Scalable and robust engine for processing and acting on streaming data
Data-driven applications need to process data as it arrives, acting on data in real-time to be able to deliver information to applications and users as quickly as possible. However, existing technologies, born in a batch processing world, were not designed for this.
Streamlio address this need with proven, best-of-breed technology for real-time and streaming data processing. Building on the the streaming processing capabilities in Apache Pulsar and the Streamlio team’s experience as co-creators of the Apache Heron* real-time engine at Twitter, Streamlio enables enterprises to deliver data-driven applications that can meet demanding SLAs at any scale.
Streamlio’s event-based design makes it possible to process data immediately, as it arrives. Tune latency and throughput to meet demanding SLAs with end-to-end latencies in tens of milliseconds
Building on Apache Pulsar, the streaming solution proven in production at Yahoo, and experiences building and operating Apache Heron, the streaming data processing engine in production at Twitter, the Streamlio solution can handle 100s of billions of messages per day.
With a multi-tenant architecture that ensures workload isolation, the ability to tune throughput and latency for individual workloads, and stream-native processing jobs, the Streamlio platform can meet demanding SLAs even as data and workloads scale
Automated monitoring and healing features regulate data flow and processing to automatically address bottlenecks and ensure resiliency even as workload demands fluctuate and failures occur
Support for multiple processing semantics–at least once, at most once, effectively once–together with support for a serverless-inspired approach to stream-native processing makes it easy to meet diverse development needs