This article examines the massive 2015 WeChat Red Packet traffic, detailing its core functions, system challenges, cross‑region networking, lossy service design, set‑model construction, concurrency handling, and data‑consistency strategies that kept the platform stable under extreme load.
This article explains the challenges of maintaining data consistency across multiple services in distributed e‑commerce architectures and presents six practical solutions—including business integration, the eBay BASE pattern, Qunar's approach, Mogujie's design, Alipay's DTS, and Nongxin's scheme—highlighting their advantages, drawbacks, and implementation details.
This article describes the design, architecture, and key features of a Linux‑based distributed file system that emphasizes horizontal scalability, data consistency, fault tolerance, and efficient handling of both small and large files without traditional metadata services.
Data ConsistencySynchronizationdistributed file system
The article explains how MySQL replication parameters such as expire_logs_days, relay-log-recovery, and relay-log-info-repository affect binlog cleanup, SQL and I/O thread consistency, and why crashes can produce duplicate‑key errors, then offers configuration fixes including the critical relay‑log cleaning option and the super_read_only setting.
The article explains how multi‑datacenter architectures prevent total service loss, improve latency by placing services near users, and balance the CAP trade‑offs through models like AC, CP, and AP, while outlining practical design, sharding, monitoring, and failover strategies for large‑scale backend systems.
ZooKeeper achieves strong data consistency across its server cluster by implementing the Paxos consensus algorithm, where clients propose changes, servers vote on proposals with monotonically increasing IDs, and a leader coordinates writes to ensure only one operation is committed at a time.
Data ConsistencyDistributed ConsensusLeader Election
This article explains the challenges of ensuring data consistency across multiple databases during operations like Alipay transfers, introduces local and distributed transactions, details the two‑phase commit protocol, and shows how reliable messaging can replace heavyweight distributed transactions for high‑concurrency systems.
The article explains how to maintain data consistency when updating related tables by using local transactions for single‑node databases, distributed two‑phase commit for multi‑node systems, and reliable message‑queue patterns—including deduplication techniques—to avoid the performance pitfalls of traditional distributed transactions.
This article explains how to migrate live service data between systems without downtime, covering migration types, a four‑stage process, practical examples with MySQL and HBase, and key techniques for ensuring consistency and smooth cut‑over.
To prevent data inconsistencies when transferring funds—like moving money from Alipay to Yu'ebao—the article explores local transactions, the limitations of two-phase commit in distributed systems, and how reliable message queues can achieve eventual consistency without sacrificing performance.
This article explores how to maintain data consistency across services by using local transactions, two‑phase commit distributed transactions, and reliable message‑queue patterns, illustrating each method with an Alipay‑to‑YuEBao transfer example and discussing their trade‑offs.
The article examines the challenges of scaling a single‑server data service, compares data partitioning and replication, explains consistency models, and surveys distributed transaction protocols such as 2PC, 3PC, Paxos, and Dynamo's NWR model, highlighting their trade‑offs in availability, consistency, and performance.