▸case-16 When a customer cancels a booking, three separate aggregates—Reservation, Payment, and Notification—must execute sequential compensating steps. Should individual aggregates directly mutate each other's database records, or how should this cross-aggregate process be managed? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-03 We are building an internal lookup microservice for employee badges that performs simple create, read, update, and delete (CRUD) operations on a PostgreSQL table with 50 total records. We need strict immediate transaction consistency for single-row updates and have no audit log or temporal query requirements. Outline the recommended system architecture and data storage pattern. | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-20 Our mobile client app requires displaying a complex dashboard showing 30-day spending trends, top category breakdown, and loyalty points summary. Calculating this on demand by replaying events on every app launch is too slow. How should CQRS projection building address this read performance requirement? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-04 Our banking core transaction ledger mandates strict ACID immediate consistency across all balances synchronously in every database write, with absolute zero tolerance for eventual consistency. The underlying infrastructure cannot support event store operations or asynchronous background projections. Provide the architectural recommendation for this ledger service. | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-22 An enterprise inventory tracking system processes thousands of stock adjustments per item per day. What threshold or strategy should be defined to determine when an aggregate state snapshot should be recorded? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-19 Explain the standard architectural flow inside an event-sourced domain aggregate when processing an incoming user request to update an account balance. | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-07 A developer accidentally published an event with incorrect pricing data into our primary event store production database yesterday. They propose executing an SQL UPDATE or DELETE statement on the event store table to fix the corrupt payload. How should corrupt or bad data in an event store be corrected? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-06 We need to fix a bug in our customer analytics reporting query that reads from our event stream. Engineers want to drop and re-create the projection tables directly in our live production environment during low-traffic hours to get the updated metrics immediately. Provide an operational guide on how projection rebuilding should be deployed safely. | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-21 After a user submits an order, they immediately view their order list, but the new order is missing for 200 milliseconds because the read projection model update is asynchronous. How should the system handle this consistency model? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-08 We are designing a multi-step fulfillment saga that coordinates inventory reservation, payment processing, and shipment dispatch across three microservices. The process must survive server restarts and process crashes midway through execution without losing state. What execution model or framework pattern should be used for cross-aggregate saga orchestration? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-10 We are modifying the schema of our OrderPlaced event to break a full shippingAddress string into structured street, city, and postalCode fields. Developers suggest overwriting old event schemas in place or replacing old JSON payloads in the database. How should schema evolution and backwards compatibility be structured in event-sourced systems? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-02 Our team needs to redesign an enterprise healthcare audit service to move away from direct record updates toward recording every modification as historical facts. Provide a step-by-step implementation guide detailing domain aggregate boundaries, event handling practices, schema evolution handling, and projection maintenance rules. | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-11 Our event-driven e-commerce architecture triggers multiple downstream microservices and projections whenever a customer places an order. When investigating missing payments or delayed notifications, engineers struggle to trace which incoming request generated which background events. What field standard should be included in all event metadata for end-to-end distributed tracing? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-14 Our compliance team needs to run historical audits that query what a user's account balance and active permissions were at 2:15 PM on March 14th, 2023. Standard database tables only store current state. Describe how to structure the write model to support temporal time-travel queries. | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-05 We have a basic customer contact preference form on a content website that only needs to store the latest email address and notification checkbox. The system has minimal business logic and CRUD operations are completely sufficient. Should we implement an event store with read model projections, or stick to a standard database? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-12 Due to network retries in our message broker, projection workers occasionally receive the exact same PaymentReceived event twice. The current projection handler increments account balances every time it receives a message, causing double-counting. How should event handler methods be designed to handle duplicate deliveries? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-01 Design an event-driven architecture for managing order state transitions while keeping a complete historical record and separating write models from read models. Please provide an architectural specification detailing the boundary definitions, immutable record structures, read model generation, and multi-step workflow handling. | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-09 Our user account aggregate stream has accumulated over 50,000 historical events over 5 years. Replaying the entire event stream from event ID 1 every time a user logs in is causing severe latency spikes. What architectural mechanism should be added to keep aggregate hydration fast? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-13 An engineering team is designing a monolithic event named UserEverythingUpdated that bundles profile changes, password hashes, payment settings, notification preferences, and application telemetry into a single 500KB JSON object emitted every minute. Provide feedback on event design best practices. | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-15 A high-throughput food delivery app handles 100,000 search queries per second while processing 1,000 order creation commands per second. Combining read and write operations on the same relational database table is causing deadlocks. How should CQRS be applied to decouple these operational profiles? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-17 Our security team is writing governance rules for our database storage engine underlying the event store. What strict rule must be enforced regarding write, update, and delete privileges on committed event tables? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
▸case-18 A developer is writing event definitions for a banking system and proposes event names like CalculateInterestCommand and UpdateUserAddressIfValid. How should domain events be named and conceptualized compared to commands? | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |