Quick Start

Prerequisites

• Docker Desktop
  • ZooKage officially supports macOS
  • Some components might not start on a machine with Apple silicon + Rosetta 2
• 8+GB RAM
  • You need more when you launch more components
• Kubernetes is enabled

Instructions

Set up a Hadoop cluster

$ git clone --branch v0.2.5 git@github.com:zookage/zookage.git
$ cd zookage
$ ./bin/up
namespace/zookage created
job.batch/package-hadoop created
job.batch/package-hive created
job.batch/package-tez created
...
All resources are ready!

Log in to a client container

With ./bin/ssh, you can ssh into client-node. Various command-line tools such as hdfs, yarn, beeline, etc. are available.

$ ./bin/ssh
zookage@client-node-0:~$

Submit a job

You can run any commands on client-node.

zookage@client-node-0:~$ beeline
...
0: jdbc:hive2://hive-hiveserver2:10000/defaul> SELECT 1;
...
+------+
| _c0  |
+------+
| 1    |
+------+
1 row selected (6.432 seconds)

Access web UI

You can access various web UIs such as YARN ResourceManager through your browser.

The list of available endpoints is here.