bytehouse builds a new generation of "elastic" cloud data warehouse to help enterprises reduce costs
2024-09-19
한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina
in today's internet applications, business traffic often has great uncertainty.
for example, during promotional events such as "618" and "double 11", the number of visits to e-commerce platforms will explode, which may be ten or even a hundred times the usual amount, causing the system to be unable to bear such a high load and crash. this requires the underlying database to have an "elastic" mechanism that can automatically scale resources according to traffic conditions to cope with dynamic changes in business needs.
elasticity is the core feature of cloud-native architecture. as a cloud-native data warehouse launched by volcano engine, bytehouse has high performance and can support users' real-time data analysis and offline analysis of massive data. on the other hand, based on elastic scaling capabilities, users can expand or shrink capacity based on time, resource load and other conditions, reducing the burden of manual management, improving resource utilization and further saving costs.
in traditional infrastructure architecture, enterprises usually need to configure enough resources for peak loads, which results in resources being idle most of the time, causing great waste. however, bytehouse's cloud-native elasticity allows enterprises to dynamically adjust resources according to actual needs, allocate resources only when needed, and realize on-demand use. they are automatically paused when not in use, and no computing layer fees are charged during the pause period, thus reducing costs.
in addition, bytehouse cloud-native elasticity can ensure that resources are used most efficiently. when the application load is low, the system can automatically recycle excess resources and allocate them to other applications or services in need.
so, how does bytehouse achieve cloud-native elasticity?
at the storage level, bytehouse adopts a serverless architecture with low cost and unlimited scalability. at the computing level, bytehouse is based on the paas model and uses containerization to achieve statelessness or weak state, packaging the entire computing group into tenants and applications and presenting them to users, ensuring that there will be no resource requisition conflicts or performance degradation between tenants, and allowing computing resources to be elastically pulled up and elastically expanded and contracted within seconds.
because of the paas approach to computing resources, bytehouse allows users to effectively avoid excessive resource consumption caused by non-standard sql, and the pricing model uses resource usage (cpu) to ensure that user bills are predictable.
bytehouse's elastic scaling capabilities are also widely used in advertising, gaming, finance and other scenarios. for example, a well-known chinese game manufacturer built an integrated real-time data warehouse platform based on bytehouse, which has the capabilities of real-time data access, real-time etl data processing, real-time dimension table association and real-time data services. it can not only support 200,000+ qps high-concurrency point query, but also improve performance by more than 2 times. in terms of resource usage, it reduces costs by 30% compared with the previous architecture.
nowadays, reducing costs and increasing efficiency has always been a topic of great concern to enterprises. in addition to elastic scaling, bytehouse will optimize capabilities such as cold and hot isolation and data compression to further reduce resource costs while continuously ensuring high performance of data query and analysis.
