Build production data pipelines on Google Cloud using Rust — predictable latency, single-digit-megabyte containers, and errors that fail at compile time instead of 3 a.m. This course shows engineers how to read from Cloud Storage, query BigQuery (REST jobs.query for small results, Storage Read API for million-row Arrow scans), and deploy distroless handlers to Cloud Run with sub-100 millisecond cold starts. You'll learn the gcloud CLI and Cloud Shell workflow, choose the right GCS client crate stack (google-cloud-storage, tonic, tokio), and configure Pub/Sub push subscriptions with idempotent content-hash handlers and backpressure controls. Production discipline comes through cargo-audit, cargo-deny, secure-by-design defaults, and CI gates on GitHub Actions. By the end, you'll have a working pattern for shipping a Rust ETL handler that survives at-least-once delivery, distroless image scans, and concurrent load — all on the GCP services you already pay for.
What you'll learn
Read data from Cloud Storage using Rust
Query BigQuery with Rust
Deploy distroless containers on Cloud Run
Configure Pub/Sub with idempotent content-hash handlers
Implement production discipline with cargo-audit and cargo-deny
Course objectives
Understand the workflow of gcloud CLI and Cloud Shell
Choose appropriate GCS client crate stack for Rust
Build an ETL handler that supports at-least-once delivery