An open-source real-time stream processing framework combining high-throughput event processing with low-latency SQL-like streaming queries.
Tigon is an open-source real-time stream processing framework that combines high-throughput event processing with low-latency SQL-like streaming capabilities. It solves the challenge of handling diverse real-time data requirements by providing both scalable persistent processing and flexible in-memory stream processing in a unified framework. The project emerged from a collaboration between Cask Data and AT&T to create a comprehensive solution for modern streaming applications.
Data engineers and developers building real-time streaming applications on Hadoop ecosystems who need both high-throughput processing and low-latency analytics. Organizations with large-scale data flows requiring exactly-once processing semantics and SQL-like query capabilities.
Developers choose Tigon because it uniquely combines two complementary approaches: Cask Data's reliable high-throughput processing and AT&T's flexible low-latency streaming engine. This integration provides a complete solution that handles diverse streaming requirements without forcing compromises between throughput and latency.
High Throughput Real-time Stream Processing Framework
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Provides reliable event processing with exactly-once semantics using Java APIs, ensuring data consistency in high-throughput scenarios as highlighted in the README.
Features a streaming SQL engine that allows filtering, grouping, and joining data streams in-memory, simplifying complex stream logic for real-time analytics.
Runs natively as a YARN application and integrates tightly with HDFS and HBase, making it ideal for existing Hadoop deployments, as emphasized in the key features.
Offers fault-tolerant architecture with horizontal scalability and built-in debugging, logging, and monitoring tools for production use, ensuring robust operation.
Only supported on *NIX systems like Linux and Mac OS X, with no Windows support, limiting deployment options and accessibility for some teams.
Requires multiple prerequisites including JDK, GCC, G++, and Maven, and involves building from source or managing YARN clusters, adding significant initial overhead.
Heavily dependent on Hadoop ecosystem components, which can be a barrier for teams using alternative data platforms or modern cloud services, reducing flexibility.
The last major release appears to be version 0.2.1 from 2014, and documentation links might not be current, raising concerns about active maintenance and community support.