https://github.com/nleiva/aws-testbed
Create reproducible Linux environments in the cloud
https://github.com/nleiva/aws-testbed
ansible aws containerlab docker fedora ubuntu
Last synced: 2 months ago
JSON representation
Create reproducible Linux environments in the cloud
- Host: GitHub
- URL: https://github.com/nleiva/aws-testbed
- Owner: nleiva
- Created: 2021-06-22T18:12:30.000Z (almost 4 years ago)
- Default Branch: main
- Last Pushed: 2021-09-17T17:30:22.000Z (over 3 years ago)
- Last Synced: 2024-12-28T12:15:46.936Z (4 months ago)
- Topics: ansible, aws, containerlab, docker, fedora, ubuntu
- Language: Shell
- Homepage:
- Size: 923 KB
- Stars: 3
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Automating a testing environment setup on AWS
[](https://github.com/nleiva/aws-testbed/actions/workflows/ansible-lint.yml)
Cloud resources are expensive, you only want to pay for them while your are using them.
If you only run tests or work on demos for a couple of hours per day, it makes a sense to shutdown all resources for the rest of the day to save on costs.
If you want to resume work or start fresh, your setup should ideally be re-created consistently and then destroyed with a single click of a button.
That's where automation comes in.
A [Playbook](create-EC2-testbed.yml) example is included in this repository to illustrate how a Testbed environment can be automatically be provisioned for a particular use-case.
In this scenario, we create a virtual machine (VM) in AWS that is ready to run containerized multi-vendor network topologies.
The diagram below is a representation of the end-goal, that shows the AWS resources that need to be present otherwise created.
It also hints about Software requirements that will be installed, as well as providing pre-built network topology samples.
More details -> [How to consistently run temporary workloads on AWS and save money](https://nleiva.medium.com/how-you-can-consistently-run-temporary-workloads-in-the-cloud-37140b4b5e55)
## Requirements
Ansible 2.9+ needs to be [installed](https://docs.ansible.com/ansible/latest/installation_guide/intro_installation.html#installing-ansible-with-pip) in your computer to run this example. [Python3](https://wiki.python.org/moin/BeginnersGuide/Download) as well.
```bash
python -m pip install --user ansible
```
### Python libraries
As we will interact with AWS, we need a couple of Python libraries to be present in the system.
```bash
pip install --user -r requirements.txt
```
### Ansible Collections
We will also need the Ansible [Amazon AWS Collection](https://github.com/ansible-collections/amazon.aws#amazon-aws-collection).
```bash
ansible-galaxy collection install -r collections/requirements.yml
```
## Creating the testbed
Follow these steps to provision the testbed.
1. Clone this repository: `git clone https://github.com/nleiva/aws-testbed.git`
2. Make your [AWS account credentials](https://docs.aws.amazon.com/general/latest/gr/aws-sec-cred-types.html#access-keys-and-secret-access-keys) (`AWS_ACCESS_KEY_ID` and `AWS_SECRET_ACCESS_KEY`) available as environment variables (`export`).
```bash
export AWS_ACCESS_KEY_ID='...'
export AWS_SECRET_ACCESS_KEY='...'
```
3. Run the [Playbook](create-EC2-testbed.yml) and wait a couple of minutes while a VM is provisioned and the software is installed:
```bash
⇨ ansible-playbook create-EC2-testbed.yml -v
TASK [Print out SSH access details] ***********************************************************************************************************************************************************
ok: [testbed-fedora34] => {
"msg": "ssh -i testbed-private.pem [email protected]"
}
RUNNING HANDLER [configure_instance : Reboot machine] *****************************************************************************************************************************************
changed: [testbed-fedora34] => {"changed": true, "elapsed": 22, "rebooted": true}
PLAY RECAP ************************************************************************************************************************************************************************************
localhost : ok=25 changed=7 unreachable=0 failed=0 skipped=1 rescued=0 ignored=0
testbed-fedora34 : ok=33 changed=22 unreachable=0 failed=0 skipped=0 rescued=0 ignored=0
```
You can now access the VM as displayed in the logs, for example: `ssh -i testbed-private.pem [email protected]`.
### Linux distribution
You can select either Fedora (34) or Ubuntu (20.04) by passing the variable `aws_distro` to the [Playbook](create-EC2-testbed.yml). By default it selects `fedora`.
You would run the [Playbook](create-EC2-testbed.yml) as follows, if you preferred to run an `ubuntu` machine.
```bash
ansible-playbook create-EC2-testbed.yml -v --extra-vars "aws_distro=ubuntu"
```
### Instance type
You can select any instance type you prefer, based on your vCPU/Memory requirements, and price constrains. By default it selects `t2.micro`. Check out [On-Demand Plans for Amazon EC2](https://aws.amazon.com/ec2/pricing/on-demand/) for reference. Some examples:
Instance name | On-Demand hourly rate | vCPU | Memory
--- | --- | --- | ---
t3.medium | $0.0416 | 2 | 4 GiB
m5.large | $0.096 | 2 | 8 GiB
t3.xlarge | $0.1664 | 4 | 16 GiB
m4.xlarge | $0.2 | 4 | 16 GiB
r5.xlarge | $0.252 | 4 | 32 GiB
m5.2xlarge | $0.384 | 8 | 32 GiB
r5.2xlarge | $0.504 | 8 | 64 GiB
c5.metal | $4.08 | 96 | 192 GiB
You would run the [Playbook](create-EC2-testbed.yml) as follows, if you preferred to run a `m5.large` instance.
```bash
ansible-playbook create-EC2-testbed.yml -v --extra-vars "instance_type=m5.large"
```
To run the Containerlab's Firecracker micro-VMs [mode of operation](https://containerlab.srlinux.dev/manual/kinds/cvx/#mode-of-operation), you need nested virtualization support, which regular AWS instances don't provide ([Cloud Provider Instances with KVM support](https://ignite.readthedocs.io/en/stable/cloudprovider/)), so you would need to select a [bare metal instance type](https://aws.amazon.com/about-aws/whats-new/2018/05/announcing-general-availability-of-amazon-ec2-bare-metal-instances/).
Note1: `a1.metal` is the cheapest bare metal instance. However, its architecture is `arm64`, not `x86_64` (it affect the image type you select).
Note2: Rebooting a [bare metal instance](https://aws.amazon.com/about-aws/whats-new/2018/05/announcing-general-availability-of-amazon-ec2-bare-metal-instances/) takes considerably longer than a regular EC2 instance.
### Volume type
The volume type is also configurable, check out [Amazon EBS pricing](https://aws.amazon.com/ebs/pricing/). The default is `gp2`.
Volume Type | Price
--- | ---
General Purpose SSD (gp3) - Storage | $0.08/GB-month
General Purpose SSD (gp2) Volumes | $0.10 per GB-month of provisioned storage
To select a different type, use the variable `volume_type`.
```bash
ansible-playbook create-EC2-testbed.yml -v --extra-vars "instance_type=m5.large volume_type=gp3"
```
## Running a network topology
Once in the VM, you can run any of the examples of the [lab folder](lab) in the VM's `$HOME` directory. [Containerlab](https://github.com/srl-labs/containerlab) is already installed and does all the magic here. For example, a simple topology with two [FRR](https://frrouting.org/) routers connected back-to-back as described in [lab/frr](lab/frr/topology.yml) can be instantiated as follows:
```bash
cd lab/frr
sudo clab deploy --topo topology.yml
```
The routers are [pre-configured](lab/frr/router1/frr.cfg) with a BGP session running between them. To access router1 for example:
```bash
docker exec -it clab-mylab-router1 vtysh
```
[More details](lab/README.md) about Topology creation.
## Deleting the test VM
As important as creating the VM, is being able to delete it. You can do this by running:
```bash
ansible-playbook delete-EC2-testbed.yml -v
```
For a specific Linux distribution:
```bash
ansible-playbook delete-EC2-testbed.yml -v --extra-vars "aws_distro=ubuntu"
```
## Deleting all remaining AWS resources
This is optional. You can remove VPC, Subnet, Security Group, SSH Key Pair, and Internet Gateway, by setting the variable `delete` to `true` in the previous Playbook. For example:
```bash
ansible-playbook delete-EC2-testbed.yml -v --extra-vars "aws_distro=ubuntu delete=true"
```
If you used a non-default prefix, then:
```bash
ansible-playbook delete-EC2-testbed.yml -v -e "aws_prefix=another delete=true"
```
## Alternative Ansible execution (optional)
### Ansible Navigator
You can alternatively run it with Ansible Navigator:
```bash
ansible-navigator run create-EC2-testbed.yml --ee false --penv AWS_ACCESS_KEY_ID,AWS_SECRET_ACCESS_KEY
```
### Run in Playbook in an Execution Environment
#### Create Execution Environment image with Ansible builder
Image details in [execution-environment.yml](execution-environment.yml). This is optional, you can pull the image from [quay.io/nleiva/aws_ee_image](https://quay.io/repository/nleiva/aws_ee_image).
```bash
ansible-builder build -t aws_ee_image
```
#### Run Execution Environment
```bash
ansible-navigator run create-EC2-testbed.yml --extra-vars "instance_type=t3.medium"
```