Skip to main content
Version: 3.6 (unsupported)

Back up and restore ScalarDB or ScalarDL data in a Kubernetes environment

This guide explains how to backup and restore ScalarDB or ScalarDL data in a Kubernetes environment. Please note that this guide assumes that you are using a managed database from a cloud services provider as the backend database for ScalarDB or ScalarDL. The following is a list of the managed databases that this guide assumes you might be using:

  • NoSQL: does not support transactions
    • Amazon DynamoDB
    • Azure Cosmos DB for NoSQL
  • Relational database (RDB): supports transactions
    • Amazon RDS
      • MySQL
      • Oracle
      • PostgreSQL
      • SQL Server
    • Amazon Aurora
      • MySQL
      • PostgreSQL
    • Azure Database
      • MySQL
      • PostgreSQL

For details on how to back up and restore databases used with ScalarDB in a transactionally consistent way, see A Guide on How to Backup and Restore Databases Used Through ScalarDB.

Perform a backup​

Confirm the type of database and number of databases you are using​

How you perform backup and restore depends on the type of database (NoSQL or RDB) and the number of databases you are using.

NoSQL or multiple databases​

If you are using a NoSQL database, or if you have two or more databases that the Multi-storage Transactions or Two-phase Commit Transactions feature uses, please see Back up a NoSQL database in a Kubernetes environment for details on how to perform a backup.

Single RDB​

If you are using a single RDB, please see Back up an RDB in a Kubernetes environment for details on how to perform a backup.

If you have two or more RDBs that the Multi-storage Transactions or Two-phase Commit Transactions feature uses, you must follow the instructions in Back up a NoSQL database in a Kubernetes environment instead.

Restore a database​

For details on how to restore data from a managed database, please see Restore databases in a Kubernetes environment.