Vagrant Cassandra

This is a sample project to show how to get a simple Cassandra cluster up and running with Vagrant. It is not secure or production ready.

Acknowledgement

It is a modified vagrant-cassandra project.

Prerequisites

First usage

(install latest Ruby)
vagrant plugin install vagrant-omnibus
gem install librarian-chef
librarian-chef install
vagrant up

Instead of vagrant up you can use provision-parallel script for faster provisioning. However, sometimes there is a race condition which results in failure in provisioning one of the VMs. You can then retry provisioning only that failed VM by executing vagrant up <node> --provision.

After that you should be able to interact with the Cassandra nodes.

vagrant@node2:~$ cqlsh
Connected to Test Cluster at 127.0.0.1:9042.
[cqlsh 5.0.1 | Cassandra 2.1.20 | CQL spec 3.2.1 | Native protocol v3]
Use HELP for help.
cqlsh>

vagrant@node2:~$ nodetool status
Datacenter: datacenter1
=======================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
--  Address         Load       Tokens  Owns (effective)  Host ID                               Rack
UN  192.168.200.11  54.88 KB   256     65.6%             9d3e59d5-27b8-4d45-9ac1-1e1b3a9c6df1  rack1
UN  192.168.200.13  111.8 KB   256     66.9%             c8e91f28-42d2-452e-8b27-311e70b34987  rack1
UN  192.168.200.12  122.65 KB  256     67.4%             f5ceccc8-3e40-406b-9a7c-92c1b6cb4bb5  rack1

Basic Vagrant commands

vagrant up - boots up all VMs (and provisions them if it's their first boot)
vagrant up <node> - boots up only the specified node (there are node1, node2 and node3)
vagrant halt - gracefully shutdowns all VMs
vagrant halt <node> - gracefully shutdowns only the specified node
vagrant provision - applies changes from Vagrantfile
vagrant ssh <node> - logs to the specified node via SSH

Partitions may be simulated by SSH-ing to particular node(s) and bringing down eth1 interface. Unplugging the network cable from VirtualBox GUI or VBoxManage should yield the same result.

Changelog

2018-12-08

update Cassandra from 2.1.0 to 2.1.20
update Oracle JDK7 to Oracle JDK 8
update java cookbook
add cassandra binaries to PATH
add parallel provisioning scripts for Windows and Unix

2014-10-11

upgrade cassandra-chef-cookbook to 2.8.0
set cassandra version to 2.1.0
switch back to cassandra::tarball
remove cassandra restart post provisioner

2014-09-13

upgrade Ubuntu from 12.04 to 14.04
upgrade Chef to from 11.8.2 to 11.16.0
upgrade Cassandra from 2.0.3 to 2.0.9
upgrade java from 1.7.0_51 to 1.7.0_67
increased RAM from 512 to 1024
switch from cassandra::tarball to cassandra::datastax
add apache2 license
add explicit versions to Cheffile
add ssl_verify_mode = :verify_peer configuration to remove chef warning

kwencel / vagrant-cassandra Goto Github PK