/////////////////////////////////////////////////////////////////////////////// Copyright (c) 2000, 2026, Oracle and/or its affiliates. Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl. /////////////////////////////////////////////////////////////////////////////// = Introduction :description: What Oracle Coherence CE does, where it fits, and how to get started :keywords: coherence, distributed data, caching, microservices, computation, analytics, vector search // DO NOT remove this header - it might look like a duplicate of the header above, but // both headers serve a purpose, and the docs will look wrong if it is removed. == Introduction Oracle Coherence Community Edition (CE) is an open source, distributed data platform for applications that need fast access to shared state and dependable processing at scale. It brings data and application logic together: store data across a cluster, run operations on the members that own it, and add capacity as your workload grows. You can start by using Coherence to reduce database load and speed up data access. The same platform also supports applications that manage their primary working state in Coherence, distribute business processing, react to events, or search application data by meaning. Distributed maps, queries, computation, messaging, and persistence give you several ways to build around the same data. This introduction describes the platform in the context of CE 26.10. The xref:docs/core/01_overview.adoc[Core Improvements] guides describe the features and changes specific to that release. == What Coherence makes possible === Fast access to shared application data Keep frequently used customer, product, account, or reference data available to many application instances without making every request return to a database. Coherence distributes the data across storage members and routes requests to the members that own the relevant entries. Applications can use a shared distributed map, a near cache that keeps recently accessed data close to a client, or a filtered view that stays current as the underlying data changes. Cache stores integrate Coherence with databases and other systems, including read-through and write-through or write-behind patterns. Choose the pattern according to which system owns the data and how updates must be coordinated. === Business processing where the data lives A service often needs to inspect and update a record together, or apply the same calculation across many records. Coherence can run that logic on storage members, reducing the amount of data an application has to retrieve and send back. Entry processors provide exclusive access while processing an entry; distributed aggregators combine work performed across the cluster. Data affinity places related records in the same partition. This supports operations across those records, including partition-local transactions, while keeping their processing together. It is useful for work such as updating an order with its associated entries or calculating a result for one customer's data. Java developers access these capabilities through familiar map concepts and APIs such as `NamedMap`, `NamedCache`, and the Repository API. The bootstrap API and configured sessions provide the application's entry point to these resources. === Answers that follow changing data Use distributed queries, indexes, streams, and aggregators to find records and calculate summaries without first copying the complete data set into an application. For dashboards and other recurring questions, CE 26.10 adds xref:docs/core/05_continuous_aggregation.adoc[Continuous Aggregation]. It maintains partial results as entries change, then combines those results when requested. Frequently queried totals, counts, and grouped summaries can avoid repeatedly scanning the same data; the guide explains the supported aggregators and the trade-off between query savings and maintenance work. Map listeners and interceptors let applications respond to data changes. Topics provide publish/subscribe messaging, and the distributed executor and concurrency APIs support background work and coordination across processes. === Search and AI over application data Coherence's vector search capabilities let applications find similar items and combine similarity queries with filters on business attributes. This makes it possible to search embeddings alongside the application records they describe. The xref:examples/guides/300-vector-store/README.adoc[Vector Store guide] introduces the available search and indexing options. Optional modules extend the platform with distributed full-text search through Apache Lucene and a retrieval-augmented generation (RAG) framework. The RAG framework provides document ingestion, chunking, vector storage, and integration with embedding and language-model providers. These modules let applications build retrieval workflows around data managed by the cluster. == How the cluster supports your application Coherence partitions data across storage members and can maintain backup copies on other members. When a member leaves or fails, the cluster promotes available backups and redistributes ownership. Adding members lets the cluster rebalance data and processing across the available capacity. This model separates the application's view of a named map from the placement of individual entries. Applications work with their data while Coherence manages ownership, routing, and partition movement. Choose backup counts and placement policies for the failures your deployment needs to tolerate. Persistence provides recovery beyond live replicas, and snapshots provide a way to capture and recover persisted state. xref:docs/core/01_unified_persistence.adoc[Unified Persistence] describes the journal persistence engine and its storage and recovery options. Planned maintenance can use rolling restarts and supported rolling upgrades; consult the release's compatibility and migration guidance before changing versions. Coherence can serve as a cache in front of another system of record, or manage persistent application data itself. That is an application architecture decision: define the authoritative data source, update path, durability requirements, and recovery process together. == Connect and operate in your environment Applications can join a Coherence cluster or connect remotely through proxies. Separate application and storage tiers when they need different deployment or scaling policies. Java, Go, Python, JavaScript, .NET, and C++ clients provide language-specific ways to access the cluster; consult each client's documentation for its supported APIs and release compatibility. Integrations with https://helidon.io/[Helidon], https://github.com/coherence-community/coherence-spring[Spring], and https://micronaut-projects.github.io/micronaut-coherence/latest/guide/[Micronaut] connect Coherence to application frameworks. You can run clusters on physical servers, virtual machines, or Kubernetes. The https://github.com/oracle/coherence-operator[Coherence Operator] manages Coherence deployments on Kubernetes. Metrics, health checks, tracing, JMX, and REST management provide visibility into the cluster. Use them to understand resource use, data distribution, requests, and service health as the application grows. CE 26.10 requires *Java 21 or later* and enables hardened mode by default. Read the xref:docs/core/04_security_hardening.adoc[Security Hardening guide] before an upgrade to prepare application metadata, review endpoint and remote-execution policies, and plan a staged migration where needed. == Choose a starting point * xref:examples/guides/050-bootstrap/README.adoc[Bootstrap Coherence] introduces the bootstrap API and sessions for a first Java application. * xref:examples/README.adoc[Guides and Tutorials] provides examples for data access, queries, processing, events, topics, and other application patterns. * xref:docs/core/01_overview.adoc[Core Improvements] covers Unified Persistence, worker pools, security, Continuous Aggregation, and Extend proxy performance. * {commercial-docs-base-url}/index.html[Oracle Coherence documentation] provides the broader configuration, development, administration, and reference material. Check edition availability when using documentation shared with commercial editions. ++++ ++++ [[community-edition]] == Community Edition and commercial editions CE is the open source edition of Oracle Coherence, released under the Universal Permissive License (UPL) 1.0. Oracle Coherence commercial editions provide additional capabilities and support. The following commercial functionality is not included in CE: * Management through the Oracle WebLogic Management Framework and WebLogic Server multi-tenancy * Deployment of Grid Archives (GARs) * HTTP session management for application servers (Coherence*Web) * GoldenGate HotCache and TopLink-based cache loaders and cache stores * Elastic Data Journal (RAM/flash journal backing maps) * Federation and WAN support * The Transaction Framework and CommonJ Work Manager These exclusions are distinct from the CE features described above. In particular, CE's partition-local transactions and Unified Persistence Journal are separate from the commercial Transaction Framework and Elastic Data Journal.