Introduction to a Lightweight Tool for Data Consistency in RPC Services During Network Timeouts

This article discusses a technical solution for maintaining data consistency in distributed systems during network timeouts. It introduces a lightweight tool that mimics successful service calls (MOCK/SPY) to bypass failed or timed-out requests, ensuring application flow continuity without real data processing.

The tool operates by intercepting method calls based on configurable rules (e.g., time ranges, method names) and returning predefined responses. It supports both mocking (DO_NOTHING_AND_RETURN_SPECIFIED_VALUE) and spying (DO_AND_RETURN_SUCCESS_REGARDLESS_OF_FAILURE) strategies. Code examples demonstrate configuration via JSON or Spring YAML, with SpEL expressions for dynamic parameter mapping.

Use cases include preventing abnormal data generation in inventory systems and accelerating fault recovery in distributed transactions. The tool is implemented as an aspect-oriented Java library with Maven dependencies, requiring annotation on target methods and optional DUCC configuration for runtime control.

Key features include dynamic configuration, precise scope definition to avoid unintended mocking, and support for both read and write operations. The article emphasizes practical implementation over theoretical solutions, offering a configurable framework for similar scenarios.