Content Feeder and CAE Feeder now cache the results of Apache Tika text and metadata extraction for indexed blob values. This improves feeding performance, if content is re-indexed with unchanged blob value, or if the same blob value is used in multiple content items. The Blueprint configures a default cache capacity of estimated 10 MB heap memory using configuration property cache.capacities.feeder.tika.heap in application.properties. Because of that, Feeders require slightly more heap memory. Caching can be disabled by setting the cache capacity to 0.

(CMS-28599)