Cassandra Image Save

Goal of this project is to create AMI, Vagrant Box and Docker base images that one could use to deploy Cassandra.

The docker image is hosted on docker hub (see DockerHub). The vagrant box is hosted on Atlas see (Atlas). The source code is hosted on GitHub (see GitHub).

Key features of image, AMI, Vagrant box

• Uses JBOD instead of RAID (not done)
• Uses ergonomics to configure Cassandra based on deployment environment (preliminary work done)
• Sets up security if requested (not done)
• Sets up TLS/SSL if requested (not done)
• Sets up HD encryption if requested (not done, or use encrypted EBS instances)
• Sets up users (not done)
• Installs Cassandra as a systemd service (not done)
• Allows cloud deploy (not done)
• Monitoring (CloudWatch, InfluxDB) (not done)
• Log aggregation (CloudWatch, ELK) (not done)
• Installs JEMalloc and configures Cassandra to use off heap no JVM (done)
• Install JNA (done)
• Configures OS (Linux) to be performant (done)

Create a vagrant box

git clone https://github.com/cloudurable/cassandra-image.git
cd cassandra-image 
vagrant up 

# Connect to vagrant box
cqlsh localhost 19042

Connect to vagrant image

vagrant ssh

Create a docker image

git clone https://github.com/cloudurable/cassandra-image.git
cd cassandra-image 
bin/start-image.sh

# Connect to docker image
cqlsh localhost 29042

Provisioning

We use packer and vagrant to create images.

Running setup scripts

## cd ~; mkdir github; cd github; git clone https://github.com/cloudurable/cassandra-image
$ cd ~/github/cassandra-image
$ pwd
~/github/cassandra-image
## Setup keys
$ bin/setupkeys-cassandra-security.sh
## Download binaries
$ bin/prepare_binaries.sh
## Bring Vagrant cluster up
$ vagrant up

Working with ansible from bastion

$ vagrant ssh bastion

First setup ssh-agent and add keys to it.

Start ssh-agent and add keys

$ ssh-agent bash
$ ssh-add ~/.ssh/test_rsa

Ansible Ping server

$ ansible node0 -m ping

Output

node0 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}

Ansible Ping servers

$ ansible nodes  -m ping

Output

node0 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}
node2 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}
node1 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}

Setting up my MacOSX to run Ansible against instances

Move to the where you checked out the project.

cd ~/github/cassandra-image

Add bastion, node0, etc. to /etc/hosts

$ cat /etc/hosts

### Used for ansible/ vagrant
192.168.50.20  bastion
192.168.50.4  node0
192.168.50.5  node1
192.168.50.6  node2
192.168.50.7  node3
192.168.50.8  node4
192.168.50.9  node5

Add keys to known_hosts to avoid prompts

$ ssh-keyscan node0 node1 node2  >> ~/.ssh/known_hosts

Start ssh-agent and add keys

$ ssh-agent bash
$ ssh-add ~/.ssh/test_rsa

Notice the ansible.cfg file and inventory.ini file in the project dir

$ cd ~/github/cassandra-image

$ cat ansible.cfg 
[defaults]
hostfile = inventory.ini

cat inventory.ini 
[nodes]
node0 ansible_user=vagrant
node1 ansible_user=vagrant
node2 ansible_user=vagrant

Ansible will use these.

Ansible Ping server

$ ansible node0 -m ping

Output

node0 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}

Ansible Ping servers

$ ansible nodes  -m ping

Output

node0 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}
node2 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}
node1 | SUCCESS => {
    "changed": false, 
    "ping": "pong"
}

More details to follow

About us

Cloudurable provides AMIs, cloudformation templates and monitoring tools to support Cassandra in production running in EC2. We also teach advanced Cassandra courses which teaches how one could develop, support and deploy Cassandra to production in AWS EC2.