Serverless GCP architectures include interactive or batch data processing. Also it's common to include machine learning evaluation and processing. Data Lake architectures are generally used when data volumes are genomic-sized. Below are examples and also links to more information about GCP libraries that are commonly used in this type of solution.
Of course you can build a serverless website using any number of patterns. I shown the three most common ones in the diagram below. I've also included other common GCP services used in web applications which provide management and/or automation capabilities.
- See section for Cloud Functions --> https://cloud.google.com/architecture/serverless-functions-blueprint
- See section for Cloud Run --> https://cloud.google.com/architecture/serverless-functions-blueprint
- Filter by service/scenario from Jump Start Solutions --> https://cloud.google.com/architecture/all-jss-guides
This example is for genomics (bioinformatics tool / library testing) with the ability scale out if needed for size of input data. More information here
This example is for genomics (bioinformatics data) with multiple pipeline types and is shown below.
- 📘 "Build a Data Lake on GCP" -- link
- 📘 "Serverless Analytics and Machine Learning with Google Cloud" -- link
- 🔨 About Knative -- link
- 🔨 KServe for serverless ML model serving -- link
