Menu
Chronicle Queue Backed Map
Chronicle Queue Backed Map serves as a high-performance, drop-in replacement for standard Java Maps, delivering superior speed and zero garbage generation. Leveraging off-heap memory, it enables efficient data sharing between processes on the same host and supports replication for seamless distribution across LAN or WAN environments.
Talk to our teamKey Features
Zero-Compromise Persistence
QBM uses an underlying Chronicle Queue as its persistence backbone. Every map mutation is first appended to the Queue before being reflected in the Map.
Deterministic, Sub-microsecond Latency
By bypassing the JVM heap entirely and using memory-mapped I/O, map read and write operations remain in the nanosecond-to-microsecond range, with no stop-the-world GC pauses.
Replication
Maps are configured as a source or sink, with Chronicle Queue Enterprise replicating the underlying queue over TCP/IP to one or more remote nodes, ensuring that data continuity and throughput are never sacrificed for distribution guarantees.
Consistent State Snapshotting
QBM supports periodic, consistent snapshots of the full Map state, capturing a point-in-time materialised view of the key-value surface without interrupting read/write operations.
Scalability for Large Datasets
Memory-mapped files let you keep vast data sets such as full order books or historical ticks in a single map, limited only by available disk. Grow effortlessly as volumes rise.
Highly Performant
Proven to handle up to 30 million updates per second on a 16-core server with a 64 MB heap. Run real-time pricing, risk, and analytics workloads without costly hardware refreshes.
Technical Information
Explore the advanced engineering principles that shape this product’s core, ensuring exceptional performance, operational resilience, and seamless scalability.
Use Cases
Discover how Chronicle helps financial institutions solve real integration, latency, and scalability challenges proven in production environments.
