Overview

Oracle Coherence is the world’s first—and the leading—in-memory data grid, with proven robustness and the richest feature set in the market. Designed for data management and grid computing, this cloud-friendly clustered software can be used to scale mission-critical applications, quickly access data while offloading backend systems, implement parallel distributed computing algorithms and event-driven architectures, and store data in microservices and AI applications. Oracle Coherence is a fast, key-value store with optional persistence, offering fault-tolerant automatic sharding, polyglot and REST interfaces, querying, transactions, eventing, and in-place processing, as well as change data capture, multi-site federation, messaging, and integrations with popular frameworks.

Oracle Coherence features

Low-latency data storage: Key-value object storage

Coherence stores frequently accessed data as serialized key-value pairs in a Java Virtual Machine (JVM) cluster with automatic sharding. Read, query, and write latencies are fast—on the order of 1 millisecond—to achieve maximum application performance and stability.

Scalability and availability: Elasticity

Coherence ensures maximum scalability and performance by providing clustered low-latency data storage that enables multiple applications, including in different languages, to rapidly read and write data and ingest asynchronous event streaming at scale. Additionally, all Coherence services provide failover and failback without any data loss through the Coherence Cluster structure.

Polyglot grid programming: Parallel in-place processing

Deploying mixed language/polyglot functions to the data grid for parallel execution in-memory allows for performance gains compared to batch or serial fetching the data from storage. Coherence’s lock-less concurrency control and efficient, atomic transactions minimize contention and latency, resulting in improved system throughput and fault tolerance.

Asynchronous event streaming: Listeners and messaging

For highly scalable and decoupled event-driven architectures, Coherence provides event models both within server processes and between servers and clients, as well as messaging, with publishers, topics, and subscribers.

Data source integration: Read/write backing maps

Coherence maps are logical structures that can read from and write to arbitrary backing data sources synchronously or asynchronously, such as a database or disk file. To ensure the data your application is operating on is current, any changes to the source database are replicated by Oracle GoldenGate HotCache as it efficiently updates the Coherence cache. Low latency is assured because the data is pushed when the change occurs in the database, ensuring you are always operating on current data.

Multi-site data federation: Federated caching

Coherence offers federated caching to link multiple clusters so that cache data is automatically synchronized between clusters. This multi-site architecture provides redundancy, offsite backup, and multiple points of access for application users in different geographical locations.

Cloud native deployment and operation: Docker/Kubernetes support

Coherence is available in Docker images and has its own Kubernetes Operator. Grafana and Kibana dashboards facilitate monitoring. Coherence is also offered in Oracle Cloud Infrastructure Marketplace for instant provisioning.

HTTP session-state management: Coherence performance for in-memory session management

Coherence*Web is an HTTP session management module dedicated to managing session state in clustered environments. Easily configure fine-grained session and session attribute scoping by way of pluggable policies. Adding Coherence*Web to an application allows user sessions to survive process failure, machine failure, even site failure in multisite active/active architectures, and requires no code changes.

WebLogic Server integration: Managed Coherence servers

The WebLogic Server ecosystem can be used to manage Coherence clusters, including the Administration Console and WLST. A grid archive type is defined alongside EARs and WARs, including a container contract with lifecycle and events.