The ways that customers use cloud resources are changing and an increasing number of customers deploy their applications and databases on the cloud. In this case, databases must be upgraded to support business upgrades and architecture upgrades.

Fast scaling

PolarDB uses an architecture in which compute is decoupled from storage. A newly added compute node can be in service within only 5 minutes. The specification upgrade of a compute node from 4 cores to 88 cores can be completed within only 10 minutes.

Auto scaling

PolarDB O Edition supports auto scaling. After you create a rule-based auto scaling policy, PolarDB automatically upgrades or downgrades your database instances each time the policy is triggered, without the need for manual intervention. This way, PolarDB processes your business workloads with minimal costs. This makes PolarDB suitable for scenarios that have clear workload peaks and troughs. PolarDB can be used together with Alibaba Cloud Auto Scaling and Container Service for Kubernetes (ACK) to manage the capacities of application servers and PolarDB databases.

Billable items for Serverless instances

You do not need to configure instances in advance. Instances are automatically scaled to meet your business requirements. You are charged only for the database space that you use.

DBaaS

PolarDB Platform as a Service (PaaS) is a lightweight Database as a Service (DBaaS) dedicated to the offline database market and the hybrid cloud database market. PolarDB PaaS provides the same user experience as public clouds. PolarDB PaaS is a one-stop platform for lifecycle management and automatic O&M of databases. PolarDB PaaS provides enterprise-grade features. It is stable, reliable, and can be deployed in a flexible manner. PolarDB PaaS also provides comprehensive and professional database services that are provided by database experts and after-sales O&M services.

Resource scheduling

The resource scheduling feature of PolarDB PaaS improves resource utilization and allows you to implement automatic O&M and prediction. This makes the resource scheduling feature suitable for small-scale data centers and large-scale hybrid cloud scenarios.