The IMS Data Storage Hierarchy - dsh-1.
Abstract
To achieve high performance, large capacity, and high availability, the IMS data base computer takes the approach of implementing the logical information management functions by means of a pipeline of parallel microprocessors and makes use of a storage hierarchy with distributed control for the physical storage and manipulation of very large databases. The data storage hierarchy (DSH) of IMS provides a very large byte addressable virtual address space accessible by the large number of processors that implement the logical information management functions of IMS. A highly parallel structure is used by the DSH to support asynchronous processing of a large number of data requests in order to attain very high throughput. Use of multiple block sizes across the storage levels and use of efficient data movement algorithms contribute to the high performance of DSH. The use of distributed control as well as multiple data redundancy gives DSH the capability to tolerate hardware failures without suffering data loss or availability of service. This report describes the design objectives and the structure of a general data storage hierarchy (DSH-1). Research issues in DSH-1 are also discussed. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1979
- Accession Number
- ADA073375
Entities
People
- Chat-yu Lam
- Stuart E. Madnick