TFE Release v202204-1 (609)
Known Issues
- [April 21, 2022] This release includes an issue with the
tfe-bootstrap
container being unable to negotiate the Docker API version causing the installation and upgrade of Terraform Enterprise to fail withError response from daemon: client version 1.41 is too new. Maximum supported API version is 1.40
affecting dockerv19.03.x
or older. This issue is fixed inv202204-2
. - [May 3, 2022]
tfe-admin health-check
fails withsh: /root/ptfe-health-check: not found
after a change to theptfe_health_check
container Entrypoint to run the process as an unprivileged user under/run/ptfe-health-check
but thealias
was still referencing/root/ptfe-health-check
. This issue will be fixed inv202205-1
. - [May 10, 2022] This release includes an issue with Audit Log tagging, causing audit logs to no longer be tagged with the [Audit Log] prefix. This issue will be fixed in 'v202205-1'.
UPCOMING APPLICATION LEVEL BREAKING CHANGE
The Terraform Enterprise May 2022 release will change the names of application containers from ptfe_*
to tfe-*
for product consistency. For example, ptfe_nginx
will be changed to tfe-nginx
. If you have downstream configuration (monitoring, log forwarding etc.) that references the older naming scheme, you will need to update container references to the new names.
APPLICATION LEVEL BREAKING CHANGES
- The demo operational mode has been removed. If you are currently running demo mode, we strongly suggest that you upgrade to Terraform Enterprise v202204-1. To do this, you must migrate your application data.
APPLICATION LEVEL FEATURES
Added a log storage memory limit to log forwarding. When the 128MB limit is reached, logs will be stored in a buffer on the filesystem until they can be forwarded.
Changed cost estimation so that it uses the HTTP proxy settings configured within Terraform Enterprise.
Added the ability to specify an 'SSO team ID' for teams that Terraform Enterprise can use to map teams to non-human readable 'MemberOf' values in SAML assertions.
Updated display of download count metrics for modules in the private registry
Added an API endpoint to fetch a workspace's current state version outputs. Refer to https://www.terraform.io/cloud-docs/api-docs/state-version-outputs#show-current-state-version-outputs-for-a-workspace for details
Updated the default
json-file
log rotation settings for all containers in order to improve the performance of support bundle generation.Previous Settings New Settings max-size: 10m
andmax-file: 3
max-size: 8m
andmax-file: 4
max-size: 10m
andmax-file: 10
max-size: 32m
andmax-file: 4
max-size: 100m
andmax-file: 10
max-size: 64m
andmax-file: 8
max-size: 100m
andmax-file: 200
max-size: 50m
andmax-file: 20
APPLICATION LEVEL BUG FIXES
- The
iact_subnet_list
setting now allows you to use,
to separate IPv4 addresses. - Fixed situation where occasionally you could not create a workspace after a workspace with the same name was deleted.
- Fixed an issue causing configuration version tarballs downloaded through the API to have a non-human-readable filename and no file extension.
- Fixed several issues with Terraform Cloud agent pools.
APPLICATION LEVEL SECURITY FIXES
- Added no-cache and no-store headers for some API responses that may contain sensitive data (2FA configuration, SSO configuration, and state versions).
- Removed credentials from health-check endpoint for an internal service.
- Adopted container updates to address reported vulnerabilities in underlying packages / dependencies.