SQL SUM Double Counting in One-to-Many JOINs
A reproducible SQL example shows why one-to-many JOINs inflate totals and how to fix double counting without hiding valid payments.
Peng
I’m Peng, a Java backend developer and the author of TechShare. My project experience includes online classroom systems and student fitness data platforms. I write about Java, distributed systems, databases, and AI engineering, focusing on implementation decisions, troubleshooting, and practical trade-offs. Articles distinguish firsthand project experience from illustrative examples and explanations based on technical documentation.
Sentinel Resilience: Limits, Circuit Breakers, and Fallbacks
A practical Java guide to containing slow dependencies with rate limits, concurrency caps, timeouts, Sentinel circuit breakers, and business-safe fallbacks.
OpenFeign in Java: Timeouts, Retries, and Idempotent Order APIs
Design OpenFeign order APIs with timeouts, bounded retries, idempotency keys, fallbacks, and connection budgets. Includes reproducible failure tests.
Distributed Workflow Ordering: Why Event Timestamps Fail
See how clock skew breaks event ordering, then compare fixed clocks, monotonic timing, Lamport clocks, and order versions in Java.
Reliable Java Messaging: Outbox, Idempotency, Recovery
Build a reliable Java payment-to-reward workflow with a transactional outbox, idempotent consumers, stable reward keys, refund handling, and reconciliation tests.
Distributed Locks in Java: Redis, Redlock, ZooKeeper, and Fencing
A practical guide to Java distributed locks: protect inventory with database invariants, use Redis leases safely, fence stale workers, and test Redlock and ZooKeeper failure cases.
Distributed Transactions in Java: Outbox, TCC, and Saga Compared
A Java order workflow compares local transactions, the transactional outbox, TCC, and Saga using concrete failure paths for inventory, payment, rewards, and refunds.
Redis Caching in Java: TTL, Invalidation, and Consistency Tests
Debug Redis cache consistency in a Java service with cache-aside reads, TTL, invalidation, stale data, and database failures. Includes repeatable failure tests.
MySQL and ClickHouse Consistency: Outbox and Recovery
Keep MySQL and ClickHouse aligned with an outbox, versioned writes, reconciliation queries, and explicit recovery for delayed or duplicate events.
MySQL to ClickHouse: Scaling a Student Fitness Platform
A student fitness platform evolves from MySQL writes to ClickHouse analytics with bounded imports, outbox delivery, and query-oriented views.