This topic describes how Database Autonomy Service (DAS) helps the database team of LAIX transform and upgrade.
Customer profile
LAIX is an education company that is driven by science and technology. This company was founded in September 2012. As an advocate of intelligent education, LAIX has developed a large-scale English phonetic database of Chinese people based on the data collected over years. Approximately 3.9 billion minutes of English conversations that consist of 52.4 billion sentences have been recorded in the database by December 31, 2020. LAIX has developed an engine for assessing oral English proficiency and writing proficiency and a self-adaptive learning system based on the database. LAIX is committed to using these systems to develop a systematic solution for English learning that helps users improve language skills such as listening, speaking, reading, and writing.
Challenges
- Various databases to be managed: Different models are required to write data to and read data from databases. To store the data of different types of business, LAIX uses a variety of database services, such as ApsaraDB RDS, ApsaraDB for Redis, PolarDB, and ApsaraDB for HBase. These databases are difficult to be managed in a centralized manner.
- Demand for precise optimization caused by fast iteration: Fast business iteration imposes great pressure on database development and O&M teams. Database administrators (DBAs) cannot use only their experience to optimize business and keep up with the pace of business iteration.
- High requirements for service continuity: LAIX has high requirements for service continuity as the number of users grows. However, the time required to troubleshoot online exceptions is unpredictable.
- Demand for cost reduction: LAIX aims to develop cost-effective systems by using the scalability of cloud-native database services.
Solutions
- LAIX realizes database autonomy based on the core features of DAS. These features include 24/7 anomaly detection, SQL diagnosis, automatic SQL throttling, automatic scaling for performance optimization, automatic SQL optimization, and intelligent stress testing. For more information, see 24/7 anomaly detection, SQL diagnosis engine, Automatic SQL throttling, Automatic scaling for performance optimization, Automatic SQL optimization, and Intelligent stress testing. LAIX also calls the DAS API to automatically troubleshoot online exceptions, optimize databases based on intelligent diagnosis, and use digitalized assessment of database capacity. For more information, see API operations that are available in DAS. In addition, LAIX develops a methodology for intelligent database O&M.
- DBAs of LAIX focus on the design of business-oriented data architectures instead of database O&M.
Benefits
- DAS frees DBAs from heavy O&M workloads.
- DAS reduces the average time required to restore an operation failure.
- DAS provides the intelligent stress testing feature to ensure that the storage capacities of instances meet the business requirements of LAIX.
Customer feedback
"DAS helps LAIX manage databases in a more professional and efficient manner. The AI-based optimized database engines provided by DAS help LAIX reduce a large amount of workloads for DBAs and make online troubleshooting more efficient. This way, database autonomy is realized." -- LAIX