Installation
This section describes the Installation procedure for the Privitar Connector.
The Connector consists of two separate plug-ins, but they are both contained in single jar:
Masking plug-in connector. (This plug-in is called,
PrivitarSinkConnector)Unmasking plug-in connector. (This plug-in is called,
UnmaskPrivitarSinkConnector)
This document will refer to the Privitar Connector plug-in to mean both plug-ins. Where necessary, it will reference a specific plug-in.
Pre-requisites
Installed Kafka 1.1+ or Confluent 4.1+ (see supported versions in Introduction)
Installed Kafka Connect 1.1+ (included from Confluent 4.1+).
Installation procedure
Create a folder in the plug-in directory used by the Kafka worker node.
Add the Privitar Connector plug-in to the new folder.
Add the necessary Token vault drivers into the new folder.
Restart the Connect worker node.
These steps need to be performed on every Kafka Connect worker node that will be using the Privitar plug-in.
Create a folder
Create a folder to store the Privitar connector plug-in. The location for plug-ins is defined by the plugin.path variable in the Kafka worker configuration file. For example:
plugin.path=/usr/local/share/kafka/plugins
Create a folder called, for example privitar-kafka-connect in that folder. This makes the path to the Privitar connector plug-in:
/usr/local/share/kafka/plugins/privitar-kafka-connect
Add the plug-in
Copy the Privitar connector plug-in JAR file into the new folder. The Privitar plug-in is supplied as an uber-JAR, so all the dependencies it requires are included in the jar - apart from the drivers that will be needed to connect to the Privitar Token vault.
Add Token Vault drivers
Drivers are required by the Privitar plug-in to connect to the Privitar Token vault. These drivers need to be added to the new Privitar plug-in folder. The drivers to include will be specific to the type of database you are using to store the Privitar Token vault.
Restart Worker nodes
Restart all Kafka Connect worker nodes from the console.
When the worker nodes are restarted they will discover all connectors, transforms, and/or converters defined within the Privitar plug-in. When we use a connector, transform, or converter, the Kafka Connect worker loads the classes from the respective plug-in first, followed by the Kafka Connect runtime and Java libraries. Kafka Connect explicitly avoids all of the libraries in other plug-ins and prevents conflicts, making it very easy to use connectors and transforms developed independently by different providers.