Overview of OceanBase Database AP deployment

You can choose a suitable product edition, deployment plan, and deployment method as needed.

Choose a suitable product edition

OceanBase Database provides two editions: Enterprise Edition and Community Edition.

• OceanBase Database Enterprise Edition: A fully self-developed enterprise-level native distributed database that provides financial-grade high availability on ordinary hardware. It sets a new standard for automatic, lossless disaster recovery across five IDCs in three regions and sets a new world record in TPC-C benchmark tests. It can support more than 1,500 nodes in a single cluster. It is cloud-native, highly consistent, and highly compatible with Oracle and MySQL.
• OceanBase Database Community Edition: A fully self-developed, open-source, and MySQL-compatible single-node distributed database that supports financial-grade high availability, transparent horizontal scaling, distributed transactions, multi-tenancy, and syntax compatibility.

For more information about the features of OceanBase Database Enterprise Edition and Community Edition, see Features of OceanBase Database Enterprise Edition and Community Edition.

Choose a suitable deployment plan

OceanBase Database supports high availability and disaster recovery at the single-node, IDC, and regional levels. You can deploy OceanBase Database in a single IDC, two IDCs, three IDCs in two regions, or five IDCs in three regions. You can also deploy the arbitration service to reduce costs.

For more information about each deployment plan, see Overview of OceanBase cluster high availability deployment plans.

Choose a suitable deployment method

Plan your resources

Before you deploy OceanBase Database, you need to plan and prepare the resources required for the deployment.

• For more information about the server configuration requirements, see Prepare servers.

Deploy a columnstore replica

To support TP-OLAP hybrid workloads (such as real-time decision analysis), OceanBase Database provides support for columnstore replicas. You can deploy a columnstore replica in an independent zone. All user tables (including replicated tables but excluding index tables, internal tables, and system tables) in a columnstore replica are stored in columnar format. OLAP workloads access the columnstore replica through an independent ODP and execute decision analysis workloads in weak-consistency read mode.

From the perspective of log streams, all user tables in the same log stream in a columnstore replica are stored in pure columnar format at the major sstable level. In addition, a columnstore replica follows the same rules as a general read-only replica in terms of replica distribution and weak-consistency read release mechanisms. Except for the major sstable storage type, a columnstore replica is similar to a read-only replica in that it does not participate in elections or log voting and has complete sstables, clogs, and memtables.

After you deploy a columnstore replica, you must configure a routing forwarding strategy and weak-consistency read requests to ensure that OLAP requests are automatically converted into weak-consistency read requests and forwarded to the corresponding columnstore replica.

Considerations

• When you use a columnstore replica, no additional operations are required for the OceanBase cluster. You can deploy the cluster in the normal architecture.

• To access a columnstore replica, you must deploy an independent ODP. We recommend that you deploy only columnstore replicas in the zone that contains columnstore replicas. When you deploy an ODP, its version must be ODP V4.3.2 or later.

• When you create a tenant, you can specify a columnstore replica in a zone by setting the locality. For more information about how to create a tenant, see CREATE TENANT.

• When you use a columnstore replica, you must configure a routing forwarding strategy and weak-consistency read requests. For more information, see Use a columnstore replica.