This topic describes the features of PolarDB-X.
PolarDB-X features the integration of centralized and distributed architectures, financial-grade high availability, distributed horizontal scaling, hybrid transaction/analytical processing (HTAP) loads, low-cost storage, and infinite scalability.
PolarDB-X is compatible with the open source MySQL ecosystem in terms of SQL syntax, transaction behaviors, and ecosystem tools. When you migrate business from MySQL to PolarDB-X, you need to only modify a small amount of code or none at all.
Financial-grade high availability
PolarDB-X adopts a multi-replica architecture and uses Paxos to ensure strong consistency between replicas (RPO = 0). Paxos requires each write operation to be confirmed by more than half of the nodes. This ensures that a cluster can continue to provide services as expected even if one of the cluster nodes fails.
Paxos ensures strong consistency among replicas. This meets the disaster recovery requirements of the finance industry.
Category | Feature | Description | References |
Zone-disaster recovery | Single-zone deployment | Three replicas are deployed in each zone to prevent failures when one node in a minority goes down. | |
Three-zone deployment | Three replicas are deployed in three zones to prevent failures when a single data center goes down. | ||
Cross-region disaster recovery | Deployment in three data centers across two regions | Five replicas are deployed in two regions to prevent failures when a single region goes down. |
Transparent distribution
PolarDB-X focuses on developing its linear scaling capabilities to provide a user experience that is similar to that of standalone databases that run the MySQL engine. Based on technologies such as distributed transaction consistency, data distribution and scheduling, global secondary index (GSI), and cross-shard query, PolarDB-X provides a distributed architecture that is transparent to users.
PolarDB-X provides binary logging to resolve the issue of data forwarding from compute nodes to data nodes in distributed databases. In distributed databases, if you restore nodes based on backup files that are created at different points in time, data inconsistency can occur. To resolve this issue, PolarDB-X ensures data consistency before backup and creates backup files based on the globally consistent data.
Category | Feature | Description |
Data distribution | Partitioned tables | Evenly distributes data to multiple nodes by using the partition rules for partitioned tables. |
GSIs | Supports multi-dimensional queries across partitions in distributed databases. | |
Manual partitioning and automatic partitioning | Provides flexible data partitioning policies to migrate existing business with a few clicks. | |
O&M compatibility | Global binary logs | Provides a protocol that is compatible with MySQL binary logs and allows you to subscribe to incremental logs based on change data capture (CDC). |
Distributed backup and restoration | Supports distributed and consistent backup and restoration. |
Integration of centralized and distributed architectures
PolarDB-X integrates centralized and distributed architectures. It provides Standard Edition (centralized architecture) and Enterprise Edition (distributed architecture) instances. You can change between the instance editions with ease.
Data nodes in PolarDB-X are isolated as centralized databases, which are fully compatible with standalone databases. If distributed scaling is required due to business expansion, you can upgrade the architecture to a distributed architecture with ease. The corresponding distributed components integrate seamlessly with the original data nodes to facilitate scaling. This eliminates the need to migrate data or modify applications.
Category | Feature | Description | References |
Standard Edition (centralized architecture) | Paxos replicas | Provides a majority-based consistency protocol. | |
Lizard transaction system | Provides a distributed transaction system developed by Alibaba Cloud with capabilities to manage transactions on the nodes of a standalone database by using the System Commit Number (SCN) mechanism. | ||
In-place upgrade | Data migration from an ApsaraDB RDS for MySQL database to a PolarDB-X Standard Edition instance | Allows you to migrate data from an ApsaraDB RDS for MySQL database to a PolarDB-X instance. | - |
Upgrade of a PolarDB-X instance from Standard Edition to Enterprise Edition | Upgrades the architecture to a distributed architecture from the smallest node. | Upgrade an instance from Standard Edition to Enterprise Edition |
HTAP
PolarDB-X provides the clustered columnar index (CCI) feature. By default, a row-oriented table in PolarDB-X has primary key indexes and secondary indexes. A CCI is a secondary index that is built on top of a column-oriented structure and is valid for all columns in a row-oriented table. A table can contain both row-oriented data and column-oriented data. In addition, cost-based optimizers and vectorized operators are developed for HTAP, and a set of SQL engines is used to support HTAP.
Category | Feature | Description | References |
Column-oriented data | CCIs | Provides row-oriented tables that have primary key indexes and secondary indexes by default. A CCI is a secondary index that is built on top of a column-oriented structure. | |
Cold data archiving | Archives historical data to Object Storage Service (OSS) based on time partitions. | ||
Read-only instances | Read-only column store instances | Provides the link isolation capability and performs query and analysis based on vectorized engines. |
Open source and multi-cloud deployment
In November 2021, the kernel engine of PolarDB-X officially went open source. To align the iterations and features of the community version with those of the commercial version, the PolarDB-X development team synchronizes the code to open source communities every three to six months.
PolarDB-X provides the lightweight management feature and tools that are compatible with the open source ecosystem. You can use open source code to implement business production and deployment, build your own applications, and perform multi-cloud disaster recovery.
Category | Feature | Description | References |
Open source | Open source kernel engine | Contains multiple distributed components, such as compute nodes (CNs), data nodes (DNs), CDC, Global Meta Service (GMS), and Columnar. | |
Lightweight management | Provides production and deployment capabilities based on Kubernetes Operator. | ||
Multi-cloud deployment | Self-managed databases built on Elastic Compute Service (ECS) instances | Allows you to deploy open source PolarDB-X on physical machines in data centers or ECS instances. | |
Multi-cloud deployment in ApsaraDB for MyBase | Manages open source PolarDB-X based on ECS and supports the centralized management of databases built on ECS instances. |
Security and stability
As a database service developed by Alibaba Cloud, PolarDB-X has received several state-level security certifications. It has been successfully applied in core systems in industries that have high security requirements, such as finance and telecommunications.
PolarDB-X provides comprehensive security measures, such as IP address whitelists, Secure Sockets Layer (SSL), transparent data encryption (TDE), backup encryption, always-confidential, SQL audit and tracing, three-role mode, and tag permissions.