Data Loading and Synchronization
GridGain Hadoop Connector supports several ways to perform initial data loading from Hadoop to GridGain and synchronization between the two stores.
Loading With GridGain Spark Loader
Use GridGain Spark Loader if you already have Spark in your architecture. Using Spark is straightforward and one of the fastest ways to load data from Hadoop to GridGain.
Loading and Schema Import with Hive Store
Hive Store better suits deployments without Spark and it can also import a Hive schema and convert it to a GridGain configuration via GridGain Nebula. Refer to Load and Sync With Hive Store for more details.
Bi-Directional Synchronization With Apache Sqoop
The Sqoop integration can be used to synchronize two stores. The synchronization can be set to send updates from GridGain to Hadoop, from Hadoop to GridGain, or in a bi-directional mode.
© 2024 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.