POLARDB Alibaba independent research and development of next-generation cloud-native relational distributed database, currently compatible with three kinds of database engines: MySQL, Oracle, PostgreSQL.
Computing power scales up to more than 1000 nuclear storage capacity up to 100T. POLARDB combines both commercial databases stable, reliable, high-performance features, but also has
open-source database simple, scalable, continuous iteration advantage, and cost only 1/10 of a commercial database.

See Video Profile

The industry's leading edge

Mass storage, Pratt & Whitney and enjoy the native cloud technology

100T maximum capacity
can be extended laterally up to 16 nodes, each node with the highest vCPUs 88
Serverless, only has to be paid to the capacity actually used

Surging performance, resilient to sudden increase in load speed

Calculated using & separate storage, distributed shared memory architecture
using the RDMA, file system user mode, Parallel Raft, 3D Xpoint such as the latest software and hardware technologies, high performance concurrent lifting up scene 6 fold
increase in computing nodes 5 minutes, rapid elastic deal with unexpected traffic load

Steady as a rock, data security is always first

All nodes share the same copy of data, data 0 standby switching loss
provided whitelist mechanism, the VPC network isolation
using physical replication, the main delay in milliseconds from

Inclusive, leading the global digital transition

Compatible with a variety of popular relational database
is fully compatible with MySQL / PostgreSQL, no need to modify the code / application
is highly compatible with Oracle

Well-built function

Huge amounts of data, scale to support hundreds of TB of data

A distributed block storage devices and file systems, easy to deal with hundreds of TB of data scale.

Massive data processing capacity

Support hundreds of TB-level data scale easily to meet the needs of business and data growth

OLTP processing capability

While supporting massive data, compatible with MySQL, Oracle, PostgreSQL strong OLTP processing capability

Business continuity, high availability service assurance

Support primary node, read-only node failure automatic recovery, and support for the master node, between automatic and manual Failover read-only nodes.

Availability of the master node and the node read-only

Automatically detecting unusual master node and a read-only node, and automatic recovery.

The master node and the node read-only function on demand Failover

The priority policy node, after detecting the fault can be automatically migrate or migrate resources designated by a user.

Reliable and secure, Snapshot backup and recovery

Snapshot mechanisms supporting block data storage device for backup and recovery of data, in conjunction with support Redo log recovery mode by way point in time.

Snapshot backup support.

Sanpshot support for fast backup and recovery.

Multi-copy backup data support.

Ensure their backup data reliability and availability.

Flexible upgrade, expansion of storage capacity needed

Upgraded to support the master node and a read-only node CPU, memory and disk. And a read-only node supports Scale Out lateral extension.

With more elasticity.

Support master node, read node CPU, memory, storage specification upgrade or downgrade.

Read-only node expansion.

Can maximum support to the 15 read-only nodes.

Different versions of the engine advantage

POLARDB for MySQL

POLARDB provide fully compatible with versions of MySQL 5.6, cloud-oriented on providing customers with a number of enterprise-class features.

Backup and Recovery

Second-level snapshot technologies to solve problems large-capacity backup database;
use the underlying storage snapshot backup, to avoid locking the overhead of traditional physical backup;
snapshot-based backup and physical Redo logs, database achieve full recovery (Point-in-Time Recovery, PITR).

Intelligent Agents

Adaptive separate read and write, an address to access all the nodes, the load balancing automatically without having to configure;
support multiple levels of the final consistency consistency / session consistency, to meet different business scenarios;
custom address, in off-line traffic load isolator to ensure the stability of the core business.

Diagnostic optimization

[CloudDBA] Examples of real-time monitoring hardware using indicators, slow SQL, and gives a variety of optimization suggestions to help users quickly locate and resolve problems;
[SQL] insight into the real-time collection SQL database full amount, provide powerful analysis, query engine that helps you find potential performance or security risk, SQL logs longest retained for five years.

POLARDB for Oracle (open beta)

Highly compatible with Oracle, Oracle migration to reduce risk and shorten the period of migration, help enterprises to quickly replace Oracle, enter the Smart era.

Oracle compatible

Compatible Compatible full] [Oracle SQL syntax, types, functions, PL / SQL, package, system view, OCI, PRO * C and the like;
built-in function 3000, and an inner package 26 is compatible with the package 317 kinds of methods, systems view 88 .
[] Advanced features compatible with Oracle partition table, heterogeneous queries, HINT and so on.

Intelligent Drive

[Learning] Anti-SQL Injection SQL and SQL misuse;
[index] recommended a key index to solve optimization problems;
when [isolation] user-level resources to address multiple business using the same database instance resource competition problem;
[AAS] Performance Insight a key insight into the macro and micro business performance issues.

Cloud native

[Table] OSS external data separation cold storage, historical data may want to deposit long;
[] seamless collaboration cloud mass calculated Drive force (ADB, MaxCompute, OSS function calculation, etc.).

POLARDB for PostgreSQL (open beta)

Fully compatible with PostgreSQL, supports the separation of computing and storage, independent scalable, pay-per storage volume. Transparent separate read and write operations (this function in development). For medium and large enterprise core business scenarios.

OLTP + OLAP mixed load

[] QPS one million single computing node with the highest 88 vCPU, providing one million QPS;
[] parallel computing business completely transparent, more than 20 times the average speed;
[a] AAS key insight into performance insight micro and macro service performance issues.

Calculation multimode

[] Ganos time and space components compatible GIS standards, support professional data processing time and space, enhance the open source compared to 50-100 times the performance of PostGIS MOD model;
[components] built-in multi-mode full-text search, image recognition, multi-dimensional look, the vector computing, industrial The timing and so on;
[properties] nosql built schemaless, KV and other nosql characteristics.