Translating current and future business needs into an architecture
This objective tests your ability to translate business requirements, current and anticipated, into a Google Cloud data architecture that balances scalability, cost, latency, and governance. You must justify service choices (e.g., BigQuery vs. Bigtable, Dataflow vs. Dataproc, Pub/Sub vs. Cloud Storage) based on data volume, velocity, variety, and access patterns rather than familiarity alone.
1 · Learn the must-know
- Distinguish batch vs. streaming needs early: Dataflow (Apache Beam) unifies both, while Dataproc suits existing Spark/Hadoop workloads needing lift-and-shift with minimal refactoring.
- For analytics at scale with SQL access patterns, BigQuery is the default choice due to serverless scaling and separation of storage/compute, but Bigtable or Spanner are better for high-throughput OLTP or NoSQL key-value access.
- Future scalability requirements should push you toward serverless, auto-scaling services (BigQuery, Dataflow, Pub/Sub) over fixed-cluster solutions (Dataproc) unless there's a strong reason for cluster control.
- Cost and performance trade-offs matter: BigQuery slot reservations vs. on-demand pricing, or Bigtable node count, should be sized against anticipated growth, not just current load.
- Data governance requirements (data residency, lineage, access control) often dictate architecture choices, e.g., using Dataplex, Data Catalog, VPC Service Controls, or CMEK, and must be mapped explicitly to compliance needs stated in the scenario.
- A common exam gotcha: don't over-engineer for hypothetical future scale if the stated requirements only justify a simpler service; conversely, don't choose a service that can't scale if growth is explicitly mentioned as a future requirement.
2 · Check your understanding
A logistics company currently loads shipment records into BigQuery through a nightly batch pipeline orchestrated by Cloud Composer, feeding a set of daily operational reports. The business now wants a live shipment-tracking dashboard within the next two quarters, and the existing daily reports must keep working without interruption during the transition. Which architecture best satisfies both the current and the new requirement?
What you have tried across GCP PDE's objectives, not a readiness score.
Designing data processing systems~22% of the exam0 of 17 tried
Ingesting and processing the data~25% of the exam0 of 11 tried
Storing the data~20% of the exam0 of 13 tried
Preparing and using data for analysis~15% of the exam0 of 11 tried
Maintaining and automating data workloads~18% of the exam0 of 15 tried
3 · Keep going
Ready for more? Take a weighted mock or try free practice questions.