DolphinDB: The second round of DolphinDB is coming
This time, let’s talk about:
New TSDB storage engine architecture and features
Full of dry goods! Not to be missed!
The content is introduced
1. Why develop a new TSDB storage engine?
With the rapid popularity of TSDB, the application scenarios of TSDB are becoming more and more complex. When writing data, it should not only meet the requirements of fast writing massive timelines, but also keep the storage cost low, and meet the requirements of automatic deduplication in out-of-order state. When reading data, it can not only read the latest state of each timeline as fast as possible, but also read the data within a period of time with high concurrency and high speed. It can also quickly complete the statistical analysis of the data within a period of time with full (or large) timeline. We developed a new storage engine, TSDB, for such diverse scenarios. TSDB supports more efficient writes and queries in the above scenarios, and supports all the computational analysis capabilities we provided previously.
2. Where does TSDB’s high performance come from?
This is going to be the highlight of the broadcast. Our storage engine is primarily based on Log Merge Structured Tree(LSMT). LSMT is a classic architecture design in database and the basis of many well-known databases. We also did a lot of performance tuning for application scenarios, resulting in TSDB achieving very high performance. As for the technical details, we will share them in detail in this live broadcast. Hope this part of the content can do database system counterparts have certain inspiration.
3. How to make better use of TSDB?
There is a classic saying in software engineering called “there is no silver bullet”, which means there is no solution to every problem in software engineering. The same can be said of database storage engines: there is no single storage engine that can achieve optimal performance in all scenarios. TSDB is certainly no exception. This part of the live broadcast will describe the skills of using TSDB in combination with financial and industrial Internet of Things scenarios, that is, how to configure the basic parameters of TSDB according to the scenarios. Proper parameter configuration can better take advantage of TSDB’s performance benefits.
Live time
Tuesday, June 29 20:00-21:00
The guest is introduced
DolphinDB storage engine development director Zhou Sin-jing
Once worked as a storage engine R&D engineer in Tencent Cloud database, published many papers in SIGMOD and VLDB, the top database conferences in the past two years, and is an expert in the field of database storage engine.
DolphinDB Senior R&D engineer Jinming Hu
He worked in Optiver (Shanghai) in the research and development of programmed trading strategies and high-frequency trading low-delay systems. He has a deep understanding and practice of machine learning and high-performance systems, and has published several papers in top conferences and journals on machine learning in recent years.
Share point
1. Why the new TSDB storage engine: Application scenario and performance introduction of TSDB
2. Where does TSDB’s high performance come from: Introduction to TSDB’s architecture design
3. How to use TSDB better: introduction to the skills of using TSDB
Participate in the way
DolphinDB assistant [13362150840] note “live” to join the activity