PolarDB for Oracle uses a three-node architecture, the multi-zone feature, and the backup and restoration feature to ensure high reliability.

Three nodes

The Data Migration Assistant (DMA) feature of PolarDB for Oracle allows you to replicate data across data centers. This helps you ensure financial-grade consistency and high reliability. At the kernel level, physical logs are synchronized based on the Raft protocol. This ensures that no data is lost when a single point of failure (SPOF) occurs. If your primary node fails to run as expected, a failover is performed. Features such as node weight configuration and node election customization allow you to use different failover configurations to meet different requirements. Data can be only read from read-only nodes. In scenarios where more read operations need to be performed than write operations, read-only nodes can be added to increase read capabilities. Log nodes are provided to retain only real-time log records. This ensures that no data is lost and that the cost is approximately equal to the cost for the traditional databases that are deployed in primary/secondary mode. Traditional databases cannot ensure strong consistency because asynchronous replication is used to synchronize data between primary and secondary databases. To solve this issue, PolarDB for Oracle uses shared distributed storage to ensure zero recovery point objective (RPO) in the event of an SPOF in the cluster.

Multi-zone deployment that supports the storage of up to six replicas

PolarDB for Oracle stores data in multiple zones. Each data record can have up to six replicas to ensure reliability across data centers.

Fast backup

PolarDB for Oracle uses the snapshot technology at the underlying distributed storage layer to back up data. This way, less than 5 minutes are required to back up 100 TB of data. This allows you to back up business data before each system upgrade or cutover to ensure the security of business data. The fast backup capabilities make PolarDB widely used in industries such as gaming.

Lock-free backup

The entire backup process is implemented at the storage layer by creating consistent snapshots. This operation is transparent to databases and does not require locks on databases. This ensures that the backup does not affect your business, improves the backup efficiency, and minimizes the impact on the entire system.