A simulation model of the ELHILL 3 system of the National Library of Medicine

Abstract

The ELHILL 3 system of the National Library of Medicine is one of the world's largest online information retrieval systems, providing access to many different files via MEDLINE, TOXLINE, CATLINE, SDILINE, CHEMLINE, AVLINE, etc. In order to determine whether cost reductions are possible for systems of online bibliographic retrieval, stochastic models were developed that are capable of projecting the operational and cost performance of different configurations handling the same known workload as ELHILL 3. The existing configuration (multiprocessor system consisting of two tightly‐coupled IBM 370/158s) was first modeled, establishing a baseline for performance comparison. A systems approach was followed in building the model. This involved decomposition of the ELHILL 3 system into quantifiable components, analysis of the components, integration of the components into a network model, and performance measurement via discrete event procedures. The model is a network of processing activities, decision points, queues, and services, and the passage of individual user transactions through the retrieval system is simulated. Alternative configurations are modeled (e.g. distributed processing systems) for ELHILL 3, and tested for cost and performance projections.

Citation

Hillman, D.J. (1980), "A simulation model of the ELHILL 3 system of the National Library of Medicine", Online Review, Vol. 4 No. 3, pp. 281-288. https://doi.org/10.1108/eb024035