infinispan

Why we included this project

Infinispan is a mature JVM data grid that has been around for a long time. It keeps an in-memory key/value store distributed across cluster nodes, so entries survive node failures and the cluster can scale up or down without losing data. You can deploy it as a hot cache in front of a slower database, as a primary in-memory NoSQL store, embedded inside your application, or as a standalone server. It can also index cached values with Lucene, which adds full-text search and, more recently, vector similarity queries to data you already hold, making it useful as the storage behind retrieval-augmented applications. JVM-heavy teams will appreciate that Keycloak and Hibernate already run on it, while anyone doing semantic search gets caching and similarity lookups in a single service.