Skip to content

Commit

Permalink
Add pre-built AMIs to Terraform and update documentation
Browse files Browse the repository at this point in the history
  • Loading branch information
clong committed Mar 10, 2019
1 parent 5978e1b commit 26140b2
Show file tree
Hide file tree
Showing 15 changed files with 302 additions and 70 deletions.
2 changes: 2 additions & 0 deletions .gitignore
Expand Up @@ -6,3 +6,5 @@ Boxes/*
.DS_Store
Terraform/*/*.tfstate
Terraform/*/.terraform
Terraform/*/*.tfvars
Terraform/*/*.lock.info
17 changes: 0 additions & 17 deletions Terraform/Method1/terraform.tfvars

This file was deleted.

19 changes: 19 additions & 0 deletions Terraform/Pre-Built_AMIs.md
@@ -0,0 +1,19 @@
# Method 1 - Build Locally and Import to AWS

This method uses Terraform to bring DetectionLab infrastructure online by using pre-built shared AMIs.

The supplied Terraform configuration can then be used to create EC2 instances and all requisite networking components.

## Prerequisites
* A machine to build DetectionLab with
* An AWS account
* An AWS user and access keys to use with the AWS CLI
* Optional but recommended: a separate user for Terraform

## Step by step guide

1. [Configure the AWS command line utility](https://docs.aws.amazon.com/polly/latest/dg/setup-aws-cli.html)
2. Copy the file at [/DetectionLab/Terraform/terraform.tfvars.example](./terraform.tfvars.example) to `/DetectionLab/Terraform/terraform.tfvars`
3. In `terraform.tfvars`, provide overrides for the variables specified in [variables.tf](./variables.tf)
4. From the `/DetectionLab/Terraform/` directory, run `terraform init` to setup the initial Terraform configuration
5. Run `terraform apply` to begin the provisioning process
19 changes: 7 additions & 12 deletions Terraform/README.md
@@ -1,21 +1,16 @@
# DetectionLab Terraform

When I considered the possible ways of building DetectionLab using Terraform, two possibilities came to mind:

### Method 1 - Building locally and exporting VMs
The general concept behind this method is to use Virtualbox or VMware to build DetectionLab. You can then use AWS's VM import capabilities to create AMIs based off of the virtual machines. Once that process is complete, the infrastructure can easily be spun up using a Terraform configuration file.
### Method 1 - Building the VMs locally and exporting them to AWS as AMIs
One method for spinning up DetectionLab in AWS is to begin by using Virtualbox or VMware to build DetectionLab locally. You can then use AWS's VM import capabilities to create AMIs based off of the virtual machines. Once that process is complete, the infrastructure can easily be spun up using a Terraform configuration file.

This method has the benefit of allowing users to customize the VMs before importing them to AWS.

The obvious downside is that it still requires local infrastructure to build the lab, and uploading large OVA files to S3 can be extremely time consuming on slower connections.

### Method 2 - Building and deploying in AWS
The alternative to building locally would be to build the lab entirely in AWS. This would mean the Packer builds would need to be modified to generate EBS volumes and the Vagrant provisioning would need to be modified to support cloud infrastructure. Virtualbox and VMware-based builds benefit from things like virtual machine guest tools for file sharing, which are obviously unavailable on AWS instances.
The instructions for deploying DetectionLab in AWS via this method are available here: [Build Your Own AMIs README](./VM_to_AMIs.md)

This method has the benefit of not requiring any local infrastructure for builds but requires a lot of work, cost, and time to convert the build process to be cloud-based.

### Progress Updates
### Method 2 - Pre-built AMIs
As of March 2019, I am now sharing pre-built AMIs on the Amazon Marketplace. The code inside of main.tf uses Terraform data sources to determine the correct AMI ID and will use the pre-built AMIs by default.

The instructions for deploying DetectionLab to AWS via Method 1 are available here: [Method 1 README](./Method1/Method1.md)
Using this method, it should be possible to bring DetectionLab online in under 15 minutes.

Progress on Method 2 will be tracked using a GitHub project that is viewable here: https://github.com/clong/DetectionLab/projects/1
The instructions for deploying DetectionLab in AWS using the pre-built AMIs are available here: [Pre-Built AMIs README](./Pre-Built_AMIs.md)
45 changes: 45 additions & 0 deletions Terraform/VM_to_AMIs.md
@@ -0,0 +1,45 @@
# Method 1 - Build Locally and Import to AWS

This method involves using Terraform to bring DetectionLab infrastructure online by first building it locally using Virtualbox/VMware and then [importing the resulting virtual machines](https://docs.aws.amazon.com/vm-import/latest/userguide/vmimport-image-import.html#import-vm-image) as AMIs on AWS.

The supplied Terraform configuration can then be used to create EC2 instances and all requisite networking components.

## Prerequisites
* A machine to build DetectionLab with
* An AWS account
* An AWS user and access keys to use with the AWS CLI
* Optional but recommended: a separate user for Terraform

## Step by step guide

1. Build the lab by following the [README](https://github.com/clong/DetectionLab/blob/master/README.md)
2. [Configure the AWS command line utility](https://docs.aws.amazon.com/polly/latest/dg/setup-aws-cli.html)
3. [Create an S3 bucket](https://docs.aws.amazon.com/AmazonS3/latest/user-guide/create-bucket.html). You will upload the DetectionLab VMs to this bucket later.

4. For the VM importation to work, you must create a role named `vmimport` with a trust relationship policy document that allows VM Import to assume the role, and you must attach an IAM policy to the role:

```aws iam create-role --role-name vmimport --assume-role-policy-document file:///path/to/DetectionLab/Terraform/vm_import/trust-policy.json```

5. Edit `/path/to/DetectionLab/Terraform/vm_import/role-policy.json` and insert the name of the bucket you created in step 3 on lines 12-13, replacing `YOUR_BUCKET_GOES_HERE` with the name of your bucket.

6. Use the create-role command to create a role named vmimport and give VM Import/Export access to it:

```aws iam put-role-policy --role-name vmimport --policy-name vmimport --policy-document file:///path/to/DetectionLab/Terraform/vm_import/role-policy.json```

7. Export the DetectionLab VMs as single file OVA files if they are not already in that format
8. [Upload the OVAs to the S3 bucket](https://docs.aws.amazon.com/AmazonS3/latest/user-guide/upload-objects.html) you created in step three
9. Edit the `logger.json`, `dc.json`, `wef.json` and `win10.json` files and modify the S3Bucket and S3Key headers to match the location of the OVA files in your S3 bucket.
10. Import the VMs from S3 as AMIs by running the following commands:
```
aws ec2 import-image --description "dc" --license-type byol --disk-containers file:///path/to/DetectionLab/Terraform/vm_import/dc.json
aws ec2 import-image --description "wef" --license-type byol --disk-containers file:///path/to/DetectionLab/Terraform/vm_import/wef.json
aws ec2 import-image --description "win10" --license-type byol --disk-containers file:///path/to/DetectionLab/Terraform/vm_import/win10.json
aws ec2 import-image --description "logger" --license-type byol --disk-containers file:///path/to/DetectionLab/Terraform/vm_import/logger.json
```
11. Check on the status of the importation with the following command:

```aws ec2 describe-import-image-tasks --import-task-ids <import-ami-xxxxxxxxxxxxxxxxx>```

12. Fill out the variables in `/path/to/DetectionLab/Terraform/terraform.tfvars`
13. Run `terraform init` to setup the initial Terraform configuration
14. `cd /path/to/DetectionLab/Terraform/Method1 && terraform apply`
115 changes: 74 additions & 41 deletions Terraform/Method1/main.tf → Terraform/main.tf
@@ -1,27 +1,3 @@
# Terraform configuration to be used with DetectionLab Method1
# Before using this, you must fill out the variables in terraform.tfvars
# Please follow the instructions in https://github.com/clong/DetectionLab/blob/master/Terraform/Method1/Method1.md

variable "region" {
default = "us-west-1"
}
variable "shared_credentials_file" {
type = "string"
}
variable "key_name" {
default = "id_terraform"
}
variable "public_key_path" {
type = "string"
}
variable "ip_whitelist" {
type = "list"
}
variable "logger_ami" {}
variable "dc_ami" {}
variable "wef_ami" {}
variable "win10_ami" {}

# Specify the provider and access details
provider "aws" {
shared_credentials_file = "${var.shared_credentials_file}"
Expand Down Expand Up @@ -129,6 +105,14 @@ resource "aws_security_group" "windows" {
cidr_blocks = "${var.ip_whitelist}"
}

# Windows ATA
ingress {
from_port = 443
to_port = 443
protocol = "tcp"
cidr_blocks = "${var.ip_whitelist}"
}

# Allow all traffic from the private subnet
ingress {
from_port = 0
Expand All @@ -147,75 +131,124 @@ resource "aws_security_group" "windows" {
}

resource "aws_key_pair" "auth" {
key_name = "${var.key_name}"
public_key = "${file(var.public_key_path)}"
key_name = "${var.public_key_name}"
public_key = "${file("${var.public_key_path}")}"
}

resource "aws_instance" "logger" {
instance_type = "t3.medium"
ami = "${var.logger_ami}"
instance_type = "t2.medium"
ami = "ami-0ad16744583f21877"
tags {
Name = "logger"
}
subnet_id = "${aws_subnet.default.id}"
vpc_security_group_ids = ["${aws_security_group.logger.id}"]
key_name = "${aws_key_pair.auth.id}"
key_name = "${aws_key_pair.auth.key_name}"
private_ip = "192.168.38.105"
# Run the following commands to restart Fleet
# Provision the AWS Ubuntu 16.04 AMI from scratch.
provisioner "remote-exec" {
inline = [
"cd /home/vagrant/kolide-quickstart && sudo docker-compose stop",
"sudo service docker restart",
"cd /home/vagrant/kolide-quickstart && sudo docker-compose start"
"sudo add-apt-repository universe && sudo apt-get update && sudo apt-get install -y git",
"echo 'logger' | sudo tee /etc/hostname && sudo hostnamectl set-hostname logger",
"sudo adduser --disabled-password --gecos \"\" vagrant && echo 'vagrant:vagrant' | sudo chpasswd",
"echo 'vagrant ALL=(ALL:ALL) NOPASSWD:ALL' | sudo tee -a /etc/sudoers",
"sudo git clone https://github.com/clong/DetectionLab.git /opt/DetectionLab",
"sudo sed -i \"s#sed -i 's/archive.ubuntu.com/us.archive.ubuntu.com/g' /etc/apt/sources.list##g\" /opt/DetectionLab/Vagrant/bootstrap.sh",
"sudo sed -i 's/eth1/eth0/g' /opt/DetectionLab/Vagrant/bootstrap.sh",
"sudo sed -i 's/ETH1/ETH0/g' /opt/DetectionLab/Vagrant/bootstrap.sh",
"sudo sed -i 's#/usr/local/go/bin/go get -u#GOPATH=/root/go /usr/local/go/bin/go get -u#g' /opt/DetectionLab/Vagrant/bootstrap.sh",
"sudo sed -i 's#/vagrant/resources#/opt/DetectionLab/Vagrant/resources#g' /opt/DetectionLab/Vagrant/bootstrap.sh",
"sudo chmod +x /opt/DetectionLab/Vagrant/bootstrap.sh",
"sudo apt-get update",
"sudo /opt/DetectionLab/Vagrant/bootstrap.sh",
"sudo pip3.6 install --upgrade --force-reinstall pip==9.0.3 && sudo pip3.6 install -r /home/vagrant/caldera/caldera/requirements.txt && sudo pip3.6 install --upgrade pip",
"sudo service caldera stop && sudo service caldera start"
]
connection {
type = "ssh"
user = "vagrant"
password = "vagrant"
user = "ubuntu"
private_key = "${file("${var.private_key_path}")}"
}
}
root_block_device {
delete_on_termination = true
volume_size = 64
}
}

resource "aws_instance" "dc" {
instance_type = "t2.small"
ami = "${var.dc_ami}"
instance_type = "t2.medium"
ami = "${data.aws_ami.dc_ami.image_id}"
tags {
Name = "dc.windomain.local"
}
subnet_id = "${aws_subnet.default.id}"
vpc_security_group_ids = ["${aws_security_group.windows.id}"]
private_ip = "192.168.38.102"
provisioner "remote-exec" {
connection = {
type = "winrm"
user = "vagrant"
password = "vagrant"
agent = "false"
insecure = "true"
}
inline = [
"powershell -command \"$newDNSServers = @('192.168.38.102','8.8.8.8'); $adapters = Get-WmiObject Win32_NetworkAdapterConfiguration | Where-Object {$_.IPAddress -match '192.168.38.'}; $adapters | ForEach-Object {$_.SetDNSServerSearchOrder($newDNSServers)}\"",
]
}
root_block_device {
delete_on_termination = true
}
}

resource "aws_instance" "wef" {
instance_type = "t2.small"
ami = "${var.wef_ami}"
instance_type = "t2.medium"
ami = "${data.aws_ami.wef_ami.image_id}"
tags {
Name = "wef.windomain.local"
}
subnet_id = "${aws_subnet.default.id}"
vpc_security_group_ids = ["${aws_security_group.windows.id}"]
private_ip = "192.168.38.103"
provisioner "remote-exec" {
connection = {
type = "winrm"
user = "vagrant"
password = "vagrant"
agent = "false"
insecure = "true"
}
inline = [
"powershell -command \"$newDNSServers = @('192.168.38.102','8.8.8.8'); $adapters = Get-WmiObject Win32_NetworkAdapterConfiguration | Where-Object {$_.IPAddress -match '192.168.38.'}; $adapters | ForEach-Object {$_.SetDNSServerSearchOrder($newDNSServers)}\"",
]
}
root_block_device {
delete_on_termination = true
}
}

resource "aws_instance" "win10" {
instance_type = "t2.small"
ami = "${var.win10_ami}"
instance_type = "t2.medium"
ami = "${data.aws_ami.win10_ami.image_id}"
tags {
Name = "win10.windomain.local"
}
subnet_id = "${aws_subnet.default.id}"
vpc_security_group_ids = ["${aws_security_group.windows.id}"]
private_ip = "192.168.38.104"
provisioner "remote-exec" {
connection = {
type = "winrm"
user = "vagrant"
password = "vagrant"
agent = "false"
insecure = "true"
}
inline = [
"powershell -command \"$newDNSServers = @('192.168.38.102','8.8.8.8'); $adapters = Get-WmiObject Win32_NetworkAdapterConfiguration | Where-Object {$_.IPAddress -match '192.168.38.'}; $adapters | ForEach-Object {$_.SetDNSServerSearchOrder($newDNSServers)}\"",
]
}
root_block_device {
delete_on_termination = true
}
Expand Down
15 changes: 15 additions & 0 deletions Terraform/outputs.tf
@@ -0,0 +1,15 @@
output "logger_public_ip" {
value = "${aws_instance.logger.public_ip}"
}

output "dc_public_ip" {
value = "${aws_instance.dc.public_ip}"
}

output "wef_public_ip" {
value = "${aws_instance.wef.public_ip}"
}

output "win10_public_ip" {
value = "${aws_instance.win10.public_ip}"
}
6 changes: 6 additions & 0 deletions Terraform/terraform.tfvars.example
@@ -0,0 +1,6 @@
region = "us-west-1"
shared_credentials_file = "/home/user/.aws/credentials"
public_key_name = "id_logger"
public_key_path = "/home/user/.ssh/id_logger.pub"
private_key_path = "/home/user/.ssh/id_logger"
ip_whitelist = ["1.2.3.4/32"]
9 changes: 9 additions & 0 deletions Terraform/vm_import/dc.json
@@ -0,0 +1,9 @@
[
{
"Description": "dc",
"Format": "ova",
"UserBucket": {
"S3Bucket": "YOUR_BUCKET_GOES_HERE",
"S3Key": "dc.ova"
}
}]
9 changes: 9 additions & 0 deletions Terraform/vm_import/logger.json
@@ -0,0 +1,9 @@
[
{
"Description": "logger",
"Format": "ova",
"UserBucket": {
"S3Bucket": "YOUR_BUCKET_GOES_HERE",
"S3Key": "logger.ova"
}
}]
27 changes: 27 additions & 0 deletions Terraform/vm_import/role-policy.json
@@ -0,0 +1,27 @@
{
"Version":"2012-10-17",
"Statement":[
{
"Effect":"Allow",
"Action":[
"s3:GetBucketLocation",
"s3:GetObject",
"s3:ListBucket"
],
"Resource":[
"arn:aws:s3:::YOUR_BUCKET_GOES_HERE",
"arn:aws:s3:::YOUR_BUCKET_GOES_HERE/*"
]
},
{
"Effect":"Allow",
"Action":[
"ec2:ModifySnapshotAttribute",
"ec2:CopySnapshot",
"ec2:RegisterImage",
"ec2:Describe*"
],
"Resource":"*"
}
]
}
15 changes: 15 additions & 0 deletions Terraform/vm_import/trust-policy.json
@@ -0,0 +1,15 @@
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Principal": { "Service": "vmie.amazonaws.com" },
"Action": "sts:AssumeRole",
"Condition": {
"StringEquals":{
"sts:Externalid": "vmimport"
}
}
}
]
}

0 comments on commit 26140b2

Please sign in to comment.