Architecture overview¶
This page introduces the Architecture section, which explains how KX Sensors is structured — from its layered, extensible design to the core processes that ingest, store, and serve data.
KX Sensors uses a layered, extensible architecture built around a core set of processes for high-volume ingestion, low-latency query, and historical storage. Understanding how these pieces fit together helps you configure, develop against, and operate the system with confidence.
Architecture concepts¶
How the application and its configuration are structured.
- Application architecture — a high-level view of how data flows through the system
- Layered architecture — the layered, extensible design of KX Sensors
- KXS repositories — the repositories that make up KX Sensors
- About systemd — how KX Sensors integrates with systemd
- KXS directory structure — where files and directories live on disk
- Configuration objects — how configuration is represented
- Overlay rules for config objects — how configuration values are merged and resolved
Key processes and service classes¶
The core processes that make up a running deployment.
- Key processes — the core KX Sensors processes and what they do
- About RT — the reliable transport layer
- About TP+ — the tickerplant and its role in the data flow
- DBW/Z database writer and EOX processing — writing data to disk and end-of-period processing
- Parallel EOD/EOI processing — parallelized end-of-day and end-of-interval processing
- Task scheduler — scheduling recurring work
Where to next¶
- Develop — customize schemas, APIs, and ingestion
- Configure and Deploy — configure, tune, and deploy for your environment
- Operate — run and monitor a live deployment