Role of ZooKeeper

The ZooKeeper utility provides configuration and state management and distributed coordination services to Dgraph nodes of the Big Data Discovery cluster. It ensures high availability of the query processing by the Dgraph nodes in the cluster.

ZooKeeper is part of the CDH package. CDH package is assumed to be installed on all CDH nodes in the BDD cluster deployment. Even though ZooKeeper is installed on all CDH nodes in the BDD cluster, it may not be running on all of these nodes. To ensure availability of a clustered Dgraph deployment, configure an odd number (at least three) of CDH nodes to run ZooKeeper instances. This will avoid ZooKeeper being a single point of failure.

ZooKeeper has the following characteristics:

It is a shared information repository that provides a set of distributed coordination services. It ensures synchronization, event notification, and coordination between the nodes. The communication and coordination mechanisms continue to work in the case when connections or Dgraph-hosting nodes fail.
It provides communication between the Dgraph-hosting nodes, ensuring that if one of these nodes fails, requests are sent to other active Dgraph nodes, until the node rejoins the Dgraph cluster (this is true if more than two instances of ZooKeeper are running in the deployment).
It provides communication between Dgraph nodes. It controls the election of the Dgraph leader node in the Dgraph cluster, in case the current leader Dgraph node fails. The newly-elected leader Dgraph node identifies the most recent version of the index, and, using ZooKeeper, informs other nodes of the current version of the index.

To summarize, in order to run, ZooKeeper requires a majority of its hosting nodes to be active. Therefore, it is recommended that ZooKeeper runs on an odd number (at least three) of the CDH nodes in the deployed Big Data Discovery cluster. You can ensure this during the installation, when running the deployment script.