DataFunCon Conference – OLAP, StarRocks, ClickHouse, and ByteHouse Technical Sessions

OLAP is a currently very active direction in big data, with new technologies and breakthroughs emerging rapidly and being widely used in reporting analysis, behavior analysis, business intelligence, and real‑time data warehouses.

Facing many new technologies, companies encounter challenges in technology selection, standardization, and multi‑scenario support; using a single engine to support as many scenarios as possible is a future trend.

The DataFunCon forum invites technical experts from Ctrip, Didi, Bilibili, and ByteDance to deeply interpret next‑generation OLAP technology applications, including report query acceleration, materialized view exploration, multi‑engine convergence, and ELT practice in data warehouses, providing valuable references for those involved in OLAP selection, platform architecture, and scenario implementation.

Speaker: Xu Peng – Ctrip, Director of Big Data Technology

Personal introduction: Focuses on distributed computing and storage, with deep research on Spark and PostgreSQL.

Talk title: StarRocks Helps Ctrip Achieve Ultra‑Fast Report Query Transformation

Outline: 1. Overall report service architecture; 2. Existing problems; 3. StarRocks features; 4. Benefits after integrating StarRocks; 5. Migration challenges and solutions.

Audience benefits: 1) Accelerated report queries with StarRocks; 2) Fast migration from Presto to StarRocks; 3) Query optimization techniques for StarRocks.

Speaker: Wang Zhibo – Bilibili, Senior OLAP Platform Engineer

Personal introduction: Senior engineer on Bilibili's OLAP platform, responsible for ClickHouse kernel refactoring and performance tuning; previously worked at Yixin and TAL, with five years of big‑data platform development experience.

Talk title: One Size Fits All: ClickHouse Practices Across Bilibili’s OLAP Scenarios

Outline: Sharing how Bilibili unified multiple OLAP engines onto ClickHouse and enhanced the open‑source engine for various scenarios, including building a next‑generation log analysis platform, strengthening user‑behavior analysis, and enabling real‑time UPSERT analysis.

Audience benefits: 1) Experience migrating logs from Elasticsearch to ClickHouse; 2) Using ClickHouse for diverse user‑behavior analyses; 3) Query push‑down optimization and table replication parameter tuning.

Speaker: Li Minghuang – Didi, Senior Software Development Engineer

Personal introduction: Apache Druid and StarRocks contributor; after a master’s degree joined Didi, participated in building Didi’s OLAP engines Presto and ClickHouse, now mainly responsible for Apache Druid and StarRocks.

Talk title: StarRocks Materialized Views in Didi’s Practice

Outline: 1) Background and development of StarRocks at Didi; 2) Concept and advantages of materialized views; 3) Synchronous materialized view practice; 4) Asynchronous materialized view practice; 5) Summary, advantages, limitations, importance in data analysis, and future directions.

Audience benefits: 1) Improve multi‑dimensional query performance via synchronous materialized views; 2) Enhance high‑cardinality distinct count real‑time dashboard performance with asynchronous materialized views.

Speaker: Wang Tao – ByteHouse (ByteDance), Senior R&D Engineer

Personal introduction: Works in ByteDance’s ByteHouse team focusing on mixed‑load workloads; previously at Huawei and Ant Group, deep research on Spark, Flink, Ray, and cloud computing.

Talk title: ELT in ByteHouse: Practice and Outlook

Outline: Current ByteHouse usage scenarios and challenges at ByteDance; shortcomings when handling ELT workloads; architectural improvements made to boost offline capabilities; future plans.

Audience benefits: 1) Understand ByteHouse’s ELT design and practice; 2) Learn the business benefits of ByteHouse’s ELT capabilities; 3) Gain insight into ByteHouse’s future ELT roadmap.

Ticket Information

DataFunCon tickets can be obtained by scanning the QR code to add Da Chuan on WeChat. Limited‑time discount: 30% off, saving ¥1500.

For full conference details, click “Read Original” to visit the official website.