▸case-11 Designing a disaster recovery plan for a mission-critical MySQL database on AWS RDS requiring an RPO under 5 minutes and an RTO under 15 minutes. The team suggests taking daily full snapshots without binary logging to save on storage costs. Evaluate if this meets the requirements and detail the required backup configuration. | pass→pass | 17,173 | 16,189 | -6% | 1 | 1 | 0% | 3,175 | 4,784 | +51% | 0 | 0 | — |
▸case-12 An e-commerce MySQL query running SELECT * FROM orders WHERE status = 'PENDING' ORDER BY created_at DESC LIMIT 50 is taking 12 seconds on a table with 50 million rows. Developers created an index on status and another separate index on created_at. Explain why MySQL is still doing a filesort/full scan and specify the composite index definition that optimizes both filtering and sorting. | pass→pass | 12,960 | 11,181 | -14% | 1 | 1 | 0% | 2,057 | 3,959 | +92% | 0 | 0 | — |
▸case-01 We are migrating our legacy PostgreSQL database on-premises to AWS Aurora PostgreSQL and need an architecture runbook. Please outline a step-by-step migration and high-availability design document covering schema versioning with Liquibase, connection pooling via PgBouncer, automated failover setup, and backup validation procedures. Format the output with clear technical sections and actionable deployment checklists. | fail→pass | 33,067 | 25,665 | -22% | 1 | 1 | 0% | 6,214 | 6,038 | -3% | 0 | 0 | — |
▸case-02 I need a security and disaster recovery compliance playbook for a multi-region Azure Cosmos DB deployment supporting a healthcare app. Provide a structured guide detailing role-based access controls, at-rest and in-transit encryption setups, cross-region replication procedures to hit low RPO/RTO targets, and routine backup restoration testing steps. | fail→fail | 32,379 | 26,278 | -19% | 1 | 1 | 0% | 6,207 | 7,118 | +15% | 0 | 0 | — |
▸case-03 We are building a React frontend application that displays user profile cards. Provide the JSX component code and CSS styles for rendering a responsive profile card grid using CSS Grid. | pass→pass | 15,641 | 16,646 | +6% | 1 | 1 | 0% | 3,788 | 5,917 | +56% | 0 | 0 | — |
▸case-04 Write a Python function using the Pandas library to calculate the 30-day moving average and exponential moving average of stock closing prices from an in-memory DataFrame. | pass→pass | 8,382 | 8,339 | -1% | 1 | 1 | 0% | 1,727 | 3,673 | +113% | 0 | 0 | — |
▸case-05 Write a Node.js Express route handler that validates incoming JSON payloads using Joi schema validation and returns 400 Bad Request if validation fails. | pass→pass | 8,505 | 8,782 | +3% | 1 | 1 | 0% | 1,799 | 3,555 | +98% | 0 | 0 | — |
▸case-06 Our high-throughput PostgreSQL 15 database on AWS RDS is experiencing heavy table bloat and high disk I/O during peak hours due to default autovacuum settings. The team suggests turning off autovacuum entirely and running a single manual VACUUM FULL every midnight. Evaluate this approach and provide the recommended configuration parameter adjustments. | pass→pass | 19,013 | 18,938 | -0% | 1 | 1 | 0% | 3,339 | 5,371 | +61% | 0 | 0 | — |
▸case-07 We need to rename a heavily queried column user_email to email_address in a PostgreSQL database without taking down our application or causing read/write errors during deployment. The engineering team plans to run a single ALTER TABLE users RENAME COLUMN statement during business hours. Explain why this breaks live app instances and detail the multi-phase deployment pattern to achieve zero downtime. | pass→pass | 15,808 | 13,920 | -12% | 1 | 1 | 0% | 3,030 | 4,373 | +44% | 0 | 0 | — |
▸case-08 Our AWS Lambda microservices frequently exhaust PostgreSQL connection limits on RDS under spike traffic, throwing 'too many clients' errors. The team proposes increasing PostgreSQL max_connections from 500 to 10,000 to handle peak concurrency. Evaluate this proposal and recommend the correct connection management solution. | fail→pass | 19,527 | 16,093 | -18% | 1 | 1 | 0% | 2,697 | 4,755 | +76% | 0 | 0 | — |
▸case-09 We are deploying a MongoDB replica set across two data centers (DC1 has 2 nodes, DC2 has 2 nodes). In the event of a network partition between DC1 and DC2, how will primary election behave, and what configuration adjustment is required to prevent a split-brain scenario or inability to elect a primary? | pass→pass | 12,559 | 13,265 | +6% | 1 | 1 | 0% | 2,370 | 4,309 | +82% | 0 | 0 | — |
▸case-10 We need to automate the provisioning of an AWS RDS PostgreSQL primary instance in us-east-1 with a cross-region read replica in us-west-2 for disaster recovery. A developer proposed manually clicking through the AWS Management Console to create the replica. Specify the IaC approach using Terraform and identify the specific resource types required. | pass→pass | 11,604 | 17,853 | +54% | 1 | 1 | 0% | 2,439 | 5,569 | +128% | 0 | 0 | — |
▸case-13 We are designing a multi-tenant SaaS application sharing a single PostgreSQL schema. To enforce data isolation between tenants, developers suggested appending WHERE tenant_id = X in every application query manually. Recommend the native database security mechanism to enforce tenant isolation at the database layer regardless of application query flaws. | pass→pass | 9,636 | 10,976 | +14% | 1 | 1 | 0% | 1,679 | 3,927 | +134% | 0 | 0 | — |
▸case-14 We are building an in-memory session store needing auto-scaling beyond 100 GB RAM and high write throughput across multiple shards. A developer recommends deploying Redis Sentinel with 1 Master and 2 Replicas. Evaluate whether Redis Sentinel fulfills multi-shard horizontal write scaling and state the appropriate Redis deployment model. | pass→pass | 12,897 | 11,536 | -11% | 1 | 1 | 0% | 2,010 | 3,623 | +80% | 0 | 0 | — |
▸case-15 Our centralized log management Elasticsearch cluster is running out of fast SSD storage because 2-year-old log indices remain on primary hot nodes. The ops team plans to run a weekly manual script to delete old indices. Outline the native Elasticsearch feature and lifecycle phases to automate moving data from hot to warm/cold storage before deletion. | pass→pass | 12,785 | 11,361 | -11% | 1 | 1 | 0% | 1,981 | 4,049 | +104% | 0 | 0 | — |
▸case-16 We operate a cluster of AWS RDS PostgreSQL instances for a global 24/7 fintech service. Operators want to apply minor version engine upgrades manually during ad-hoc hours without testing. Describe the automated patching strategy and safety controls required to perform minor engine version updates without service disruption. | fail→fail | 20,982 | 14,709 | -30% | 1 | 1 | 0% | 3,206 | 4,047 | +26% | 0 | 0 | — |
▸case-17 We need real-time change data capture (CDC) from an operational SQL Server database to stream row changes into a data lake without placing heavy polling query load on the primary DB. Developers suggested running a continuous SQL script SELECT * FROM table WHERE updated_at > NOW() - INTERVAL 5 SECONDS every second. Explain why this approach fails under scale and specify the log-based CDC architecture. | pass→pass | 18,256 | 20,683 | +13% | 1 | 1 | 0% | 2,691 | 5,181 | +93% | 0 | 0 | — |
▸case-18 In an Apache Cassandra cluster deployed across 3 data centers with a Replication Factor of 3 per DC, write operations are failing or slow because the application uses LOCAL_QUORUM for reads and ALL for writes. Evaluate this consistency strategy for cross-data-center resilience and latency, and recommend the adjusted consistency levels for strong consistency with high availability. | pass→pass | 15,835 | 20,292 | +28% | 1 | 1 | 0% | 2,604 | 5,146 | +98% | 0 | 0 | — |
▸case-19 Our team manages schema updates by sharing raw .sql migration files over Slack and manually applying them to production servers. This caused schema drift and deployment failures. Detail how to integrate an automated, version-controlled schema migration tool into a CI/CD pipeline. | pass→pass | 20,156 | 19,012 | -6% | 1 | 1 | 0% | 3,204 | 4,940 | +54% | 0 | 0 | — |
▸case-20 We are migrating a distributed global application to Google Cloud Spanner. Developers defined the primary key of a high-write inserts table as an auto-incrementing integer sequence (INT64). Explain why sequential primary keys create severe performance bottlenecks in Cloud Spanner and specify the correct primary key design pattern. | pass→pass | 16,533 | 17,648 | +7% | 1 | 1 | 0% | 2,598 | 3,851 | +48% | 0 | 0 | — |
▸case-21 To comply with PCI-DSS requirements, our PostgreSQL production database needs complete audit logging for all schema modifications (DDL) and access permission changes (GRANT/REVOKE). The security team suggested enabling statement logging log_statement = 'all' for all queries including SELECT. Evaluate the operational impact and recommend the appropriate audit extension. | pass→pass | 14,660 | 22,855 | +56% | 1 | 1 | 0% | 2,644 | 5,476 | +107% | 0 | 0 | — |
▸case-22 Our AWS DynamoDB table has predictable, steady-state read and write workloads 24/7 (averaging 5,000 WCU and 10,000 RCU with minimal spikes). Currently, it runs on On-Demand capacity mode. Explain why On-Demand mode is financially inefficient for predictable workloads and specify the cost optimization steps. | pass→pass | 20,198 | 17,559 | -13% | 1 | 1 | 0% | 3,007 | 4,892 | +63% | 0 | 0 | — |