HashiCorp Developer AI Private beta now available Sign up today
Terraform
Terraform Home

You are viewing documentation for version v202308-1. View latest version.

Operational Mode Data Storage Requirements

Terraform Enterprise data storage requirements differ based on the operational mode you choose for your instance. If you are following one of the reference architectures , we recommend referring to it while preparing your data storage services.

External Services Mode

External Services mode has the following requirements:

  • PostgreSQL Requirements
  • Any S3-compatible object storage service, GCP Cloud Storage or Azure blob storage meets Terraform Enterprise's object storage requirements. You must create a bucket for Terraform Enterprise to use, and specify that bucket during installation. Depending on your infrastructure provider, you might need to ensure the bucket is in the same region as the Terraform Enterprise instance.
    • Disable any lifecycle rules that would delete, archive, or transition objects in this container. Terraform Enterprise expects to manage all data the object storage service, so any lifecycle moves may result in unexpected data inconsistencies.

Active/Active Mode

Active/Active mode has the same requirements as External Services, with the additions of an external Redis server, and a fully automated install method. Specifically, Active/Active operational mode has the following additional requirements:

  • Redis server version 6.x or 7.x (recommended)
  • Redis Cluster is not supported.

Mounted Disk

If you choose to use the Mounted Disk operational mode, Terraform Enterprise will manage its own PostgreSQL database and object storage using a separate directory on the host, with the intention that the directory is configured to store its data on an external disk, such as EBS, iSCSI, etc.

We strongly suggest following the guidelines below for mounted disk storage.

Supported Mounted Disk Types

The following are supported mounted disk types:

  • AWS EBS
  • GCP Zonal Persistent Disk
  • Azure Disk Storage
  • iSCSI
  • SAN
  • Physically connected disks as in non-cloud hardware

These disk types provide the necessary reliability and performance for data storage and retrieval in Terraform Enterprise.

Unsupported Mounted Disk Types

The following are generally not supported mounted disk types:

  • NFS
  • SMB/CIFS

Terraform Enterprise's storage device or service must be highly reliable and high-speed in both I/O and connectivity to meet performance requirements. Device types in the supported list will usually meet these requirements, but many standard NAS and other device types will not perform at the level required. Only use a NAS or other device type not in the supported list if you are certain it can accommodate these requirements. For more information about high-speed and highly available storage please see your storage vendor.

Mounted Disk Types Not Listed Here

If the type of mounted disk you wish to use is not in either of the above lists, please contact your HashiCorp representative for clarification on whether that type is supported.

Minimum Disk Size

Terraform Enterprise's minimum disk size is 40GB.

Depending on your cloud or storage application, you may need to confirm the disk has been resized to at least 40GB.

For example, with RedHat-flavor (RHEL, CentOS, Oracle Linux) images in Azure Cloud, the storage disk must be resized above the 30GB default after initial boot with fdisk, as documented in the Azure knowledge base article How to: Resize Linux osDisk partition on Azure.

Database Maintenance

There are three CLI commands available as of v202005-2 to facilitate management of the PostgreSQL database that runs on the host as part of the Mounted Disk operational mode:

  • replicated admin db-backup: This will run a pg_dump and store the backup in /backup/ptfe.db on the host.
  • replicated admin db-restore: This will run a pg_restore using /backup/ptfe.db as it's data source.
  • replicated admin db-reindex: This will run a REINDEX against the application database. Note: A reindex can take anywhere from minutes to hours to complete, depending on the size of your database. Running this command locks the database and prevents any other action against it.

These commands will only display output if there is an error. Please contact support if you have any questions or issues with these commands.