Tags → #nats
-
Building a GPU SaaS Platform - Reliable Serverless Execution
Part 21: add queue retry policy, dead-letter handling, execution state classification, and runtime behavior changes for serverless GPU execution.
-
Building a GPU SaaS Platform - Invocation Result Store
Part 17: add a ScyllaDB-backed result-store consumer for durable serverless invocation metadata and control-plane result lookup.
-
Building a GPU SaaS Platform - Worker Lifecycle
Part 16: add activator-owned worker lifecycle management, prewarm pools, idle scale-down, and metrics-driven worker state.
-
Building a GPU SaaS Platform - Worker Sidecar
Part 15: add a serverless worker sidecar, define a UDS framework contract, and return results and metrics through NATS.
-
Building a GPU SaaS Platform - Activator Dispatch
Part 14: add a dedicated activator that consumes ingress invocations, selects or creates GPUUnit workers, and publishes worker-targeted dispatch messages.
-
Building a GPU SaaS Platform - Queue-First Ingress
Part 13: record the runtime-side serverless contract on GPU units and enqueue invocations durably through NATS JetStream before any worker executes them.