# Official certification doc - [Professional Cloud Architect official page](https://cloud.google.com/certification/cloud-architect) - [Cloud Architecture Center](https://cloud.google.com/architecture/) ## Scenarios - [EHR Healthcare](https://services.google.com/fh/files/blogs/master_case_study_ehr_healthcare.pdf) - [Helicopter Racing League](https://services.google.com/fh/files/blogs/master_case_study_helicopter_racing_league.pdf) - [Mountkirk Games](https://services.google.com/fh/files/blogs/master_case_study_mountkirk_games.pdf) - [TerramEarth](https://services.google.com/fh/files/blogs/master_case_study_terramearth.pdf) # Other info - I use the course _Google Cloud Professional Architect: Get Certified 2021_ on Udemy by [Dan Sullivan](https://udemy.com/user/dan-sullivan-3/) - I also use the [The Google Cloud Developer's Visual Notes](https://github.com/priyankavergadia/GCPSketchnote) # Products to know - [Cloud resource hierarchy](https://cloud.google.com/resource-manager/docs/cloud-platform-resource-hierarchy) ## Security - [[Identity and Access Management]] - [[Web Security Scanner]] ## Service communication - [[Cloud Pub_Sub]] - queue ## Running services - [[Google Cloud Compute Engine]] - VMs - [[Google Kubernetes Engine]] - [[Istio]] - [[Anthos]] - [[Anthos#Cloud Run|Cloud Run]] - [[App Engine]] - [[Cloud Functions]] ## Storage ![[options to move data in google cloud.png]] - [[Cloud Storage]] - [[Cloud Memorystore]] ## Monitoring and sysadmin - Google Operations Suite (formerly Stackdriver) - [[Deployment manager]] - [[Cloud Shell]] - [[Transfer Appliance]] - [[Cloud Logging]] for events - [[Cloud Monitoring]] for timeseries data. ### The four golden signals of monitoring 1. Latency - time to complete a request. Differentiate between latency for successful calls and failed calls. 2. Traffic - demand on a system. 3. Errors - rates of errors. Differentiate between failed response and wrong response 4. Saturation - measure of the capacity in use. Usually you focus on the most contested resources. For example, CPU, disk, memory. ## Network - [[Network and VPC]] ## Data pipeline - A nice blog post to differenciate the different products: [Dataproc vs. Dataflow vs. Dataprep: What is the difference?](https://wisdomplexus.com/blogs/dataproc-vs-dataflow-vs-dataprep/) - [[Cloud Dataflow]] - Robust Batch with a graph. Can work at row level - [[Cloud Data Fusion]] - GUI based ETL. No code and simpler. - [[Cloud Dataproc]] - Spark Hadoop - [[Dataprep by Trifacta]] - explore new dataset - [[Vertex AI]] - ML/AI tools - [[Datastudio]] ## Databases - [[Cloud SQL]] - relational - [[Cloud Spanner]] 30TB+ SQL - [[Big Table]] - wide column - [[Cloud Firestore]] - document - [[BigQuery]] - analytics - [[Cloud Datastore]] ## Orchestration - [[Cloud Composer]] - Apache Airflow using DAGs in Python - [[Cloud Workflows]] - [Spinnaker](https://spinnaker.io/docs/concepts/) - [[Cloud Tasks]] ## Security - [[Data Loss Prevention 1]] - [[Firewall]] - Data at rest is always encrypted - Data in transit inside the Google network is authenticated (using ALTS) but not encrypted. Data going on the Internet is encrypted. - Encryption keys ![[Encryption keys options in Google Cloud.png]] - VPC service controls manage the flow between perimeters, and are context aware. For example, limiting the transit of data outside of a safe zone ignoring the identity (user) permissions. - on-prem can be a perimeter