Rule Builder > Decision Tables and Trees > Scalability |
Scalability |
PREVIOUS NEXT |
ILOG JRules can handle arbitrarily large decision tables and trees; however, as the size of the decision table or tree grows, performance can be degraded based on the available memory of the host machine (during both editing and execution). Most decision tables or trees will never experience a degradation; however, there are actions you can take if you expect your decision table or tree to grow to thousands of branches.
The most straightforward solution is to partition your table or tree into two or more tables or trees based on the conditions. For example, rather than build one large Patient table or tree, you might build two smaller Male Patient and Female Patient tables or trees. In addition, you can speed up the performance by disabling the overlap checking, which can be very time consuming.
This type of up-front planning will ensure that your application scales as the number of rows / branches in your decision table / tree increases.
Customer Support | Copyright © 1987-2004 ILOG S.A. All rights reserved. Legal terms. | PREVIOUS NEXT |