Terraform Enterprise configuration reference

This topic contains reference information about the configurations you can specify for Terraform Enterprise.

Introduction

Specify environment variables to configure the operational mode, connections to external systems, and other settings. Specify the variables in a configuration file for your runtime environment, such as a Docker Compose file, Kubernetes Helm chart, or in the env block of your Nomad job. Refer to Create deployment configuration overview for additional information.

Application settings

TFE_CAPACITY_CONCURRENCY

Maximum number of Terraform runs that can execute concurrently on each Terraform Enterprise node. Defaults to 10.

TFE_CAPACITY_CPU

Maximum number of CPU cores a Terraform run is allowed to use. Set to 0 for no limit. Defaults to 0. Ignored if TFE_RUN_PIPELINE_DRIVER is set to nomad.

TFE_CAPACITY_MEMORY

Maximum amount of memory (MiB) a Terraform run is allowed to use. Defaults to 2048. Ignored if TFE_RUN_PIPELINE_DRIVER is set to nomad.

TFE_ENCRYPTION_PASSWORD

Encryption password used to encrypt and decrypt the internal Vault root token and unseal key.

Required when TFE_VAULT_USE_EXTERNAL is false.

TFE_HOSTNAME

Hostname where Terraform Enterprise is accessed (e.g., terraform.example.com).

Required always.

TFE_LICENSE

The raw HashiCorp license (e.g., 02MV4UU4...MSEKE6T2). Defaults to "".

Required when TFE_LICENSE_PATH is unset.

TFE_LICENSE_PATH

Path to a file containing the raw HashiCorp license. Must be mounted inside the container (e.g., /opt/terraform-enterprise-license). Defaults to "".

Required when TFE_LICENSE is unset.

TFE_LICENSE_REPORTING_OPT_OUT

Whether to opt out of reporting licensing information to HashiCorp. Defaults to false.

TFE_USAGE_REPORTING_OPT_OUT

Whether to opt out of reporting usage information to HashiCorp. Defaults to false.

TFE_RUN_PIPELINE_DRIVER

Driver for Terraform runs. Must be one of nomad, docker or kubernetes. Defaults to docker. If set to nomad, the TFE_CAPACITY_CPU and TFE_CAPACITY_MEMORY settings are ignored.

Required for all deployment options.

TFE_OPERATIONAL_MODE

Determines how Terraform Enterprise stores and retrieves data. This variable is required for all runtimes except Kubernetes.

You can set one of the following values:

• disk: Runs Terraform Enterprise, PostgreSQL database, optional Redis data store, and S3-compatible storage in the same container. In this mode, Terraform Enterprise manages the storage objects. By default, this mode limits the number of Terraform Enterprise instances to one.
• external: Runs Terraform and optional Redis store in the same container. In this mode, Terraform Enterprise can manage the Redis store, but you are responsible for managing the PostgreSQL database and S3-compatible storage as external services. By default, this mode limits the number of Terraform Enterprise instances to one.
• active-active: Runs Terraform Enterprise in its own container. In this mode, you are responsible for managing the PostgreSQL, optional Redis data store, and S3-compatible storage as external services. As a result, you can deploy more than one instance of Terraform Enterprise. Terraform Enterprise operates in active-active mode when it is deployed to Kubernetes.

Refer to Configure the operational mode for additional information.

TFE_RUN_PIPELINE_IMAGE

Container image used to execute Terraform runs. Leave blank to use the default image that comes with Terraform Enterprise. Defaults to "".

TFE_DISK_PATH

Path where Terraform Enterprise stores application data. Must be backed by persistent and performant storage such as a container volume or block storage. Defaults to /var/lib/terraform-enterprise.

Required when TFE_OPERATIONAL_MODE is disk on Docker runtime.

TFE_BACKUP_RESTORE_TOKEN

The authentication token for the backup/restore API. Leave blank to have it be automatically generated. Defaults to "".

TFE_NODE_ID

A unique identifier for this node. Useful to identify nodes in an active/active installation. Leave blank to have it be automatically generated. Defaults to "".

TFE_HTTP_PORT

Port application listens on for HTTP. Default 80.

TFE_HTTPS_PORT

Port application listens to HTTPS on. Default 443.

Database settings

TFE_DATABASE_HOST

The PostgreSQL server to connect to in the format HOST[:PORT] (e.g., db.example.com or db.example.com:5432). If only HOST is provided then the :PORT defaults to :5432.

Required when PostgreSQL server is externalized.

TFE_DATABASE_NAME

Name of the PostgreSQL database to store application data in.

Required when PostgreSQL server is externalized.

TFE_DATABASE_PARAMETERS

PostgreSQL server parameters for the connection URI. Used to configure the PostgreSQL connection (e.g., sslmode=require).

TFE_DATABASE_PASSWORD

PostgreSQL password.

Required when PostgreSQL server is externalized.

TFE_DATABASE_USER

PostgreSQL user.

Required when PostgreSQL server is externalized.

TFE_DATABASE_RECONNECT_ENABLED

Whether to enable database reconnection in case of a failover. Defaults to 'false'. We recommend setting this configuration to true when connecting to an external PostgreSQL cluster.

This setting is not applicable for Replicated.

Redis settings

TFE_REDIS_HOST

The Redis server to connect to in the format HOST[:PORT] (e.g., redis.example.com or redis.example.com:). If only HOST is provided then the :PORT defaults to :6379.

Required when Redis server is externalized.

TFE_REDIS_USE_AUTH

Must be true or false. true indicates Redis server is configured to use TFE_REDIS_PASSWORD and TFE_REDIS_USER (optional) for authentication. Defaults to false.

TFE_REDIS_USE_TLS

Must be true or false. true indicates to use TLS to access Redis. Defaults to false.

TFE_REDIS_USER

Redis server user. This value is required when TFE_REDIS_USE_AUTH is set to true, and Redis server is configured to use username for authentication. There is no need to set this config when running Redis server internally as credential is not required.

TFE_REDIS_PASSWORD

Redis server password. This value is required when TFE_REDIS_USE_AUTH is set to true. There is no need to set this config when running Redis server internally as credential is not required.

Network settings

TFE_IACT_SUBNETS

Comma-separated list of subnets in CIDR notation that are allowed to retrieve the initial admin creation token via the API (e.g., 10.0.0.0/8,192.168.0.0/24). Leave blank to disable retrieving the initial admin creation token via the API from outside the host. Defaults to "".

If you do not set this value, you must gain access to the container or pod command line and run curl http://localhost:80/admin/retrieve-iact to retrieve the initial admin token.

TFE_IACT_TIME_LIMIT

Number of minutes that the initial admin creation token can be retrieved via the API after the application starts. Defaults to 60.

TFE_IACT_TRUSTED_PROXIES

Comma-separated list of proxy IP addresses that are allowed to retrieve the initial admin creation token via the API. Leave blank to disable retrieving the initial admin creation token through a proxy. Defaults to "".

http_proxy

Configures the proxy address to use for HTTP requests, for examplehttp://proxy.example.com:8080. Do not set a value blank to disable using a proxy server. Defaults to "".

https_proxy

Configures the proxy address to use for HTTPS requests, for example http://proxy.example.com:8080. Leave blank to disable using a proxy server. Defaults to "".

no_proxy

Specifies a list of domains that instances, such as S3, are allowed to connect directly to without going through the proxy, for example localhost,127.0.0.1. Defaults to "".

Observability settings

TFE_LOG_FORWARDING_CONFIG_PATH

The path to a file containing valid Fluent Bit [OUTPUT] configuration. The contents of the file you specify are appended to your existing Fluent Bit configuration if you set TFE_LOG_FORWARDING_ENABLED to true. Refer to Enable log forwarding for more information.

This setting is not applicable for Kubernetes.

TFE_LOG_FORWARDING_ENABLED

Whether to enable log forwarding. Defaults to false.

This setting is not applicable for Kubernetes.

TFE_METRICS_ENABLE

Whether to enable metrics collection. Defaults to false.

TFE_METRICS_HTTP_PORT

The HTTP port that metrics will be exposed on. Defaults to 9090.

TFE_METRICS_HTTPS_PORT

The HTTPS port that metrics will be exposed on. Defaults to 9091.

Object storage settings

TFE_OBJECT_STORAGE_TYPE

Type of object storage to use. Must be one of s3, azure, or google.

Required when object storage is externalized.

Azure blob Storage settings

TFE_OBJECT_STORAGE_AZURE_ACCOUNT_KEY

Azure Blob Storage access key.

Required when TFE_OBJECT_STORAGE_TYPE is azure and TFE_OBJECT_STORAGE_AZURE_USE_MSI is false.

TFE_OBJECT_STORAGE_AZURE_ACCOUNT_NAME

Azure Blob Storage account name.

Required when TFE_OBJECT_STORAGE_TYPE is azure.

TFE_OBJECT_STORAGE_AZURE_CLIENT_ID

Client ID of a user-assigned Managed Service Identity. Leave blank to use the system-assigned Managed Service Identity. Defaults to "".

TFE_OBJECT_STORAGE_AZURE_CONTAINER

Azure Blob Storage container name.

Required when TFE_OBJECT_STORAGE_TYPE is azure.

TFE_OBJECT_STORAGE_AZURE_ENDPOINT

Azure Storage endpoint. Useful if using a private endpoint for Azure Storage. Leave blank to use the default Azure Storage endpoint. Defaults to "".

TFE_OBJECT_STORAGE_AZURE_USE_MSI

Whether to use Managed Service Identity (MSI) for authentication. Defaults to false.

Google Cloud Platform storage

TFE_OBJECT_STORAGE_GOOGLE_BUCKET

Google Cloud Storage bucket name.

Required when TFE_OBJECT_STORAGE_TYPE is google.

TFE_OBJECT_STORAGE_GOOGLE_CREDENTIALS

Google Cloud Storage JSON credentials. Must be given as an escaped string of JSON or Base64 encoded JSON. Leave blank to use the attached service account. Defaults to "".

TFE_OBJECT_STORAGE_GOOGLE_PROJECT

Google Cloud Storage project name.

Required when TFE_OBJECT_STORAGE_TYPE is google.

S3-compatible Storage

TFE_OBJECT_STORAGE_S3_ACCESS_KEY_ID

S3 access key ID.

Required when TFE_OBJECT_STORAGE_TYPE is s3 and TFE_OBJECT_STORAGE_S3_USE_INSTANCE_PROFILE is false.

TFE_OBJECT_STORAGE_S3_BUCKET

S3 bucket name.

Required when TFE_OBJECT_STORAGE_TYPE is s3.

TFE_OBJECT_STORAGE_S3_ENDPOINT

S3 endpoint. Useful when using a private S3 endpoint. Leave blank to use the default AWS S3 endpoint. Defaults to "".

TFE_OBJECT_STORAGE_S3_REGION

S3 region.

Required when TFE_OBJECT_STORAGE_TYPE is s3.

TFE_OBJECT_STORAGE_S3_SECRET_ACCESS_KEY

S3 secret access key.

Required when TFE_OBJECT_STORAGE_TYPE is s3 and TFE_OBJECT_STORAGE_S3_USE_INSTANCE_PROFILE is false.

TFE_OBJECT_STORAGE_S3_SERVER_SIDE_ENCRYPTION

Server-side encryption algorithm to use. Leave blank to have the AWS SDK use a default value. Set to AES256 to use AES-256 encryption, the AWS S3 default since January 5, 2023. Set to aws:kms to use an AWS KMS key to encrypt data in which case you must specify TFE_OBJECT_STORAGE_S3_SERVER_SIDE_ENCRYPTION_KMS_KEY_ID. Defaults to "".

TFE_OBJECT_STORAGE_S3_SERVER_SIDE_ENCRYPTION_KMS_KEY_ID

KMS key ID to use for server-side encryption. Leave blank to use AWS-managed keys. Defaults to "".

TFE_OBJECT_STORAGE_S3_USE_INSTANCE_PROFILE

Whether to use the instance profile for authentication. Defaults to false.

TLS settings

TFE_TLS_CA_BUNDLE_FILE

Path to a file containing TLS CA certificates to be added to the OS CA certificates bundle. Leave blank to not add CA certificates to the OS CA certificates bundle. Defaults to "".

TFE_TLS_CERT_FILE

Path to a file containing the TLS certificate Terraform Enterprise will use when serving TLS connections to clients.

Required always.

TFE_TLS_CIPHERS

TLS ciphers to use for TLS. Must be valid OpenSSL format. Leave blank to use the default ciphers. Defaults to "".

TFE_TLS_ENFORCE

Whether or not to enforce TLS, Strict-Transport-Security headers, and secure cookies. Defaults to false.

TFE_TLS_KEY_FILE

Path to a file containing the TLS private key Terraform Enterprise will use when serving TLS connections to clients.

TFE_TLS_VERSION

TLS version to use. Must be one of tls_1_2, tls_1_3, or blank. Leave blank to use both TLS v1.2 and TLS v1.3. Defaults to "".

Vault settings

TFE_VAULT_ADDRESS

Address of the external Vault server (e.g., https://vault.example.com:8200). Defaults to "".

Required when TFE_VAULT_USE_EXTERNAL is true.

TFE_VAULT_CLUSTER_ADDRESS

Cluster URL of the internal Vault server on this node (e.g., http://192.168.0.1:8201). Must be reachable across nodes. Defaults to http://{{ GetPrivateIP }}:8201.

Required when TFE_OPERATIONAL_MODE is active-active.

TFE_VAULT_DISABLE_MLOCK

Disable mlock for internal Vault. Defaults to false.

TFE_VAULT_NAMESPACE

Vault namespace. External Vault only. Leave blank to use the default namespace. Defaults to "".

TFE_VAULT_PATH

Vault path when AppRole is mounted. External Vault only. Defaults to auth/approle.

TFE_VAULT_ROLE_ID

Vault role ID. External Vault only.

Required when TFE_VAULT_USE_EXTERNAL is true.

TFE_VAULT_SECRET_ID

Vault secret ID. External Vault only.

Required when TFE_VAULT_USE_EXTERNAL is true.

TFE_VAULT_TOKEN_RENEW

How often, in seconds, to renew the Vault token. External Vault only. Defaults to 3600.

TFE_VAULT_USE_EXTERNAL

Whether to use external Vault. Defaults to false.

Docker driver settings

TFE_RUN_PIPELINE_DOCKER_EXTRA_HOSTS

Comma-separated list of extra hosts in the format HOST:IP,HOST:IP to set in the /etc/hosts file within the container used to execute Terraform runs. Leave blank to not set any extra hosts. Defaults to "".

TFE_RUN_PIPELINE_DOCKER_NETWORK

Network where the container used to execute Terraform runs will be created. The network must already exist, it will not be created automatically. Leave blank to use the default network. Defaults to "".

TFE_DISK_CACHE_PATH

Path where Terraform Enterprise caches Terraform binaries. The volume specified in TFE_DISK_CACHE_VOLUME_NAME must be mounted to this path. Defaults to /var/cache/tfe-task-worker.

Required when TFE_RUN_PIPELINE_DRIVER is docker.

TFE_DISK_CACHE_VOLUME_NAME

Container volume name backing the TFE_DISK_CACHE_PATH.

Required when TFE_RUN_PIPELINE_DRIVER is docker.

Kubernetes driver settings

TFE_RUN_PIPELINE_KUBERNETES_IMAGE_PULL_SECRET_NAME

The name of an ImagePullSecret in the [namespace]-agents namespace to use when pulling the custom source tfc-agent image. If an ImagePullSecret is required to access a private repository you must create the secret within the [namespace]-agents namespace after this helm chart has installed, but before attempting a plan or apply. See Prerequisites for instructions for creating ImagePullSecrets.

TFE_RUN_PIPELINE_KUBERNETES_KUBECONFIG_PATH

The path to a Kubernetes configuration file.

TFE_RUN_PIPELINE_KUBERNETES_NAMESPACE

The Kubernetes namespace to create ephemeral run containers. Defaults to default.

TFE_RUN_PIPELINE_KUBERNETES_OPEN_SHIFT_ENABLED

A feature toggle to enable or disable compatibility with an OpenShift runtime environment. Defaults to false.

TFE_RUN_PIPELINE_KUBERNETES_DEBUG_ENABLED

Boolean flag that will delay the deletion of jobs from the cluster for their inspection.

TFE_RUN_PIPELINE_KUBERNETES_DEBUG_JOBS_TTL

Time in seconds after which the jobs delayed by the TFE_RUN_PIPELINE_KUBERNETES_DEBUG_ENABLED flag will get deleted. Defaults to 86400, 1 day.

TFE_RUN_PIPELINE_KUBERNETES_POD_TEMPLATE

This is a base64-encoded custom pod template in JSON format, equivalent to Kubernetes' corev1.PodTemplateSpec, used in a job. The values here will override corresponding values from Terraform Enterprise configuration. Note that only one container is allowed in the pod template.

Nomad driver settings

TFE_RUN_PIPELINE_NOMAD_NAMESPACE

The Nomad namespace where Terraform Enterprise agent job starts. Defaults to tfe-agents.

TFE_RUN_PIPELINE_NOMAD_AGENT_JOB_ID

Specifies the ID of the batch job that runs the Terraform Enterprise agent. Defaults to tfe-agent-job.

TFE_RUN_PIPELINE_NOMAD_WORKER_TIMEOUT

Max timeout in seconds used by Terraform Enterprise workers to connect with Nomad and start an agent job. Defaults to 60.

Proxy configuration

Terraform Enterprise relies on the native application runtime for setting proxy values. For configuring proxies in Docker, refer to the official Docker proxy configuration guide. For Kubernetes, refer to the Kubernetes proxy configuration guide. For Nomad deployments, refer to the reference architecture guide

Replicated to flexible deployments configuration mapping

The following table describes how the Replicated configuration maps to the configuration settings for non-Replicated deployments. Refer to settings for more information about the variables.