Neo4j HA Cluster Deployment Using Ansible

Ansible playbook for Neo4j HA Cluster on AWS/Eucalyptus. The AMI/EMI used should be an Ubuntu Cloud Image. This was tested using Ubuntu 12.04 Precise. This playbook deploys the following:

• a Neo4j HA cluster
• an HA proxy server that serves as a single point to access the Neo4j HA cluster

Pre-reqs

• Set up Ansible following the documentation here - http://ansible.cc/docs/gettingstarted.html#getting-ansible

• Set up /etc/ansible/hosts with the following information:

[local]
127.0.0.1

• Populate vars/ec2-config with either Eucalyptus/AWS information. vars/ec2-config contains the following variables:

keypair: <EC2/Eucalyptus Keypair>
ec2_access_key: <EC2_ACCESS_KEY>
ec2_secret_key: <EC2_SECRET_KEY>
ec2_url: <EC2_URL>
instance_type: m1.small
security_group: <AWS/Eucalyptus Security Group>
image: <AMI/EMI>

Any config changes for HA Proxy or the Neo4j cluster can be found under the vars directory.

Usage

After cloning the repository, and populating the config files under vars directory, run the following command:

 ansible-playbook ansible-neo4j-cluster/neo4j-cluster.yml
 --private-key=<AWS/Eucalyptus Private Key file> --extra-vars "node_count=3"

To change the number of nodes for the cluster, change node_code to equal that number (e.g. node_code=7).

After the playbook finishes, there will be an URL provided to access the cluster - similar to the example below:

TASK: [Display HAProxy URL] *********************
changed: [23.22.248.75] => {"changed": true, "cmd": 
"echo \"HAProxy URL for Neo4j - http://ec2-23-22-248-75.compute-1.amazonaws.com/webadmin/#/info/org.neo4j/High%20Availability/\" ", "delta": "0:00:00.006835", "end": "2013-03-30 19:54:31.104320", 
"rc": 0, "start": "2013-03-30 19:54:31.097485", "stderr": "", 
"stdout": "HAProxy URL for Neo4j - http://ec2-23-22-248-75.compute-1.amazonaws.com/webadmin/#/info/org.neo4j/High%20Availability/"}

To get the status of the Neo4j cluster, run the following command:

curl -H "Content-Type:application/json" -d '["org.neo4j:*"]' 
http://ec2-23-22-248-75.compute-1.amazonaws.com/db/manage/server/jmx/query

For more information, please refer to the documentation for configuring/managing the Neo4j cluster - http://docs.neo4j.org/chunked/milestone/operations.html

Cluster Membership

As described in the following URL:

http://docs.neo4j.org/chunked/milestone/ha-configuration.html

Cluster membership is determined by using the ha.initial_hosts variable in the neo4j.properties file on each node. This variable contains all the members of the cluster. Other variables that are defined on each node use the instance's launch index to derive unique values. The instance ami launch index is obtained to through the metadata service. The following variables in the template files use the ami launch index:

• neo4j-server.properties.j2
• org.neo4j.server.webserver.port={{ ansible_ec2_ami_launch_index|int + 7474 }}
• org.neo4j.server.webserver.https.port={{ ansible_ec2_ami_launch_index|int + 7573 }}
• neo4j-wrapper.conf.j2
• wrapper.java.additional.3=-Dcom.sun.management.jmxremote.port={{ ansible_ec2_ami_launch_index|int + 3637 }}
• neo4j.properties.j2
• online_backup_server={{ ansible_ec2_local_ipv4 ~ ":" }}{{ ansible_ec2_ami_launch_index|int + 6362 }}
• ha.server_id={{ ansible_ec2_ami_launch_index|int + 1 }}
• ha.server={{ ansible_ec2_local_ipv4 ~ ":" }}{{ ansible_ec2_ami_launch_index|int + 6001 }}
• ha.cluster_server={{ ansible_ec2_local_ipv4 ~ ":" }}{{ ansible_ec2_ami_launch_index|int + 5001 }}
• remote_shell_port={{ ansible_ec2_ami_launch_index|int + 1234 }}

For more information regarding Neo4j HA setup, please reference the following tutorial - http://docs.neo4j.org/chunked/milestone/ha-setup-tutorial.html.