The following section provides an overview of how to use Presto.

E-MapReduce versions 2.0 and later support Presto. Presto can be used in E-MapReduce by checking the Presto software box when you select a mirror image.

After you create a cluster, log on to the master node. The Presto software can be found in the /usr/lib/presto-current directory, and the PrestoServer processes can be viewed using the jps command.

Presto processes can be divided into coordinator and worker processes. The coordinator process is started on the master node (the HA cluster is the master node whose hostname starts with emr-header-1), and the worker process is started on the core node. The service process configuration can be found in the /usr/lib/presto-current/etc directory. Coordinator uses coordinator-config.properties, whereas worker uses worker-config.properties. Other configuration files are shared. The web port is set as 9090.

By default, Presto services are supported by Hive. You can connect Hive's metastore on the cluster to read Hive table information and query it. The cluster is pre-installed with Presto CLI and can execute the following command to check Hive tables:
presto -server localhost:9090 -catalog hive -schema default -user hadoop -execute 'show tables'
Note There is a delay of several seconds when Hive tables are synchronized.