Show TOC

Model CompressionLocate this document in the navigation structure

Use

While Embedded Search is running, the system compresses the models that you have created in the modeler in the background by default.

Model compression is used for models that have a 1:1 assignment between the superordinate and subordinate nodes. The attributes of the nodes that fulfill this prerequisite are stored in a single TREX index.

Fewer TREX indexes are generated when using model compression, which reduces memory usage on the TREX server and improves search performance.

You can define the assignments of a model in the modeler in the Define Structure step using the fields Cardinality and Inverse Cardinality.

You can see in the administration cockpit whether or not model compression is used for a specific model. Expand a search object connector until the physical indexes are displayed: Nodes of a search object connector that are stored in a single index have the same index ID at the Physical Indexes hierarchy level.