Weights & Biases Azure Module

This is a Terraform module for provisioning a Weights & Biases Cluster on Azure. Weights & Biases Server is our self-hosted distribution of wandb.ai. It offers enterprises a private instance of the Weights & Biases application, with no resource limits and with additional enterprise-grade architectural features like audit logging and single sign-on.

About This Module

Pre-requisites

This module is intended to run in an Azure account with minimal preparation, however it does have the following pre-requisites:

Terraform version >= 1

Credentials / Permissions

How to Use This Module

Cluster Sizing

By default, the type of kubernetes instances, number of instances, redis cluster size, and database instance sizes are standardized via configurations in ./deployment-size.tf, and is configured via the size input variable.

Available sizes are, small, medium, large, xlarge, and xxlarge. Default is small.

All the values set via deployment-size.tf can be overridden by setting the appropriate input variables.

• kubernetes_instance_type - The instance type for the EKS nodes
• kubernetes_min_node_per_az - The minimum number of nodes in the EKS cluster
• kubernetes_max_node_per_az - The maximum number of nodes in the EKS cluster
• redis_capacity - The instance type for the redis cluster
• database_sku_name - The instance type for the database

Examples

We have included documentation and reference examples for additional common installation scenarios for Weights & Biases, as well as examples for supporting resources that lack official modules.

• Route

Requirements

Name	Version
terraform	~> 1.0
azapi	~> 1.0
azurerm	~> 3.17
helm	~> 2.6
kubernetes	~> 2.23

Providers

Name	Version
azapi	~> 1.0
azurerm	~> 3.17

Modules

Name	Source	Version
app_aks	./modules/app_aks	n/a
app_lb	./modules/app_lb	n/a
cert_manager	./modules/cert_manager	n/a
clickhouse	./modules/clickhouse	n/a
cron_job	./modules/cron_job	n/a
database	./modules/database	n/a
identity	./modules/identity	n/a
networking	./modules/networking	n/a
pod_identity	./modules/identity	n/a
redis	./modules/redis	n/a
storage	./modules/storage	n/a
vault	./modules/vault	n/a
wandb	wandb/wandb/helm	1.2.0

Resources

Name	Type
azapi_resource_list.az_zones	data source
azurerm_subscription.current	data source

Inputs

Name	Description	Type	Default	Required
allowed_ip_ranges	Allowed public IP addresses or CIDR ranges.	list(string)	[]	no
allowed_subscriptions	List of allowed customer subscriptions coma seperated values	string	""	no
app_wandb_env	Extra environment variables for W&B	map(string)	{}	no
azuremonitor	# To support otel azure monitor sql and redis metrics need operator-wandb chart minimum version 0.14.0	bool	false	no
blob_container	Use an existing bucket.	string	""	no
bucket_path	path of where to store data for the instance-level bucket	string	""	no
clickhouse_private_endpoint_service_name	ClickHouse private endpoint 'Service name' (ends in .azure.privatelinkservice).	string	""	no
clickhouse_region	ClickHouse region (eastus2, westus3, etc).	string	""	no
cluster_sku_tier	The Azure AKS SKU Tier to use for this cluster (https://learn.microsoft.com/en-us/azure/aks/free-standard-pricing-tiers)	string	"Free"	no
controller_image_tag	Tag of the controller image to deploy	string	"1.14.0"	no
create_private_link	Use for the azure private link.	bool	false	no
create_redis	Boolean indicating whether to provision an redis instance (true) or not (false).	bool	false	no
database_availability_mode	n/a	string	"SameZone"	no
database_sku_name	Specifies the SKU Name for this MySQL Server. Defaults to null and value from deployment-size.tf is used	string	null	no
database_version	Version for MySQL	string	"5.7"	no
deletion_protection	If the instance should have deletion protection enabled. The database / Bucket can't be deleted when this value is set to true.	bool	true	no
disable_storage_vault_key_id	Flag to disable the customer_managed_key block, the properties 'encryption.identity, encryption.keyvaultproperties' cannot be updated in a single operation.	bool	false	no
domain_name	Domain for accessing the Weights & Biases UI.	string	null	no
enable_database_vault_key	Flag to enable managed key encryption for the database. Once enabled, cannot be disabled.	bool	false	no
enable_storage_vault_key	Flag to enable managed key encryption for the storage account.	bool	false	no
external_bucket	config an external bucket	any	null	no
kubernetes_cluster_oidc_issuer_url	OIDC issuer URL for the Kubernetes cluster. Can be determined using kubectl get --raw /.well-known/openid-configuration	string	""	no
kubernetes_instance_type	Instance type for primary node group. Defaults to null and value from deployment-size.tf is used	string	null	no
kubernetes_max_node_per_az	Maximum number of nodes for the AKS cluster. Defaults to null and value from deployment-size.tf is used	number	null	no
kubernetes_min_node_per_az	Minimum number of nodes for the AKS cluster. Defaults to null and value from deployment-size.tf is used	number	null	no
license	Your wandb/local license	string	n/a	yes
location	n/a	string	n/a	yes
namespace	String used for prefix resources.	string	n/a	yes
node_max_pods	Maximum number of pods per node	number	30	no
node_pool_num_zones	Number of availability zones to use for the node pool when node_pool_zones is not set. If neither are set, 3 zones will be used	number	2	no
node_pool_zones	Availability zones for the node pool	list(string)	null	no
oidc_auth_method	OIDC auth method	string	"implicit"	no
oidc_client_id	The Client ID of application in your identity provider	string	""	no
oidc_issuer	A url to your Open ID Connect identity provider, i.e. https://cognito-idp.us-east-1.amazonaws.com/us-east-1_uiIFNdacd	string	""	no
oidc_secret	The Client secret of application in your identity provider	string	""	no
operator_chart_version	Version of the operator chart to deploy	string	"1.3.4"	no
other_wandb_env	Extra environment variables for W&B	map(any)	{}	no
parquet_wandb_env	Extra environment variables for W&B	map(string)	{}	no
redis_capacity	Number indicating size of an redis instance. Defaults to null and value from deployment-size.tf is used	number	null	no
size	Deployment size	string	"small"	no
ssl	Enable SSL certificate	bool	true	no
storage_account	Azure storage account name	string	""	no
storage_key	Azure primary storage access key	string	""	no
subdomain	Subdomain for accessing the Weights & Biases UI. Default creates record at Route53 Route.	string	null	no
tags	Map of tags for resource	map(string)	{}	no
use_internal_queue	Uses an internal redis queue instead of using azure queue.	bool	false	no
wandb_image	Docker repository of to pull the wandb image from.	string	"wandb/local"	no
wandb_version	The version of Weights & Biases local to deploy.	string	"latest"	no
weave_wandb_env	Extra environment variables for W&B	map(string)	{}	no

Outputs

Name	Description
address	n/a
aks_max_node_count	n/a
aks_min_node_count	n/a
aks_node_instance_type	n/a
client_id	n/a
cluster_ca_certificate	n/a
cluster_client_certificate	n/a
cluster_client_key	n/a
cluster_host	n/a
database_instance_type	n/a
fqdn	The FQDN to the W&B application
oidc_issuer_url	n/a
private_link_resource_id	n/a
private_link_sub_resource_name	n/a
standardized_size	n/a
tenant_id	n/a
url	The URL to the W&B application

Upgrading from 3.x to 4.x

3.0.0 introduced autoscaling to the AKS cluster and made the size variable the preferred way to set the cluster size. Previously, unless the size variable was set explicitly, there were default values for the following variables:

• kubernetes_instance_type
• kubernetes_node_count
• redis_capacity
• database_sku_name

The size variable is now defaulted to small, and the following values to can be used to partially override the values set by the size variable:

• kubernetes_instance_type
• kubernetes_min_node_per_az
• kubernetes_max_node_per_az
• redis_capacity
• database_sku_name

For more information on the available sizes, see the Cluster Sizing section.

If having the cluster scale nodes in and out is not desired, the kubernetes_min_node_per_az and kubernetes_max_node_per_az can be set to the same value to prevent the cluster from scaling.

Upgrading from 2.x to 3.x

When upgrading from 2.x to 3.x, the following changes are required:

1. Add the azapi provider to the required_providers block:

terraform {
  required_providers {
    azapi = {
      source  = "azure/azapi"
      version = "~> 1.0"
    }
  }
}

2. Add the azapi provider to the provider block:

provider "azapi" {
    # azapi provider configuration should be the same as azurerm provider configuration
}