to an estimated 27,200 petabytes by 2010. In this paper, we review the requirements for long-term storage of data and describe an innovative approach for developing a highly scalable and flexible archive storage system using commercial off-the-shelf (COTS) components. Such a system is expected to be capable of preserving data for decades, providing efficient policy-based management of the data, and allowing efficient search and access to data regardless of data content or location.

A bit-parallel search algorithm for allocating free space

File systems that allocate data contiguously often use bitmaps to represent and manage free space. Increases in the size of storage to be managed creates a need for efficient algorithms for searching these bitmaps. We present an algorithm that exploits bit-parallelism, examining all bits within a processor word at the same time, to improve search performance. Measurements of our implementation show that these techniques lead to a 14 times increase in the rate at which bitmap pages can be searched on a 64-bit processor. Trace-driven experiments indicate that overall allocation performance increases by a factor of 3 to 6 on a 32-bit processor. As processors mature, registers become wider and the degree of bit-level parallelism increases, which makes the performance improvements of our search algorithm more substantial.