Focused on technical sharing in the data field, covering a tech stack that includes Hadoop, Spark, Flink, Kafka, Fluss, Paimon, Iceberg, StarRocks, ClickHouse, ES, Milvus, and more. Welcome to follow.
As AI automates coding and model tuning, technologists must shift from pure technical mastery to understanding business needs and user pain points, because AI lacks human insight and the real value lies in connecting technology to practical solutions.
This article explains how missing Flink operator UIDs cause savepoint recovery failures when job topology changes, demonstrates the fix with semantic UID naming conventions, and shows why random UUIDs harm maintainability compared to meaningful identifiers like 'kafka-order-source-uid'.
This guide details the four index types in StarRocks 3.3—Bitmap, Bloom Filter, N-Gram Bloom Filter, and Full-Text Inverted Index—covering their core principles, supported data types, creation syntax, ideal use cases (low/high cardinality, fuzzy/long-text queries), anti-patterns, and real-world optimization examples with performance metrics.
This guide explains how to recover accidentally dropped databases, tables, and partitions in StarRocks using the RECOVER command, covering the 1-day default retention window, configuration options, limitations like TRUNCATE and FORCE drops, and practical step-by-step examples with renaming strategies.
This article outlines production-tested Flink operator naming conventions emphasizing business-logic clarity, concise action-object-result formats, type-specific patterns for map/filter/join/window/sink, DAG-stage prefixes, and pairing name() with uid() for state recovery, while avoiding technical jargon and dynamic variables.
This article provides a comprehensive reference of Elasticsearch 7.x+ RESTful API commands covering index management, document CRUD, query DSL (match, term, bool, aggregations), SQL access, cluster health, index aliases, reindexing, and _cat APIs with practical examples.
In StarRocks compute-storage separation clusters, Stream Load achieves 20-30x faster bulk data ingestion than INSERT by leveraging batch-oriented design, direct compute-node access, single-transaction metadata handling, large-file storage writes, and concentrated resource scheduling, avoiding INSERT's per-row overheads.
This guide explains Elasticsearch cluster architecture, write and read workflows, bulk and async ingestion, data streams, high availability via shard allocation awareness and cross-cluster replication, data skew detection and remediation, and performance tuning for writes, reads, and cluster configuration.
This article explains Elasticsearch's core architecture including distributed sharding, near real-time search, inverted indexes, and key concepts like indices, documents, mappings, and replicas, along with typical use cases and the Elastic Stack ecosystem.
This article compares StarRocks' integrated (BE) and separated (CN + object storage) architectures, detailing write flows like Stream Load with MemTable and compaction, read flows with local vs cached data, and scalability trade-offs.
