GeCached

GeCached is a caching system designed to accelerate delivery of geospatial data by storing frequently requested geographic responses. It targets tile requests, coordinate and feature queries, and raster or vector data delivered by map servers and GIS applications. The goal is lower latency and reduced load on the origin data store.

GeCached typically integrates with existing geospatial backends such as PostGIS, SpatiaLite, or geospatial data services. Clients request data via a defined API, and GeCached serves cached results when available, automatically refreshing entries based on configured TTLs or data-change events. It supports both vector and raster data paths and can operate in both on-premises and cloud environments.

Key features include configurable time-to-live and eviction policies, namespace isolation for multi-tenant deployments, and cache invalidation triggered by data updates or explicit purge operations. It uses spatially informed cache keys, often incorporating geohash, tile coordinates, or bounding box parameters to minimize cache misses and collisions.

In deployment, GeCached can function as a stateless service behind a load balancer, with a distributed cache layer such as Redis or Memcached, enabling horizontal scaling. It provides REST and gRPC interfaces for integration, plus client libraries in common programming languages.

While designed to improve performance, GeCached introduces considerations around cache staleness, data consistency, and cache warm-up times. Proper monitoring, metrics, and cache tuning are essential to balance freshness against latency and storage costs. The project aims for interoperability with standard GIS data formats and protocols.