Develocity Performance & Scalability

Raw build data is consumed and acknowledged by the backend service immediately so that even under high load, transmitting Build Scan data is consistently fast.

Develocity native support for Kubernetes allows each microservice to scale independently to maintain optimal performance. Configuring horizontal scaling provides full control over the number of replicas dedicated to each microservice.

High traffic pages within Develocity are backed by materialized views in the Postgres database, allowing for quick load times and fast searches across the most common data attributes of Build Scans.

For any given project, datasets don’t change significantly across multiple builds. By analyzing and segregating data that tends to change versus data that doesn’t, the size of data sets per build can be minimized.

Support for intelligent replication allows remote cache nodes to be distributed geographically and located in low-latency network proximity to developers working around anywhere in the world.

Develocity integrates sophisticated caching and preemptive parallel computation techniques to make querying build analytics data faster, without relying on external data processing systems like Spark or Hadoop.

Build Scan data is intelligently queued for processing and analysis. Users who are actively debugging builds will receive higher priority data processing, delivering Build Scan results quickly when they’re needed most.

Post-optimization, any data sent across the wire to the backend service is compressed before uploading, reducing the overall amount of bandwidth needed and processing latency.