This comprehensive guide introduces Elasticsearch’s core concepts, installation, REST API operations, and practical examples of indexing, updating, deleting, bulk processing, searching, filtering, and aggregations, helping readers build scalable, near‑real‑time search and analytics solutions for diverse use cases.
This article introduces Elasticsearch, an open‑source distributed search and analytics engine, explaining its architecture, core concepts such as clusters, nodes, shards, replicas, indices, inverted indexes, documents and fields, and how these components enable fast, scalable searching and data analysis.
Learn how to install, configure, and connect Kibana with Elasticsearch, explore data using Discover, craft powerful queries with Lucene syntax, create visualizations and dashboards, and monitor your logs—all illustrated with step‑by‑step screenshots and practical tips for effective data analysis.
This article explains query separation as a strategy to accelerate slow large‑table queries by duplicating data to a dedicated query store, outlines when to adopt it, compares synchronous, asynchronous, and binlog approaches, discusses storage choices such as MongoDB, HBase, Elasticsearch, and addresses consistency and MQ challenges.
Didi’s Elasticsearch platform, built on ES 7.6 and deployed on physical machines with containerized gateway and control layers, provides a multi‑tenant, high‑performance search service—featuring a user console, operational controls, ZGC‑based latency reductions, cost‑saving compression, custom security, real‑time cross‑datacenter replication, and a roadmap toward ES 8.13.
This article examines how vector search using Milvus handles sentence and paragraph queries compared to traditional Elasticsearch, demonstrating the advantages of embedding‑based semantic matching for document search scenarios through practical experiments and visual results.
During flash‑sale events like JD or Taobao, a tiny fraction of products generate the majority of traffic, causing server overload; this guide explains how to identify and collect hotspot product data in real time using an OpenResty‑Lua front‑end, Kafka for messaging, and Elasticsearch for storage and analysis.
This article details Didi's comprehensive big data cost governance framework, covering its data architecture, asset management scoring, Hadoop and Elasticsearch cost optimization methods, and practical insights on organizational processes and incentives for effective cost control.
This article describes the challenges of synchronizing billions of order records to Elasticsearch and presents the design, architecture, and key technical details of a generic data‑sync service (ECP) that supports multiple data sources, dynamic rate limiting, retry strategies, SPI‑based extensibility, environment isolation, health‑check, fault recovery, smooth migration, and elegant logging.
This article walks through upgrading the open‑source Mall e‑commerce system to Spring Boot 3 and JDK 17, covering dependency version updates, migration from SpringFox to SpringDoc, new Spring Data Elasticsearch usage, revised Spring Security configuration, Docker deployment tips, and essential code changes.
This guide shows how to deploy a lightweight Elasticsearch instance alongside a proxy and Kibana with Docker‑Compose, configure SLS credentials, create index patterns, and use Kibana’s query and visualization features to explore logs stored in Alibaba Cloud Log Service.
When faced with a business requirement to filter up to 100 000 records from a pool of tens of millions and then sort and de‑duplicate them, this article explores four technical solutions—multithreaded ClickHouse pagination, Elasticsearch scroll‑scan, a combined Elasticsearch‑HBase approach, and RediSearch with RedisJSON—detailing their design, implementation, performance testing, and trade‑offs.
This article explains the challenges of synchronizing billions of order records to Elasticsearch, outlines the design of the ECP service that automates heterogeneous data integration, and details its implementation including multi‑source reading, SQL parsing, dynamic rate limiting, retry mechanisms, SPI‑based extensibility, environment isolation, health‑check, smooth migration, and logging.
Facing the complexities of running stateful services like Elasticsearch and ClickHouse in production, Bilibili’s infrastructure team detailed their migration to Kubernetes, describing the architectural design, custom operators, storage provisioning with LVM, network configuration, high‑availability strategies, observability, and the resulting cost, quality, and efficiency gains.
This guide explains how to monitor Kubernetes events, differentiate normal and warning events, and use tools like kube-eventer and kube-event-exporter to collect, alert on, and analyze cluster events through webhook, Kafka, Logstash, and Elasticsearch, enabling comprehensive observability and troubleshooting.
This article provides a comprehensive, hands‑on guide to installing and configuring the ELK logging stack—including Elasticsearch, Logstash, Kibana, and Filebeat—using Docker on an Ubuntu VM, covering architecture, command‑line setup, configuration files, troubleshooting tips, and future extension options.
This article examines a MySQL slow‑query incident caused by the optimizer using the primary clustered index instead of an appropriate secondary index, explains the underlying index structures, and presents solutions such as FORCE INDEX and migrating complex queries to Elasticsearch for long‑term performance improvement.
This article details Didi's comprehensive big‑data asset governance platform, covering its architectural layers, Hadoop and Elasticsearch governance practices, health‑score models, lifecycle recommendations, and future plans for automated and intelligent data governance to reduce cost and manual effort.
This article explains how traditional keyword search with Elasticsearch often yields inaccurate or incomplete results for document retrieval, introduces vectorization and semantic search using NLP embeddings, and demonstrates a practical workflow that combines these techniques with the Milvus vector database to achieve more accurate and efficient document search.
JD’s Mini‑Program Data Center integrates data collection, storage, and real‑time analysis using Flink, ClickHouse, and Elasticsearch to provide comprehensive monitoring, user behavior insights, and scalable analytics for mini‑programs across JD’s ecosystem, enabling precise operations and future AI‑driven enhancements.
The author recounts taking the Elastic Certified Engineer exam in March 2024, outlines the official syllabus, shares ten hands‑on exam tasks with code snippets, and provides practical tips and personal reflections on preparation and execution.
This tutorial explains how to let users upload PDF, Word, or TXT files, install the Elasticsearch Ingest Attachment Processor Plugin, create an ingest pipeline and index mapping, convert files to Base64, and perform fuzzy searches with highlighted results using Spring Boot and Java code examples.
This article walks through the end‑to‑end design and implementation of a searchable system that scales from millions to hundreds of millions of products, covering business background, data‑center architecture, ES cluster sizing, multi‑room data sync, RocketMQ‑Flink pipelines, and multi‑layer reconciliation to achieve high QPS and strong consistency.
This article explains how JD ElasticSearch has integrated the high‑performance ZSTD compression algorithm, details the motivations behind its adoption, presents benchmark results comparing it with LZ4 and best_compression, and provides step‑by‑step instructions and code snippets for configuring and using the new jd_zstd codec in Elasticsearch.
Bilibili migrated its Elasticsearch and ClickHouse clusters to Kubernetes using custom operators, CRDs, LVM‑based local storage, MacVLAN networking, and pod anti‑affinity, achieving higher resource utilization, isolation, and automation that cut server count, reduced latency spikes, and lowered operational costs dramatically.
This guide walks through creating a Spring Boot backend that accepts PDF, Word, and TXT uploads, extracts their text using Elasticsearch's ingest‑attachment plugin, stores metadata in MySQL, and provides fuzzy search and highlighted results via Elasticsearch queries.
This article compiles a comprehensive Java backend interview guide covering network protocols, HTTP characteristics, core data structures and algorithms, MySQL logging mechanisms, Elasticsearch full‑text search, Kafka reliability and deduplication, common micro‑service components, load‑balancing algorithms, and essential Spring concepts such as IoC, AOP, transactions, and MVC workflow.
Learn how to construct near‑real‑time Elasticsearch indexes for petabyte‑scale datasets by comparing MySQL limitations, leveraging inverted indexes, using Hive and wide tables, and employing binlog‑based pipelines with Canal and Otter to achieve second‑level index updates.
This article details the design of a low‑latency video recommendation system, covering requirements, overall architecture, data pipelines, consistency handling, write smoothing, and performance optimizations such as multi‑level caching and Elasticsearch tuning.
This guide explains Elasticsearch cluster architecture, node roles, and metadata storage, then details step‑by‑step recovery procedures for master‑node loss, complete master outage, data‑node failures, shard allocation problems, corrupted shards, translog issues, and missing segment files, including relevant API commands and tool usage.
Didi replaced its Elasticsearch‑based log platform with ClickHouse, redesigning architecture into isolated Log and Trace clusters, using hourly‑partitioned MergeTree tables and aggregating views to handle petabyte‑scale writes, diverse low‑latency queries, and high QPS, achieving over 400 nodes, 40 GB/s throughput, 30 % cost savings and four‑fold query latency reduction.
This article explains Elasticsearch fundamentals—including clusters, nodes, shards, replicas, segments, routing, indexing, updates, deletions, search flow, pagination methods, and visualization tools—while offering practical optimization recommendations for high‑performance and scalable deployments.
This article provides a comprehensive overview of Elasticsearch, covering its distributed architecture, key features such as JSON RESTful APIs and multi‑tenant support, core functionalities like full‑text search and aggregations, comparisons with Solr, advanced mapping techniques, various query DSLs, suggestion mechanisms, and practical zero‑downtime reindexing strategies.
This tutorial explains how to integrate Elasticsearch into a Spring Boot application, covering project setup, dependency addition, configuration, entity and repository creation, service and controller implementation, and testing to enable efficient full‑text search for large data sets.
This course provides comprehensive, hands‑on training on ElasticSearch, Logstash, Kibana, and the ElasticStack ecosystem, covering advanced development, cluster design, performance tuning, security, and real‑world integration techniques for large‑scale data processing.
This article presents a comprehensive case study of how an IT operations team reduced weekly ticket volume from over 150 to fewer than 20 by classifying issues, redesigning workflows, introducing automation tools, deploying a knowledge‑base pre‑position strategy, and integrating an intelligent question‑answering chatbot.
This article documents the step‑by‑step migration of a Java campus‑blog project from JDK 8 + Spring Boot 2.6 to JDK 17 and Spring Boot 3.2, explaining why the upgrade is worthwhile, detailing code and configuration changes, and sharing solutions to the numerous compatibility issues encountered.
The article explains how to sort image filenames containing numbers in Elasticsearch by either using a painless _script to extract the numeric part at query time or, more efficiently, by preprocessing filenames with an ingest pipeline that creates a numeric field for fast sorting, recommending the pipeline for performance‑critical use cases.
The article describes NetEase Cloud Music’s public‑opinion analysis platform, which integrates external and internal data streams into a layered architecture—ingestion, processing, storage in Elasticsearch, visualization, and monitoring—and employs GPT‑based analyzers for clustering, sentiment, summarization, and intelligent alerts while optimizing costs and planning automated GPT‑driven reports.
This guide walks through integrating Spring Data Elasticsearch into a Spring Boot application, covering entity mapping annotations, repository creation, service layer implementation, and unit tests to perform create, read, update, delete, and fuzzy search operations on Elasticsearch indices.
This article compares ClickHouse and Elasticsearch, highlighting ClickHouse's superior write throughput, query speed, and lower server costs, then provides detailed deployment steps for ClickHouse, Zookeeper, Kafka, and FileBeat to build a cost‑effective big‑data analytics platform.
This article introduces the core components of the ELK stack, compares three typical deployment architectures—including Logstash‑only, Filebeat‑assisted, and Kafka‑backed designs—and provides concrete configuration examples and troubleshooting tips for multiline merging, timestamp handling, and module‑level log filtering.
This article details the design and implementation of a highly available membership system that handles billions of users and peak traffic of over 20,000 TPS, covering Elasticsearch dual‑center master‑slave clusters, traffic‑isolated three‑cluster architecture, Redis caching with distributed locks, and dual‑center MySQL partitioning.
This article introduces Easy-Es, an ORM framework built on Elasticsearch's RestHighLevelClient, explains its architecture and advantages, and provides step‑by‑step instructions—including Maven/Gradle setup, configuration, entity and mapper creation, and full CRUD test examples—so developers can quickly integrate powerful search capabilities into Spring Boot applications.
Facing data security and cost challenges in SaaS, the author evaluates ClickHouse versus Elasticsearch, highlighting ClickHouse’s superior write throughput, query speed, lower storage and CPU usage, and provides detailed deployment guides for Zookeeper, Kafka, FileBeat, and ClickHouse to build a cost‑effective private analytics platform.
This article introduces Easy-Es, an Elasticsearch ORM that mimics MyBatis‑Plus syntax, explains its key features, shows how to integrate it into a Spring Boot project, and provides comprehensive code examples for index management, CRUD, simple and advanced product searches, recommendation and aggregation functionalities.
This article details the end‑to‑end design, implementation, and operational results of a smart customer‑service platform that automates FAQ capture, leverages both Elasticsearch and LLM‑based models, and provides a low‑code, multi‑team backend for rapid issue resolution.
Easy-Es is a Java ORM framework built on Elasticsearch's RestHighLevelClient that simplifies search‑engine development, offers Mybatis‑Plus‑like usage, provides configuration and dependency details, and includes complete code examples for creating indexes and performing CRUD operations within Spring Boot applications.
This guide demonstrates how Elasticsearch 8.11's new ES|QL query language matches the traditional DSL syntax, covering basic searches, sorting, wildcard matching, aggregations, REST API nuances, and practical examples that show equivalent DSL and ES|QL statements with visual results.
Bilibili created an AI‑powered customer‑service platform that integrates a WeChat Work chat UI, a conversation state machine, Elasticsearch and LLM‑based FAQ retrieval, and a low‑code admin console, replacing a thousand‑person manual support team, handling thousands of queries, boosting resolution rates and enabling cross‑team reuse.
This article compiles practical Elasticsearch recommendations covering query caching, filter contexts, pagination, aggregation strategies, index mapping, shard design, and scripting best practices, providing developers with actionable insights to improve search performance, reduce resource consumption, and prevent common operational issues.
The article explains how Elasticsearch stored scripts can be created and invoked within ingestion pipelines to decode Base64 and hexadecimal data, showing reusable Painless scripts, pipeline configuration, and the advantages of modular, maintainable transformations without inline code duplication.
This article presents a concise Spring Cloud microservice architecture diagram, explains each component such as DNS, CDN, LVS, Nginx, Nacos, Sentinel, Redis, and Elasticsearch, and offers practical guidance on their roles, selection criteria, and implementation details for interview preparation.
This article explores the rise of generative AI, outlines popular models like ChatGPT, DALL‑E, and Google Bard, examines their limitations, and then delves into Elasticsearch’s Relevance Engine and vector capabilities, demonstrating how to store, index, and query dense embeddings with practical code examples.
This article analyzes Elasticsearch shard imbalance incidents, explains the built‑in shard balancing algorithm and its configuration parameters, demonstrates weight calculations with source code, and proposes practical improvements—including shard count adjustments and a custom load‑aware balancing tool—to achieve more effective cluster load distribution.
This article introduces the ELK stack’s core components, compares four typical deployment architectures—including Logstash‑only, Filebeat‑based, and Kafka‑enhanced setups—discusses their trade‑offs, and provides practical configurations and solutions for multiline log merging, timestamp handling, and module‑specific filtering.
This article compares ClickHouse and Elasticsearch for log analytics, highlighting ClickHouse's superior write throughput, query speed, and lower server costs, then provides a detailed, cost‑effective deployment guide covering Zookeeper, Kafka, FileBeat, ClickHouse installation, and visualization with ClickVisual, plus optimization tips.
This article records a systematic test of Elasticsearch refresh latency, revealing that update operations, a high proportion of deleted documents, and the soft‑delete setting significantly increase refresh time, while the large‑segment strategy and disabling soft delete can reduce latency without harming overall performance.
This article compares ClickHouse and Elasticsearch for log analytics, highlighting ClickHouse’s superior write throughput, query speed, and lower server costs, then details a cost‑effective deployment architecture—including Zookeeper, Kafka, FileBeat, and ClickHouse setup—and shares optimization tips and visualization using ClickVisual.
The Weimob Technology Salon session on "Elasticsearch in Weimob's Practice" shares practical usage recommendations, monitoring setups with Prometheus and Grafana, field‑type guidance, and solutions to common operational challenges, offering developers actionable insights for high‑performance search deployments.
This article compares Elasticsearch's ngram analyzer and the newer wildcard field for fuzzy searching, detailing configuration steps, performance trade‑offs, storage impact, and practical test results to help engineers choose the optimal approach for their use case.
The guide outlines production‑ready Elasticsearch best practices, covering query tuning such as using shard request cache, filter context, size‑0 aggregations and composite aggregations; write strategies like auto‑generated IDs, bulk API sizing and refresh handling; optimal shard counts, explicit mappings with disabled unnecessary features, and general advice to use explicit index names and stored scripts.
This guide walks through configuring a Spring Boot 2.4.12 project with Elasticsearch 7.8, creating index templates, implementing bulk write operations, and testing the setup with sample code and endpoints, enabling automatic index creation and efficient data ingestion.
This tutorial walks through setting up a unified real‑time ELK log analysis platform on Ubuntu, covering installation and configuration of Logstash, Elasticsearch, Kibana, integration with Spring Boot and Nginx logs, and managing services with Supervisor for reliable operation.
RediSearch, a low‑memory Redis module, offers comparable full‑text search capabilities to Elasticsearch, with faster indexing and query performance on modest hardware; this article explains its features, compares benchmarks, shows Docker installation, and provides Java/Jedis code examples for creating, querying, and managing indexes.
This article explains how to use Filebeat together with Graylog to collect, ship, store, and analyze logs from multiple environments, covering tool introductions, configuration files, Docker deployment, Spring Boot integration, and practical search syntax for effective log monitoring.
This article explains the concept of query separation, its suitable scenarios, implementation methods—including synchronous, asynchronous, and binlog approaches—storage options, MQ considerations, and how it can dramatically improve query performance for massive tables.
A detailed investigation of an Elasticsearch timeout incident reveals how an overly aggressive node_concurrent_recoveries setting caused CPU saturation, disk I/O spikes, and shard relocation overload, and outlines the steps taken to isolate the faulty node and restore cluster performance.
This tutorial walks you through the fundamentals of MySQL binlog replication, installing and configuring Canal, setting up Elasticsearch, Kibana, and the IK analyzer, and then demonstrates both full and incremental data synchronization from MySQL to Elasticsearch.
This article explains how Elasticsearch tokenizes text, why small spelling errors or plural forms can miss matches, and how synonyms improve search recall, comparing the trade‑offs of applying synonyms during indexing versus at query time, with practical code examples and tips for managing large synonym lists.
This article explains the ELK stack components, compares three typical deployment architectures—including Logstash‑only, Filebeat‑based, and a Kafka‑backed version—and provides detailed configuration examples for multiline log merging, timestamp replacement, and module‑based log filtering.
The team traced read‑request timeouts to a single overloaded Elasticsearch node where an excessively high node_concurrent_recoveries setting caused many simultaneous shard recoveries and disk‑watermark‑driven relocations, and resolved the issue by lowering concurrent recoveries, enabling adaptive replica selection, and adjusting allocation settings.
This article introduces Elasticsearch, explains its hot‑warm‑cold lifecycle management, demonstrates vector search and built‑in NLP capabilities, and describes how the 360 Zhihui Cloud Platform integrates these features with practical test cases and new visualization tools.
Alibaba Cloud unveiled AI-driven upgrades to its OpenSearch and Elasticsearch services, highlighting LLM‑based conversational search, three‑fold vector retrieval speed gains, and up to 70% cost reductions through serverless architectures and extensive performance optimizations.
This step‑by‑step guide shows how to download, configure, and launch SkyWalking with Spring Boot, connect it to Elasticsearch, instrument a sample service and Redis calls, and view real‑time monitoring data through the SkyWalking UI and Kibana.
This article explains why Elasticsearch is fast, outlines its key features, and provides a complete Spring Boot 2.4 tutorial—including dependencies, configuration, data model, repository usage, and test code—for building a distributed search service.
This article explains the concept of full-text retrieval, introduces Elasticsearch as a popular open‑source search engine built on Apache Lucene, and provides detailed step‑by‑step installation instructions for both traditional setups and Docker containers, including required environment configuration and common troubleshooting.
At QCon 2023, Baidu’s BES team detailed how their cloud‑native Elasticsearch service has been engineered for large‑scale vector search, describing architecture, C++ plugin integration, memory‑saving storage tricks, HNSW/IVF optimizations, filter strategies, and real‑world multimodal video and LLM knowledge‑base deployments.
This article explains the principles of vector databases, outlines the engineering practices of Baidu Intelligent Cloud BES for large‑scale vector retrieval, discusses optimization techniques such as HNSW, IVF and filter integration, and presents real‑world AI use cases and future development directions.
The data loss in a Flink‑Kafka job was caused by a mis‑configured bootstrap.servers list that mixed production and pre‑release Kafka clusters, leading different subtasks to connect to different clusters, resulting in inconsistent partition discovery and offset fetching, which omitted several partitions until the list was corrected.
This article introduces Easy-Es, a Java ORM framework built on Elasticsearch's RestHighLevelClient that simplifies query construction and CRUD operations, provides step‑by‑step setup including Maven dependencies, configuration, entity and mapper definitions, controller examples, and advanced features such as condition builders, index management, logging, and aggregation queries.
This tutorial explains how to set up the environment, configure Elasticsearch and Logstash, create the necessary MySQL tables, and use a Logstash pipeline to continuously sync MySQL records into an Elasticsearch index, while also covering common pitfalls and troubleshooting steps.
This guide explains how to optimize Elasticsearch 7.8 performance by adjusting the refresh interval, replica count, JVM heap settings, bulk API usage, storage devices, mapping design, and query patterns, providing practical commands and visual examples for each step.
This guide walks you through installing Kibana, configuring its connection to Elasticsearch, creating index patterns, using Discover for searches, mastering Lucene‑based query syntax, building visualizations, assembling dashboards, and monitoring logs, all illustrated with clear screenshots and code examples.
Learn how to integrate Elasticsearch 7.8 with Spring Boot 2.4, covering essential dependencies, index management, document CRUD, bulk operations, and advanced queries such as pagination, sorting, filtering, highlighting, aggregations, and grouping, all demonstrated with concise Java High-Level REST client code examples.
In a high‑traffic e‑commerce system, sharding makes cross‑shop queries inefficient, and adding terms aggregations in Elasticsearch caused queries to take dozens of seconds, but using the "execution_hint":"map" option dramatically reduces aggregation latency.
This article explains why MySQL LIKE queries struggle with large datasets, introduces Elasticsearch and Sphinx as powerful PHP search engine alternatives, and provides step‑by‑step PHP code examples for setting up indices, adding documents, and performing full‑text searches.
The article explains query separation as a technique to improve performance of massive business tables by storing write‑optimized data and query‑optimized data separately, discusses when to apply it, various implementation methods, storage options, and operational considerations such as MQ handling and consistency.
This article compares Elasticsearch and ClickHouse in terms of write throughput, query speed, and server cost, then provides a detailed step‑by‑step deployment guide for Zookeeper, Kafka, FileBeat, and ClickHouse, including common issues and their solutions.
This tutorial walks through configuring Canal in cluster mode to capture MySQL row‑based binlog events, publish them to a RocketMQ topic, and consume the ordered messages to update an Elasticsearch index, covering MySQL, RocketMQ, Elasticsearch mappings, and the required Canal properties.
This article analyzes the challenges of logging in distributed services, compares the traditional ELK stack with Baidu's Tianyan platform, and details Tianyan's architecture, data collection, high‑throughput transmission, storage, retrieval, resource isolation, dynamic cleanup, and best‑practice recommendations, complete with code examples and performance insights.
This tutorial explains why centralized log collection is essential for micro‑service systems, introduces the ELK stack (Elasticsearch, Logstash, Kibana), provides step‑by‑step installation on Ubuntu, shows how to configure Logstash shipper and indexer pipelines, and demonstrates integration with Spring Boot and Nginx logs for real‑time monitoring.
The article explains how our driver‑passenger matching service migrated from load‑balanced Elasticsearch access to a direct RestClient, then automated node discovery using the built‑in Sniffer and SniffOnFailureListener, detailing its scheduling, request logic, and how this eliminates manual IP management while keeping the client in sync with cluster topology.
This tutorial walks you through setting up a unified ELK (Elasticsearch, Logstash, Kibana) log analysis platform on Ubuntu, covering component installation, configuration for Spring Boot and Nginx logs, Grok parsing, and using Supervisor to run the services as background daemons.
This tutorial demonstrates how to use Alibaba's Canal in cluster mode with Zookeeper to capture MySQL row‑based binlog events, publish them via RocketMQ, and consume them with a custom listener that transforms inserts, updates, and deletes into synchronized documents in an Elasticsearch index.
This tutorial explains how to set up a unified ELK (Elasticsearch, Logstash, Kibana) real‑time log analysis platform on Ubuntu, covering component installation, Logstash shipper and indexer configurations, integration with Spring Boot and Nginx logs, and background management with Supervisor.
This step‑by‑step tutorial shows how to configure Alibaba's Canal to capture MySQL binlog changes, route them through RocketMQ, and index the data into Elasticsearch, covering cluster mode, MySQL and Elasticsearch setup, RocketMQ topic creation, Canal properties, and consumer implementation.
The team traced near‑100 % CPU spikes in an Elasticsearch cluster during stress tests to frequent JIT deoptimizations triggered by a switch‑based experiment routing logic, and resolved the issue by replacing the switch with a function map, which eliminated deoptimizations and stabilized CPU usage.
The article details the JD Mini Program platform, its data‑center development milestones, comprehensive business panorama, technical architecture, data collection, storage, and analysis pipelines—including Flink‑based real‑time monitoring, ClickHouse custom analytics, and Elasticsearch user‑behavior insights—while outlining current challenges and future AI‑driven enhancements.
In this interview, Baidu Smart Cloud’s big‑data lead Li Li explains how the team combined Elasticsearch with the open‑source Faiss engine, built custom columnar storage, and iteratively optimized their vector‑search service to support billions of high‑dimensional vectors for large‑model applications.
