aws-vertica

Python script using Fabric and Boto to manage and deploy a Vertica cluster into AWS.

It creates a standalone VPC environment. One node is designated as the Internet Gateway and the bootstrap node where the administrative commands are run.

Prerequisites

Python Packages:

    pip install fabric
    pip install boto

Boto Configuration File:

   /etc/boto.cfg or ~/.boto
   [Credentials]
   aws_access_key_id=<your_key>
   aws_secret_access_key=<your_key>

Other Config

Copy .fabricrc.sample to .fabricrc, and update the config as follows:

• region: your ec2 region
• ami: a CentOS 6 AMI either PV or HVM depending on your instance_type
• instance_type: ec2 instance type. take note of whether it supports PV or HVM, and adjust your ami accordingly.
• cluster_name: name you want your cluster tagged as
• key_pair: the name of your ec2 key pair
• key_filename: identity file to ssh into ec2
• local_public_key: your public key to set up passwordless ssh
• use_community_edition_license: 1 or 0, whether to use the community license or not
• local_license_path: local path to your vertica license, that will be copied to your nodes
• db_user, db_name, db_pw: user, database name, and password for vertica
• vertica_rpm_s3_url: upload the Vertica rpm you donwloaded from my.vertica.com to s3 and put the s3:// url here.

Commands

check whats going on

       fab -c .fabricrc print_status

deploy a new cluster

      fab -c .fabricrc deploy_cluster:total_nodes=3

deploy a new cluster using an existing elastic ip for bootstrap/gateway instance

      fab -c .fabricrc deploy_cluster:total_nodes=3,eip_allocation_id=eipalloc-xxxxxx

deploy to an existing vpc cluster, it will consider the gateway node to be the bootstrap if there are existing nodes in the cluster, it will attempt to bring the number of nodes in the cluster to total_nodes

      fab -c .fabricrc deploy_cluster:total_nodes=3,vpc_id=vpc-xxxxxxxx

#terminate cluster fab -c .fabricrc terminate_cluster:vpc_id=vpc-xxxxxxxx