Demo of Docker/Kubernetes on AWS

Preparation

To deploy HPCC cluster on Kubernetes/AWS we need AWS Client and Kubernetes package

Install and Configure AWS Client

Install AWS Clinet as http://docs.aws.amazon.com/cli/latest/userguide/installing.html#install-bundle-other-os

Configure AWS client with Access Key Secrete Access Key and default region:

aws configure

AWS Access Key ID [****************]: 
AWS Secret Access Key [****************4ifT]: 
Default region name [ap-southeast-1]: 
Default output format [None]: 

To test it run:

aws ec2 describe-regions

Install Kubernetes

Reference http://kubernetes.io/docs/getting-started-guides/aws/

Download the Kubernetes to local directory:

export KUBERNETES_PROVIDER=aws; wget -q -O - https://get.k8s.io | bash

Installed Kubernetes should be in Kubernetes in current directory.

It is helpful to create env file with:

export KUBERNETES_PROVIDER=aws
export KUBE_AWS_ZONE=ap-southeast-1b #change this to your select zone
export PATH=<Kubernetes install directory>/platforms/darwin/amd64:$PATH
# Uncomment following if you want to setup a namespace instead of using default one: "default"
#export CONTEXT=$(kubectl config view | grep current-context | awk '{print $2}')
#kubectl config set-context $CONTEXT --namespace=<your namespace name, for example,hpcc-kube> > /dev/null 2>&1

# Change the size based on your need. Following default will set master to m3.medium and node to t2.micro
export MASTER_SIZE=
export NODE_SIZE=
export NUM_NODES=4
export AWS_S3_REGION=ap-southeast-1 #It is better to match KUBE_AWS_ZONE

# Additional settings can be found in Kubernetes/cluster/aws/config-default.sh

You can source this env file before run any Kubernetes command or add it to profile

Start Kubernetes on AWS

Make sure you source the env file created above and go to Kubernetes/cluster and run

./kube-up.sh

It may ask you to install addional packages.

When it finish you should see the master and nodes (minion) in AWS Console

Quick Start

Create a Cluster

Go to demo-on-aws/bin directory and run

./create-all.sh

Verify all pod ready:

kubectl get pods

Configure the Cluster

Login to hpcc-ansible node:

kubectl exec -i -t hpcc-ansible -- bash -il

export TERM=xterm
stty rows 50 cols 120

Go to /opt/hpcc-tools and run

./config_hpcc.sh

This should collect ips and generate environment.xml and start HPCC cluster with Ansible playbook

Even "host_key_checking=false" is set in /etc/ansible/ansible.cfg but it seems not work. use "export ANSIBLE_HOST_KEY_CHECKING=False" instead.

Access ECLWatch

Get esp load balancer ip

kubectl get service esp

NAME      CLUSTER-IP     EXTERNAL-IP        PORT(S)    AGE
esp       10.0.150.205   ae45d4af45995...   8010/TCP   25m


kubectl get service esp -o json | grep ae45d4af45995
                    "hostname": "ae45d4af4599511e698bd02c3e1be8c5-90448396.ap-southeast-1.elb.amazonaws.com"

host ae45d4af4599511e698bd02c3e1be8c5-90448396.ap-southeast-1.elb.amazonaws.com

ae45d4af4599511e698bd02c3e1be8c5-90448396.ap-southeast-1.elb.amazonaws.com has address 52.220.16.128
ae45d4af4599511e698bd02c3e1be8c5-90448396.ap-southeast-1.elb.amazonaws.com has address 54.179.135.114

You should use either ip with port 8010 to access ECLWatch

Destroy the Cluster

./destroy-all.sh