Deploying cacheless search

When you deploy cacheless search, you must choose between a product-scoped deployment strategy and a full database deployment strategy.

Full database deployment

Teamcenter generates bounding boxes for all geometric entities in batches, ignoring items without geometry. It places all update candidates in the indexer queue. Update candidates are any item revisions with bounding boxes.

Product-scoped deployment

You mark one or more products as indexable, and Teamcenter processes only the relevant items. No indexes are created for old and inactive programs, saving system resources.

The feasibility of processing the complete database depends on the size and number of bounding boxes to process. As a guideline, product-scoped deployment is probably the best strategy if less than 30% of your data is in active use.

To obtain a baseline time, you can test a representative sample of data and then estimate the time needed to process all the bounding boxes and indexes. For example, if the entire database cannot be processed in one weekend, you may prefer the product-scoped deployment strategy. Even with this strategy, the initial preparation and processing of a single product may take hours or even days, depending on the number of components in the product.

An alternative strategy is to use a product-scoped deployment when you first populate the database, that is, to index your data one product at a time. You can then switch to a full database deployment when the database is populated with all products.