Terraform Enterprise 2.0.6 is available

The Terraform Enterprise team would like to announce the release 2.0.6 is now available.

Some key highlights and important bits from the release notes:
Features

  1. Redis mTLS connections now support password authentication. You can now enable both certificate-based mutual TLS and Redis ACL credentials at the same time.

Improvements

  1. Task worker logs now include project_name and a destroy run indicator in their metadata. Operators who consume task worker log output can now identify which project a run belongs to and whether the run is a terraform destroy operation.
  2. Two new configuration options, TFE_OBJECT_STORAGE_AZURE_UPLOAD_PART_SIZE and TFE_OBJECT_STORAGE_AZURE_UPLOAD_CONCURRENCY, are now available for Azure Blob Storage. These options bring Azure upload configuration to parity with the existing AWS S3 options, allowing operators to tune upload performance for large artifacts.

Bug Fixes

  1. Terraform Enterprise now reaps zombie processes left behind by VCS operations. When Terraform Enterprise runs as PID 1 and performs Git-based operations such as VCS-triggered runs, child Git processes could accumulate as zombie processes and never be cleaned up. This issue has been resolved.
  2. Background migrations now correctly handle workspaces that have non-standard Terraform version strings. Previously, a database migration used to populate Explorer data could fail when encountering version strings that do not follow standard versioning, causing Explorer data to stop being updated. This issue has been resolved.
  3. Terraform Stacks now respects the no_proxy environment variable configuration. Previously, some Stacks runs were ignoring no_proxy configuration, causing all traffic to be routed through the configured proxy. This issue has been resolved.

And we’ve got the usual security improvements / patches.