This topic describes the benefits of PolarDB.
Ease of use
PolarDB is compatible with a variety of popular relational database engines. It is fully compatible with MySQL and PostgreSQL, and is highly compatible with Oracle syntax, with little or no code and application modification.
- Separation of computing and storage: Compute nodes share storage resources. You only pay for compute nodes when you add read-only nodes, which greatly reduces scale-out costs.
- Serverless storage: You do not need to manually configure storage space, because the storage space is automatically scaled based on the data volume. You only need to pay for the database capacity that you have used.
- With an improved database kernel, PolarDB supports physical replication, RDMA protocol, and shared distributed storage, which greatly improves performance.
- An PolarDB cluster contains one primary node and up to 15 read-only nodes. The cluster meets performance requirements in high concurrency scenarios, particularly suitable for scenarios where read requests are far more than write requests.
- An PolarDB cluster shares storage among the primary node and read-only nodes. To apply data changes to nodes in the cluster, you only need to change data once.
Storage capacity for hundreds of terabytes of data
PolarDB uses distributed block storage and a file system to allow automatic scale-up of database storage capacity, regardless of the storage capacity of each node. This enables your database to handle up to hundreds of terabytes of data.
High availability, reliability, and data security
- Supports shared distributed storage to eliminate data inconsistency in the secondary database caused by asynchronous primary-secondary replication. This ensures zero data loss if a single point of failure occurs in a database cluster.
- Supports a multi-zone architecture. Data replicas are available across multiple zones for database disaster recovery and backup.
- Provides various security measures for your database access, storage, and management. These measures include setting an IP whitelist for database access, using VPC for network isolation, and creating multiple replicas for data storage.
Rapid elastic scaling to handle workload spikes
- Configuration upgraded or downgraded within 5 minutes
PolarDB supports rapid CPU and memory expansion by using container virtualization and shared distributed block storage.
- Nodes added or removed within 5 minutes
PolarDB can dynamically add or remove nodes to help you improve performance and reduce costs. You can use cluster endpoints to mask changes at the underlying layer. In this case, applications are unaware of the addition or removal of nodes.
Based on the snapshot technology of underlying distributed storage, PolarDB requires only a few minutes to back up a database with TB-level data. During the entire backup process, no lock is required, which delivers higher efficiency and minimizes the negative impact.