What is GCP architecture?

Google cloud platform (GCP) provides many services to fulfill your cloud computing needs. These include infrastructure and platform services that help businesses develop, deploy and scale their cloud applications.

The GCP architecture encompasses how the various components of the cloud are interconnected within the application. The architecture includes the following elements:

• How the cloud is accessed (client or device)

• Backend platform

• How the services are delivered

GCP architecture pillars

• System design: This forms the basis of the GCP architecture. It involves understanding the structure, modules, interfaces, and data needed to fulfill the requirements of cloud systems.

• Operational excellence: This includes the deployment and management of cloud workloads efficiently.

• Cost optimization: Controls your costs and improves your business value.

• Security: Ensures solid security and privacy of your application data workloads by enhancing security controls and adhering to requirements.

• Reliability: Builds resilient and highly available workloads to prevent downtime as much as possible.

• Performance optimization: Achieves your cloud application's optimal performance and speed by adjusting your cloud resources.

GCP architecture diagram

A GCP architecture diagram gives a high-level view of your architecture and determines how the different cloud components will interact. It is understandable by both the developers and stakeholders. The architecture diagram is beneficial for documenting plans, changes, and troubleshooting.

Example

Here is a sample architecture diagram for building a streaming pipeline to analyze video.

Explanation

1. Client uploads source video files to Cloud storage bucket.

2. The client is automatically notified through Pub/SubPublisher/Subscriber is a messaging service provided by GCP. It allows decoupled and asynchronous communication between independent applications or components within a distributed system. when a file is uploaded.

3. Dataflow pipeline:

   1. Reads file metadata

   2. Loads file into memory

   3. Sends its segments to video intelligence API

   4. Stores annotations in BigQuery

4. Pub/Sub sends the output message

5. Consumer application displays the output message.