0 views collected around this technical thread.
The article recounts a McDonald's China backend interview, explaining IoC/DI concepts, MyBatis benefits, JVM garbage‑collection and memory layout, MySQL engine differences, concurrency locking strategies, RPC advantages over HTTP, HTTPS basics, essential Docker/Kubernetes commands, and a typical Git workflow.
China Railway’s 12306 trial of an automatic ticket‑ordering feature lets users pre‑select up to 60 trains, set priority and pre‑pay the highest fare for Spring Festival travel, with the system submitting orders automatically, handling refunds, and addressing priority sorting, scheduling, payment accuracy, and server load through extensive stress testing and a gray‑release strategy.
The article examines five common rate‑limiting algorithms—Fixed Window, Sliding Window, Sliding Log, Leaky Bucket, and Token Bucket—detailing their principles, pros and cons, and providing complete C++ implementations to help developers choose the best approach for controlling traffic bursts and ensuring system stability.
Reflecting on his 2023 technical writings, the author—prompted by the TV series 'Reply 1988'—structures his insights into chapters covering Domain‑Driven Design lessons illustrated by a Forrest Gump metaphor, business‑architecture note‑taking, judgment models drawn from everyday driving, and personal projects like the VSEF framework and the “曲径通幽” series, emphasizing curiosity, knowledge sharing, and continual growth.
The article details how Dewu’s new product onboarding workflow was refactored into a Spring StateMachine‑driven architecture that separates channel‑specific logic via a Strategy‑pattern operation layer, defines clear enum status domains, and uses guards, actions, and choice states to achieve maintainable, extensible, and idempotent state transitions.
The article surveys common global ID generation strategies for distributed systems—database‑managed approaches like MySQL auto‑increment and MyCat‑Zookeeper, and Java‑based methods such as UUID, Snowflake, Leaf, Redis INCRBY and direct Zookeeper—comparing their trade‑offs and recommending Leaf (segment or Snowflake mode) or MyCat‑Zookeeper for reliable, scalable identifiers.
QR‑code login lets a already‑authenticated mobile app scan a code shown on a PC, using a server‑issued token tied to the user’s account and device ID; the scan creates a temporary token, the user confirms, and a permanent token is issued to the PC, enabling password‑less, secure authentication.
The article describes a full‑platform automated test execution engine that replaces legacy offline‑script/Jenkins setups with a lightweight, visual, component‑based solution supporting low‑/zero‑code case creation, parallel execution, protected environment variables, dynamic placeholders, system‑method libraries, and extensible script and case workers for scalable, efficient testing.
The article explains VSEF’s three‑layer architecture—entry, kernel, and dependencies—detailing how evolutionary themes, component selection, and design patterns such as Template Method and Strategy enable a stable core while allowing flexible, domain‑specific extensions through both market and self‑built modules.
In a shared Elasticsearch cluster, the team used slow‑log analysis to pinpoint costly queries caused by unnecessary fuzzy matches and integer‑mapped low‑cardinality fields, then optimized them by converting matches to filters and remapping those fields to keyword, re‑indexing, which cut latency from over 100 ms to under 10 ms and eliminated slow‑query alerts.
The article describes Vivo’s overseas sales team’s multi‑language document system, detailing its low‑maintenance i18n design that leverages the existing configuration center, a MCMS UI, and a runtime JAR to streamline project creation, translation, testing, publishing, and real‑time front‑end text updates across web and decoupled architectures.
The article describes a three‑layer community advertising placement system for Xianyu, detailing its four core processes (recruitment, selection, construction, delivery), resource slots, material templates, task definition, data storage, real‑time monitoring, and successful case studies, while outlining future algorithmic enhancements.
HermesX unifies Xianyu’s fragmented push pipelines into a single flow, optimizes content relevance, applies a four‑layer fatigue model, and offers a low‑code management UI, delivering up to 35% less abnormal traffic, 65% lower compute use, and 69% reduced peak latency while paving the way for richer content pools and finer user segmentation.
The article introduces load balancing as a vital solution for high‑traffic systems, explains vertical and horizontal scaling, classifies balancers by hardware, layer and algorithm, and details common algorithms—random, weighted, round‑robin, least‑active, IP‑hash, and consistent hash—with code examples and usage guidance.
The paper presents a cluster‑aware Guava cache component for Alibaba’s Xianyu platform that mitigates downstream service failures by adding asynchronous reload, cluster‑wide key invalidation, and size reporting, enabling automatic fallback to refreshed local data and improving latency, with future plans for a management console, tiered storage, and disk‑backed caching.
The article details how Xianyu’s 2020 messaging failures—lost messages, wrong avatars, and order status errors—were traced to duplicate IDs, push‑logic mismatches, and client bugs, and solved by introducing global UUIDs, ACK‑based retries, hierarchical conversation models, hybrid storage caching, and real‑time monitoring, boosting delivery reliability above 99.9%.