Templates for VMworld 2017 Demo App on AWS

Software required to use these templates: Ansible, Ansible EC2 dynamic inventory script (included in repository), Terraform

Limitations

• You'll need to ensure the correct AMIs are available in the region where you want to turn up the application. Currently the AMIs are available in us-west-1 and us-east-1 only.

Instructions for Use

1. Clone this repository to your local system.

2. Create a file named terraform.tfvars in the repository (don't worry, Git will ignore it) and populate it with values for the following variables:

   akey (your AWS access key) skey (your AWS secret access key) awsregion (the AWS region to use; currently, only us-west-1 or us-east-1) keypair (the name of an AWS SSH keypair to inject into instances) vpcname (the name you'd like assigned to the new VPC) cidr (the network block without subnet mask)

3. Run terraform validate to ensure there are no errors in the Terraform configuration. If any errors are reported, fix them before proceeding.

4. Run terraform plan to evaluate the current infrastructure and prepare a plan for creating the desired configuration.

5. Run terraform apply to create the desired infrastructure.

6. Few steps prior to configuring the new instances:

   *Obtain the VPC ID from the terraform state file (look for vpc_id in the file - i.e. "vpc-88e58aec").
   *Add the line instance_filters = vpc-id=<value> to the hosts/ec2.ini file. This allows the env to only see that VPC. *in ec2.ini, change the cache_path line to remove the ~/ from the beginning of the path. This will ensure a unique cache for each environment *in the ansible.cfg file, there is a reference to the SSH private key it should use. Change this to match the SSH keypair you’re injecting into the instances.

7. When Terraform is completed, wait a couple of minutes and then run hosts/ec2.py --refresh-cache to refresh the dynamic Ansible inventory.

8. Run ansible-playbook configure.yml to configure the EC2 instances with the dynamically-assigned IP addresses.

*** SKIP to Step 10 if you plan on using RDS for the DB Backend ***

9. Restart services:

   • Restart HAProxy on the Web tier instances.

     • The HAProxy instances have an issue - service haproxy restart DOES NOT work
     • use the following: sudo ps aux | grep haproxy - note the process id sudo kill xxx copy the ps line for haproxy from the output and use sudo to rerun it it should be something like: sudo /usr/sbin/haproxy -f /etc/haproxy/haproxy.cfg -D -p /var/run/haproxy.pid

   • Restart Django on the App tier instances.

     • sudo ps aux | grep manage - note all the process ids - should be three
     • sudo kill xxx xxx xxx
     • in the ubuntu folder run ./startdjango.sh

   • Restart HAProxy on the DB Load Balancer instance. (sudo service haproxy restart works)

You should be ready to go!

10. Create RDS Instance of mySQL

    • Launch DB Instance
    • Step 1 - MySQL / MySQL Community Edition
    • Step 2 - Dev/Test MySQL
    • Step 3
      • Defaults for:
        • DB Engine
        • License Model
        • DB Engine Version
      • Select accordingly for:
        • DB Instance Class - db.t2.micro
        • Multi-AZ Deployment - No
        • Allocated Storage - 25 GB
      • Fill in the following under Settings
        • DB Instance Identifier - #whateveryouwant
        • Master Username - db_app_user
        • Master Password - #whateveryouwant Inspire!123!
        • Confirm Password - #whateveryouwant Inspire!123!
    • Step 4
      • ???

11. Modify Django settings on ALL App Server to point to RDS

    • update fitcycle/fitcycle/settings.py with new database connection details

      DATABASES = { 'default': { 'ENGINE': 'django.db.backends.mysql', 'NAME': 'prospect', 'USER': 'db_app_user', 'PASSWORD': 'whateveryouwant', 'HOST': 'url of RDS instance', --- i.e ccio-east-01.cn3iid7bdnce.us-east-1.rds.amazonaws.com } }

    • Restart Django on the App tier instance.

      • sudo ps aux | grep manage - note all the process ids - should be three
      • sudo kill xxx xxx xxx
      • in the ubuntu folder run ./startdjango.sh

12. Restart Web Server

    • Restart HAProxy on the Web tier instances.
      • The HAProxy instances have an issue - service haproxy restart DOES NOT work
      • use the following: sudo ps aux | grep haproxy - note the process id sudo kill xxx copy the ps line for haproxy from the output and use sudo to rerun it it should be something like: sudo /usr/sbin/haproxy -f /etc/haproxy/haproxy.cfg -D -p /var/run/haproxy.pid

13. Create tables and clear database on RDS

    • sudo mysql -u db_app_user -p -h ccio-west-01.cy1up9w2u6z9.us-west-1.rds.amazonaws.com
      • use prospect;
        • paste and execute the code in the "sqldump.txt"
      • truncate polls_prospect;
      • alter table polls_prospect AUTO_INCREMENT = 1;
      • select * from polls_prospect;
    • result should be "empty"

14. Delete unused instances from the vpc

    • DB Load Balancer
    • Db1
    • Db2

You should be ready to go!

Using Regions Other than US-East-1

In the file testvpc.tf there is a section that looks like this:

variable "images" {
  type = "map"
  default = {
    web="ami-a7b886b1"
    mngt="ami-dfb886c9"
    db="ami-1dbb850b"
    dblb="ami-71bd8367"
    app="ami-75bd8363"
  }
}

The AMI IDs currently in testvpc.tf are for US-East-1; if you want to use a different region, copy the AMIs to that region and adjust the AMI IDs in this section accordingly.

Running Multiple Instances

To run multiple instances of this application, make multiple copies of this repository and follow the instructions for each.

Destroying the Infrastructure

Simply run terraform destroy to tear down the AWS infrastructure.