Point-in-Time Recovery
By using the point in time recovery (PITR) feature, you can restore cluster status to any point above the low watermark point.
Limitations
-
Point in time recovery can only be performed for data above the low watermark (600000ms by default). Older data can only be recovered by restoring a snapshot.
-
Recovery of large amounts of data may require additional storage space, as both versions will be temporarily kept available in case the recovery fails.
Performing Point in Time Recovery
To start point in time recovery, use the recovery tables start command and specify the table or tables to recover. Table names can be schema-qualified.
recovery tables start --tables Person,PUBLIC.Accounts --timestamp 2024-09-10T10:53:00+01:00
Once replication is started, it will return the operation ID.
Monitoring Recovery
Point in time recovery may take a long time. You can monitor current replication status by using the recovery tables state command and passing the ID of data recovery to it:
recovery tables state 944e4abe-b655-4283-9d78-449b5186a7ac
© 2025 GridGain Systems, Inc. All Rights Reserved. Privacy Policy | Legal Notices. GridGain® is a registered trademark of GridGain Systems, Inc.
Apache, Apache Ignite, the Apache feather and the Apache Ignite logo are either registered trademarks or trademarks of The Apache Software Foundation.