Running data movement from Hadoop

You can start both import and export operations by running the fdm.sh script from Hadoop.

Before you begin

Ensure that you have installed the IBM® Fast Data Movement software on Hadoop, as described in Installing or upgrading the data movement feature on Hadoop.

During installation, you are asked to provide the paths to Hadoop /lib and /conf files. These details are then used by fdm.sh script when running data movement.

Once you have installed the Fast Data Movement package on Hadoop, you can find fdm.sh in /fastDataMovement/.

Procedure

Depending on whether you are running an import or an export transfer, prepare the XML configuration file using one of the available templates, as described in Preparing configuration XML files. Each file defines a single transfer job, that is single or multiple tables from or to a particular database.
Optionally, you can validate connection configuration as described in Validating data movement configuration.
From the command line, run the fdm.sh and provide the import or export XML configuration file as the -conf parameter. For example:
```
fdm.sh -conf fq-import-conf.xml
```
Tip: You can also overwrite any parameters included in the configuration file with -D parameter:
```
fdm.sh -conf fq-import-conf.xml -D fq.tables=import_test -D
      fq.append.mode=overwrite
```

Results

Data transfer is performed based on the properties that you provided.