Millions of Records Per Second
1 — Acquire
Unstructured, structured, local, distributed, in-motion, at-rest, live or historical — SQLstream Blaze instantly welcomes all data, from all available sources, in all formats, and at all speeds.Learn more
2 — Analyze
Create contextual business insights through the perfect mixture of live and historical data streams.Learn more
- Development and updates take minutes, not months
- Operations can be run concurrently and continuously, without interrupting execution
- Performance scales to millions of records per second per CPU core with 1-5 millisecond latency
- ANSI-compliant SQL for query, and Java for custom ingestion/egress and for extending compute capabilities tap into existing skill sets
- Portable queries and other compute functions can be moved in out out of the system and reused in other contexts
- Operations expressed in standard ways are automatically optimized
- Implementations use existing database and data warehouse investments and can reduce hardware requirements up to 50x
- Development takes minutes: fastest time-to-market
- Rich feature set (for stream ingestion/egress, processing, querying, and automated actions) covers the entire stream processing data flow; no additional technologies needed.
StreamLab is a visual platform for interactively exploring machine data streams and building real-time streaming apps without having to write any code. An intelligent suggestions engine supports users in building complete apps from ingestion to analysis and into action, by simply accepting suggestions made by StreamLab. All of the execution takes place on s-Server, at massive scale, and deployment is immediate. With StreamLab, business analysts or Java, SQL, Python, or Scala developers can efficiently move from concept to real working application in minutes.View Demo
s-Server: Real-Time, Complex Event Processing
Amazon Kinesis Analytics is based in part on certain technology components licensed from SQLstream: s-Server. s-Server is the only stream processor to use ANSI standards-compliant SQL for querying live data streams. Processing data at over 1M events/second per CPU core on benchmarks and at over 1M events/second per server on real-world applications, s-Server is the most powerful and scalable data processing platform available. s-Server includes a suite of agents and adapters for machine data mediation and enterprise system/storage platform integration.
s-Dashboard is an HTML5 platform for building push-based, real-time visualizations on s-Server. s-Dashboard enables developers to deploy real-time dashboards for streaming analytics on any device or channel, and is pre-configured with a standard set of panels, widgets, and dashboards.
s-Studio is a stream inspection, application development, and administration console for the s-Server platform. s-Studio enables dynamic updates to live applications, adding new queries as needed or changing existing queries or views. s-Studio is built as a plug-in to the standards-compliant Eclipse framework.
SQLstream’s approach to streaming minimizes the time developers spend writing low-level code. SQLstream Blaze automatically enforces data integrity while optimizing an application’s latency and performance, simplifies application maintenance, and streamlines the integration of existing code. At the heart of SQLstream Blaze is a rich and flexible set of operators to efficiently process and analyze incoming data streams.
An Ecosystem With Broad Integration And Extensibility
Work with any data source, format or destination, concurrently, continuously, and in real-time.
Full Support for Java Functions
SQL UDXes (User Defined Transform) enable Java development of streaming operations, and inclusion of third party libraries and even entire systems in a streaming SQL query.
Auto-Optimized Time Window Operators
SQLstream Blaze implements declarative windowing meaning that windows describe the scope of the tuples taking part in windowed operations, rather than CEP-approaches that normally view windows as data structures.
Deterministic Processing for Data Accuracy and Application Integrity
SQLstream Blaze is a declarative platform based on industry standard SQL data processing semantics – the language of business data – to assure existing business rules are supported with ease and to eliminate the risks in procedural language platforms.
Enrichment of Streaming Data with Any Source
SQLstream Blaze can perform a SQL JOIN between a stream and native table, external table (for example, Oracle), or re-streamed data from a native table without compromising performance.
Full-Spectrum of Support for Proper Event Sequencing
SQLstream Blaze has native, continuous incremental, and deterministic support for missed events and lack of events while preserving order of events and enabling action on delayed events, useful for any application where record-by-record correctness for monitoring is critical such as network monitoring, security, and financial.
SQLstream Blaze continuously computes telemetry data that quantifies behavior and performance. Enabling businesses to leverage existing technology investments, this telemetry is available as a table or a stream for easy integration with any third-party monitoring tool.
Low Latency + High Performance at Scale
SQLstream Blaze provides extreme performance and throughput without requiring tuning or low-level coding.
Hardware Optimization – 3x-500x less
SQLstream Blaze can process over 1 Million 16-byte records/sec per CPU core while scaling linearly to over 16 cores on commodity hardware.
Probabilistic Model-Driven Recovery for High Availability
By redundantly computing the results of all stream operations on multiple commodity nodes, SQLstream Blaze achieves arbitrary “x9”s of availability in real-world apps without compromising performance or resorting to exotic and costly hardware.