Solution — Case Studies (GCP Professional Cloud Architect)
“Excerpt from https://www.udemy.com/course/gcp-architect-am/?couponCode=GCP-OCT”
During the exam for the Cloud Architect Certification, some of the questions may refer you to a case study that describes a fictitious business and solution concept to provide additional context to exam questions. There are four case studies as described below. The solution is narrated in a youtube video linked below.
EHR Healthcare is a leading provider of electronic health record software to the medical industry. EHR Healthcare provides their software as a service to multi-national medical offices, hospitals, and insurance providers.
Problem Statement:- Cloud Architecture for on-prem to cloud migration solution & scale problem
Solution:
Infrastructure
- Google’s IaaS and PaaS solution for data canters, Global VPC
- Multi-regional replication for DR
- Hybrid Connectivity — Cloud Interconnect (99.99% uptime SLA), Cloud VPN
- Dedicated Interconnect for high performance connection between on-premises and GCP
Applications
- Kubernetes Deployment — GKE
- Cloud and On-premise container based environments management and integration — Anthos
- Future API based integration — Apigee
- Predictions — AI Platform
- Data Ingestion — Streaming (Pub/Sub), Batch (Cloud Storage)
- Process — Dataflow, Cloud Composer
Databases
- MySQL, MS SQL — Cloud SQL
- Redis — Cloud Memorystore
- MongoDB — Cloud Firestore
Monitoring
Cloud Monitoring — Alerts and notifications, Charts and dashboards
Cloud Logging — Automatically ingest audit and platform logs, manage retention and policies.
Continuous Deployment
- Use Terraform for Infrastructure as Code
- Use Cloud Source Repositories for storing the source code
- Use Cloud Build for deployment and orchestration.
- Use Artifact Registry for container images
2. Helicopter Racing League- https://services.google.com/fh/files/blogs/master_case_study_helicopter_racing_league.pdf
Helicopter Racing League (HRL) is a global sports league for competitive helicopter racing. Each year HRL holds the world championship and several regional league competitions where teams compete to earn a spot in the world championship. HRL offers a paid service to stream the races all over the world with live telemetry and predictions throughout each race.
Problem Statement:-Cloud AI & ML, telemetry and streaming problem
Solution:
Transcoding
- Use Preemptible instance for VM based encoding solution
- Containerise the encoding solution and manage using Kubernetes engine.
TV Box Telemetry
- App Engine, Pub/Sub, Dataflow, BigQuery, Cloud Composer and Cloud Monitoring increase telemetry and create additional insights.
Live Video Latency
- Use HA configuration of Cloud VPN for connectivity between mobile data centers and google cloud.
- Use Cloud CDN for delivering content with speed, efficiency and reliability closer to the users.
- Use Cloud Storage with multi-regional buckets to serve contents.
- Use PerfKit Benchmarker to get visibility into metrics like latency, throughput, and jitter.
- Google Cloud offers Network Intelligence Center for comprehensive and proactive monitoring, troubleshooting, and optimization capabilities
AI & ML
- Use AI platform for predictive efficiency
- Use TensorFlow Deep Learning VM instances
Analytics
- Use BigQuery as a data mart for processing of large volume of data.
- Use Looker for embedded analytics.
- Big Query streaming API and ML solution can create additional insights for increasing fan engagements.
3. TerramEarth — https://services.google.com/fh/files/blogs/master_case_study_terramearth.pdf
TerramEarth manufactures heavy equipment for the mining and agricultural industries.. They currently have over 500 dealers and service centers in 100 countries. There are 2 million TerramEarth vehicles in operation currently, and they see 20% yearly growth. Their mission is to build products that make their customers more productive.
Problem Statement:- Cloud Automation, Operations and API Ecosystem problem
Solution:
Data Replication:
- Stream critical data from vehicles to Cloud Bigtable to drive analytics in real time.
- Sensor Device -> HTTPS Gateway Device -> Pub/Sub -> Dataflow -> Bigtable -> GKE (Application & Presentation)
- BigQuery partitioning by timestamp for Home base upload and unified analytics.
Data processing
- Cloud Dataflow for serverless unified (batch & stream) ETL
ML Engine or AutoML Tables
- Use Vertex AI for ML lifecycle to forecast anticipated stock needs to assist with just-in-time repairs.
Vehicles — Home Base Connected
Device management & upload
- Cloud IoT Core
- IoT devices -> Cloud Pub/Sub
- Cloud Dataflow -> Cloud Storage
Cloud Operations:
- Managed and Serverless services
- Network Intelligence Center for for monitoring, verification and optimization.
- Network Connectivity Center & Security Command Center for holistic security view.
- Cloud Monitoring for real time visibility
- Google Cloud KMS for Key Management
API’s ecosystem:
- Apigee (X) to manage and monitor APIs, it create an abstraction layer to connect to different interfaces.
- Apigee Developer Portal lets you build a self service portal for internal and external developers.
- Build and Deploy API’s on Google Kubernetes Engine
CICD:
- Use Cloud Source Repository, Artifact Repository, Cloud Build for CICD operations.
- Use Spinnaker to deploy on Kubernetes with Blue Green and Canary deployments.
Remote Workforce:
- G Suite with integrated Cloud IAM.
- Cloud Data Loss Prevention for sensitive data protection.
- Use Connected sheets with BigQuery to collaborate with integrated security controls.
4. Mountkirk Games — https://services.google.com/fh/files/blogs/master_case_study_mountkirk_games.pdf
Mountkirk Games makes online, session-based, multiplayer games for mobile platforms. They have recently started expanding to other platforms after successfully migrating their on-premises environments to Google Cloud. Mountkirk Games is building a new multiplayer game that they expect to be very popular.
Problem Statement:- Cloud Auto-scaling, Gaming analytics Problem
Solution:
- Cloud Spanner with relational features, horizontal scaling and 99.999% availability across regions.
- Google Kubernetes Engine(GKE) to deploy game’s backend as microservices
- Google Load Balancing for worldwide seamless autoscaling
- Pub/Sub, Dataflow and BigQuery for Stream analytics
- Looker for player insights and analytics
- GPUs hardware accelerators on GKE.
- Cloud Datastore for transactional game state
- Cloud Storage for storing game activity logs and analysed using BigQuery
- Cloud Pub/Sub for buffering of live and late data
- Cloud Dataflow for bulk and stream processing
- BigQuery for storage and analytics; this can also contain the 10 TB historic data
- Managed and Serverless services for dynamic scaling, minimal cost and operations.
- Cloud Operations metrics and APM functionality for proactive troubleshooting.
If you are interested in a classroom style tutorial of GCP services along with architectural framework and best practices. Check out this course : https://www.udemy.com/course/gcp-architect-am/?couponCode=GCP-OCT