ASSA search indexes with large amounts of data - Remote Matching Service - 23.1 - 23.1 - Brainware - external - Brainware/Remote-Matching-Service/23.1/Remote-Matching-Service-User-Guide/About-search-indexes/Add-a-search-index/ASSA-search-indexes-with-large-amounts-of-data - 2024-02-05 - When working with an ASSA index and large amounts of data, ASSA will distribute the data to multiple smaller ASSA indexes that contain no more than a million records each; these indexes will be searched in parallel.

Remote Matching Service User Guide

Platform
Brainware
Product
Remote Matching Service
Release
23.1
License

When working with an ASSA index and large amounts of data, ASSA will distribute the data to multiple smaller ASSA indexes that contain no more than a million records each; these indexes will be searched in parallel.

While this improves the performance, it also means a higher CPU usage for a single request. Keep this in mind when selecting the number of engine instances. For example, if you are indexing 1.5 million records each request will use two CPU cores instead of one.

If you have just eight cores available, you should not use more than four engine instances because more parallel searches could be active than there are CPU cores available.