presto + build presto cluster that will be join to exsiting hadoop cluster

0
votes

we have hadoop cluster that contain all the relevant components/services as

HDFS
YARN
mapreduce
HIVE
Tez
pig
Zookeeper

hadoop clutser contain 3 masters machines and 12 data node machines and 3 kafka

now we want to use presto to run query against data sources ( hadoop cluster / hive )

so we build a new presto cluster as the follwing

1 presto coordinator

8 presto workers

all presto cluster machines are redhat 7.2

now we want to install the presto on all OS

but we are not sure if presto can be installed immodestly after Linux scratch OS

or maybe we need to install something in the middle after the OS and before the presto ?

1
hadoopprestotrino

1 Answers

5
votes

The only requirement for Presto is a Java Virtual Machine (JVM). We recommend installing the latest OpenJDK 11 version, currently 11.0.2. After that, follow the Presto deployment instructions.

Python is required for the launcher (the script that starts the JVM), but this is normally available on a typical Linux distribution.