Replicating data into Teradata
SymmetricDS now supports capturing changes on a source database in order to load them into Teradata. With a few simple steps you can setup a Teradata load only node in SymmetricDS that will support bulk loading existing data as well as continuous replication of changes from the source database into Teradata.
Begin by setting up the source node to one of the supported database dialects in SymmetricDS.
https://www.jumpmind.com/products/symmetricds/symmetricds-overview
Load Only Nodes
Once you have your source database configured in SymmetricDS and have selected some tables for replication your ready to setup the Teradata node. Since SymmetricDS 3.9 support for a “load only” node was provided. This allows SymmetricDS to replicate changes to this database in one direction. This would support use cases where your data is only moving from source to target and not bidirectionally.
https://www.jumpmind.com/downloads/symmetricds/doc/3.9/html/user-guide.html#_load_only_node
Advantages
There are two main advantages to this approach. First, the load only node will not need any runtime tables or objects that SymmetricDS would create on a normal node. Secondly these nodes are very lightweight as they are designed to send changes through JDBC to the target. Some load only implementations also provide support for “bulk” loading as is the case for Teradata.
JDBC Driver
You will need to provide the two Teradata JDBC driver files into the /lib folder of SymmetricDS installation.
- terajdbc4.jar
- tdgssconfig.jar
If your using the Professional version you will upload through the Add Driver button during the database setup. In early 3.9 versions you could only upload a single jar file so you will still need to add the second jar in the /lib folder
Finish setting up the node and your all set for change capture.
Bulk Loading
There is a Teradata bulk loader provided in SymmetricDS that uses the Teradata fastloadcsv. To turn on the bulk loader simply set all channels that are sending data to Teradata to use the data loader type: “bulk”.
Note: Teradata tables must be empty to use the bulk loader
Review
You are now prepared to initially load data from your source database into Teradata. Also changes will be captured and sent to your Teradata database as they occur in the source to keep them in sync.
