The article recounts Ant Financial’s journey from relying on commercial databases to building its own OceanBase system, detailing the strategic, technical, and architectural challenges of self‑developed databases, distributed middleware, unit‑based design, and multi‑city fault‑tolerant solutions.
This article examines the core technologies behind NewSQL databases for OLTP workloads, contrasting them with NoSQL and traditional relational systems by exploring storage engines such as B‑Tree and LSM‑Tree, sharding strategies, replication models, and distributed transaction mechanisms like two‑phase commit.
The article recounts the evolution from early laser typesetting to the design of OceanBase, a distributed database that balances performance, cost and consistency for financial services, detailing five key technical innovations, market forces, and the broader implications for the database industry.
AliSQL X‑Cluster is Alibaba's MySQL‑compatible distributed database that integrates the X‑Paxos consensus protocol to provide strong consistency, multi‑region deployment, low‑cost replica types, asynchronous transaction commit, hotspot‑update optimizations and superior performance compared with native MySQL and Group Replication, while offering flexible online configuration and robust failover mechanisms.
This article explains why relational database middleware is essential for scaling internet‑level workloads, describes the principles of horizontal sharding and distributed primary‑key generation, and provides a comprehensive overview of Sharding‑JDBC’s architecture, core modules, performance benchmarks, and future roadmap.
Pilosa, originally built for sparse high‑cardinality user attributes, is evaluated on a dense, low‑cardinality NYC taxi dataset to see if it can serve as a general‑purpose index, with performance comparisons against Spark, PostgreSQL, Elasticsearch, and kdb+ across multiple query scenarios.
This article provides a comprehensive overview of cloud computing evolution, covering virtualization, x86 servers, cloud management platforms, container technologies, orchestration tools, resource scheduling, microservices, and distributed databases, while also noting a conference discount promotion.
In this interview, Ma Jin, lead of NetEase's distributed database DDB, discusses the system's evolution, technical challenges, performance‑optimization practices, cloud‑native proxy deployment, and future prospects for distributed databases within large‑scale applications.
This article explains how PhxSQL prioritizes strong linearizable consistency, high availability, serializable isolation, and full MySQL compatibility, and why it deliberately forgoes features such as multi‑write, sharding, and strict serializable isolation due to the high cost of distributed transactions and protocol complexity.
This article provides a comprehensive overview of distributed database systems, covering their evolution from centralized architectures, classifications, key characteristics, advantages and disadvantages, data sharding methods, allocation strategies, architectural models, and management considerations, highlighting the benefits and challenges of distributed data management.
The article reviews eight expert presentations from the 2015 SDCC Database Practice Forum, covering distributed database architectures, performance tuning, high‑availability solutions, and practical case studies from leading Chinese internet companies.
This article explains a practical method for performing pagination on a sharded table by querying each database up to the target offset, merging and sorting the results, and extracting the required page while discussing performance drawbacks and an optimization when data distribution is uniform.
This article examines the technical challenges of massive web sites—such as billions of users, extreme concurrency, and petabyte‑scale data—and explains how architectural evolution, including service separation, caching, clustering, load balancing, CDN, distributed storage, NoSQL, and micro‑service decomposition, enables scalable, highly available, and secure operations.
Apache Ignite lets developers store hot data in memory, offering partitioned or replicated caching, seamless integration with any backend, write‑through/read‑through and write‑behind modes, automatic schema mapping, and full SQL support, enabling scalable, high‑speed data access for TB‑scale workloads.
