Fault Tolerance
- Based on the Raft consensus.
- Consistent fault tolerance for event-driven applications.
- Automatic failover with zero data loss.
Get access to enterprise-level technology and support: from development to production.
Aeron Cluster
Aeron Cluster is the world’s leading framework for high-performance, in-memory, fault-tolerant transactional services.
It enables resilient 24/7 workloads in the cloud, on-prem or in a hybrid environment. Automatic failover with zero data loss takes less than a second, minimising the impact on your business and ensuring critical data is retained.
- Fault-tolerant, strongly consistent, and 24×7 with zero downtime upgrades.
- Throughput of over 1 million msg/sec with predictable latency as low as 18 microseconds on prem and sub 100 microseconds in the cloud.
- Sequences and reliably replicates transactional messages at massive scale.
- Used for capital markets systems that demand 24/7 uptime and stateful recovery with no data loss.
- Runs Exchanges, RFQ, IOI or similar workflows.
Key features
In-Memory State
- State and application logic are co-located for optimal performance leading to a more elegant programming model than other distributed or actor-based programming patterns.
Cluster Recovery
- Aeron Archive records snapshots for local and remote replay, avoiding the need for a distributed file system.
- Reliable distributed timers, inter-service messaging and remote data centre backup.
Extreme Performance
- Over 1 million msg/sec with response times as low as 18 microseconds allows Aeron Cluster to be extremely efficient and keep hardware footprint to a minimum.
Hot System Upgrade
- Snapshotting and rolling upgrades allow for 24×7 operational models, enabling upgrades with zero downtime.
Performance Charts:
Aeron Cluster
Aeron Cluster running on physical hardware and leveraging Aeron kernel bypass. In this set-up Aeron cluster is receiving a message, persisting, replicating across nodes and responding to the message.
Aeron Transport
The world’s leading low-latency message transport system for performance, reliability and observability.
- Secure streams for unicast and multicast on prem and in the cloud
- IPC message transport for messaging on a single host
- Java, C/C++ and .NET
- Predictable latency in single digit microseconds
- Throughput of over 20 million messages per second
Key features
Predictable Low Latency
- Reliable message delivery over UDP for low latency and consistency without the fanout overhead of TCP.
- Predictable round trip time (RTT) in single digit microseconds.
Brokerless
- Better performance compared to a brokered messaging solution.
- Minimise costs: reduce hardware and OPEX costs typically associated with broker based systems.
Runs Anywhere
- On-prem, in the cloud and at co-location facilities.
- Suitable for hybrid cloud setup and WAN communication.
Fully Integrated With ...
- Aeron Archive to store and replay streams.
- Aeron Cluster for fault tolerant, transactional and strongly consistent applications.
Aeron Archive
Aeron Archive allows message streams to be persisted to disk at full message rate, for replay in real-time or at a later date. This allows any service that may become disconnected to reconnect and reliably recover without message loss.
Key features
Record
Take streams and write them to disk efficiently, with or without active subscription. Spy on an active publication.
Replay
Replay from a position in a recording: either direct from an archive, or started from an archive, and merged with a real-time stream.
Replicate
Replicate recordings from a source process to a destination process; useful for near real-time backup of data.
Manage
Purge, truncate or apply CRC checks to archived data for maximum flexibility.
Performance Charts:
Aeron Transport & Archive
Aeron Transport round trip response time in a private and AWS hosted network.
