This is intended as a repository of sample terraform configs for Apstra. Each directory in this project tracks a different project that uses the terraform provider for Apstra.
A set of terraform configs that design network fabrics for AI GPU training, storage and frontend management with Apstra deploying logical devices, racks and templates. Design examples include 64 servers (512 GPUs), 128 servers, 256 servers and beyond with NVIDIA-prescribed rail-optimized topologies.
A terraform config for a Juniper Validated Design (JVD) of an AI cluster's 3 fabrics: GPU backend, storage, and frontend management. This example shows an exact small cluster design of two NVIDIA rail-optimzed groups with Juniper QFX leaf switches and high-radix PTX spines, with options for A100s and H100-based servers.
Terraform configs to pull device configs and store them in a git repo
A set of terraform configs that track the terraform lab here These were used in this podcast by @chrismarget-j and @rajagopalans
Code with a simple terraform config that shows how to integrate github, terraform cloud and apstra seamlessly. Please follow the steps in the README in the directory.
This directory contains examples from various Terraform/Apstra workshops to a variety of audiences.
- dc_inform_2023q3 - A set of terraform configs and a slide deck about gitops workflows presented by @chrismarget-j in August of 2023
- partner_demo_23q4 - A set of terraform configs demonstrating endpoint connectivity and scaling presented by @chrismarget-j and Samir Parikh in November of 2023
- The user needs to have access to an Apstra environment. This could be your own Apstra instance or a cloudlabs If you would like access to Apstra, please contact your Juniper account team.
- Terraform is already installed in the environment. If it is not, please follow the steps here
The following steps are common to all the projects in this subdirectory, except github_integration
Each of the sub-directories is a set of configs. The first step is always to enter the directories
terraform-apstra-examples % ls
README.md ai-clusters dc_inform_2023q3 terraform-lab
terraform-apstra-examples % cd ai-clusters
ai-clusters %
In the "provider" section of the provider.tf file, update the URL like so. # URL and credentials can be supplied using the "url" parameter in this file. url = "https://user:password@address_of_apstra" For instance, if the username is "admin" and the password is "goodguess" and the APSTRA url is https://3.4.5.6:34000 , the url field would look like so "https://admin:goodguess@3.4.5.6:34000" Alternatively, the APSTRA_URL environment variable can be set up similarly. If Username or Password are not embedded in the URL, the provider will look for them in the APSTRA_USER and APSTRA_PASS environment variables. sample provider.tf
terraform {
required_providers {
apstra = {
source = "Juniper/apstra"
}
}
}
provider "apstra" {
url = "https://admin:Hello3%23@3.5.8.7:30000/"
tls_validation_disabled = true
blueprint_mutex_enabled = false
}
If the username or password contains special characters they would need to be URL encoded. In the example above, the password is 'Hello3#"
From the same directory as the configs, type terraform init. This process downloads the provider from the registry, downloads any required modules (we’re not using any), and sets up the backend (we haven’t configured one, so it defaults to this folder). Type "terraform init" and press enter
Terraform will download the apstra provider and inform you that it is initialized.
% terraform init
Initializing the backend...
Initializing provider plugins...
- Finding latest version of juniper/apstra...
- Installing juniper/apstra v0.27.2...
- Installed juniper/apstra v0.27.2 (self-signed, key ID CB9C922903A66F3F)
Partner and community providers are signed by their developers.
If you'd like to know more about provider signing, you can read about it here:
https://www.terraform.io/docs/cli/plugins/signing.html
Terraform has created a lock file .terraform.lock.hcl to record the provider
selections it made above. Include this file in your version control repository
so that Terraform can guarantee to make the same selections by default when
you run "terraform init" in the future.
Terraform has been successfully initialized!
You may now begin working with Terraform. Try running "terraform plan" to see
any changes that are required for your infrastructure. All Terraform commands
should now work.
If you ever set or change modules or backend configuration for Terraform,
rerun this command to reinitialize your working directory. If you forget, other
commands will detect it and remind you to do so if necessary.
As a next step, type "terraform plan" from the same directory as the configs.
Terraform will reach out to the Apstra instance, compute the difference between the existing running state and the updated configs, and output the result.
% terraform plan
Terraform used the selected providers to generate the following execution plan. Resource actions are indicated with the following symbols:
+ create
Terraform will perform the following actions:
# apstra_logical_device.AI-Leaf_16_16x400 will be created
+ resource "apstra_logical_device" "AI-Leaf_16_16x400" {
...
... Trimmed for brevity
...
Plan: 21 to add, 0 to change, 0 to destroy.
type "terraform apply" This will deploy the logical devices, rack types and the templates into Apstra.
% terraform apply
Terraform used the selected providers to generate the following execution plan. Resource actions are indicated with the following symbols:
+ create
Terraform will perform the following actions:
# apstra_logical_device.AI-Leaf_16_16x400 will be created
...
... Trimmed for brevity
...
apstra_template_rack_based.AI_Cluster_64_DGX-A100: Creation complete after 3s [id=045520e5-c2b3-4fae-8456-64e66072c9d9]
apstra_template_rack_based.AI_Cluster_128_DGX-A100: Creation complete after 3s [id=d8f200b9-633b-427c-8b11-5ff5a9b7f33c]
Apply complete! Resources: 21 added, 0 changed, 0 destroyed.
At this point, the resources will be deployed into Apstra. You can login to Apstra and confirm to your satisfaction.