InterviewStack.io LogoInterviewStack.io

Cloud Data Architecture and Tradeoffs Questions

Designing data architectures specifically for cloud environments and evaluating platform trade offs. Topics include when to use managed relational services, managed nonrelational services, cloud data warehouses, cloud object storage, lifecycle policies, cross region replication, data residency and compliance considerations, cost versus performance trade offs, managed service operational constraints, and strategies for high availability and disaster recovery in the cloud. Candidates should be able to compare cloud service options and justify choices based on reliability, cost, and compliance.

HardSystem Design
63 practiced
Design a globally distributed analytics platform that must serve low-latency queries to users in multiple continents and support analytic ingestion in regional data centers. Define how you would handle data replication, global vs local views, consistency models, query routing, metadata management, and the tradeoffs between multi-region writes and read-latency. Provide component-level choices and justify tradeoffs.
HardTechnical
55 practiced
Create an incident runbook to troubleshoot late-arriving data for a streaming pipeline with a 30-minute SLA for dashboard freshness. The runbook should include detection thresholds, immediate mitigations (replay, backfill, degrade modes), stakeholders to notify, steps for root cause isolation, how to perform safe backfills, and communication templates for downstream consumers.
HardSystem Design
44 practiced
Design a regular disaster recovery test plan for a cloud data warehouse and associated data lake. The plan must include pre-checks, simulation of failover, data validation steps, communication runbooks, timeboxed tests, and rollback procedures. Explain how you minimize risk to production during testing and how you capture evidence for compliance.
EasyTechnical
88 practiced
Write an SQL upsert (insert-or-update) in PostgreSQL to move deduplicated records from a staging table into a fact table. Use the following schema in your example:
staging_events(staging_id serial, event_id text, user_id int, event_ts timestamp, payload jsonb)fact_events(event_id text primary key, user_id int, event_ts timestamp, payload jsonb, updated_at timestamp)
Show an upsert that keeps the newest event_ts per event_id and updates payload and updated_at accordingly. Assume Postgres on AWS RDS.
HardSystem Design
53 practiced
Design an archive and cold-analytics architecture that stores petabytes of historical data in object storage and allows analysts to run infrequent ad-hoc queries without spinning large clusters. Discuss choices like serverless query engines (Athena, BigQuery), partition pruning, compaction, caching popular results, and cost versus query latency tradeoffs.

Unlock Full Question Bank

Get access to hundreds of Cloud Data Architecture and Tradeoffs interview questions and detailed answers.

Sign in to Continue

Join thousands of developers preparing for their dream job.