Continuous Delivery known issues

These are the known issues for the Continuous Delivery 5.x release series.

Windows jumphosts are not supported when installing or upgrading Continuous Delivery

When installing or upgrading Continuous Delivery, the Bolt runner cannot be a Microsoft Windows system.

Continuous Delivery must be at version 4.29.1 before upgrading to 5.3.0 or 5.3.1

Affects version: 5.3.0 and 5.3.1

Resolved in: 5.3.2

When upgrading to Continuous Delivery 5.3.0 or 5.3.1 from 4.x, you must have upgraded Continuous Delivery to version 4.29.1 before you can migrate to version 5.3.1.

Creating or editing a pipeline with deployments causes the deployments to stop running

Affects version: 5.3.0

Resolved in: 5.3.1

Using the Continuous Delivery web UI to create or edit a pipeline that contains a deployment causes the pipeline to stop running its deployments. Jobs and Impact Analyses continue to work and pipelines continue to run, but deployments are skipped. This does not affect pipelines that do not contain deployments, pipelines managed with code, or other versions of Continuous Delivery.

Optional environment variables are not being passed to Docker containers

Affects version: 5.1.2 and earlier

Resolved in: 5.2.0

For Continuous Delivery, on the Jobs page, environment variables that are added to the Optional configuration section are unavailable in the Docker container. To work around this issue, add a secret to the job with the environment data you want to use. You can then reference the variable it creates to access your data.

A Puppet Enterprise (2023.4 or later) module is not compatible with Ruby 3.2

Affects version: 5.1.2 and earlier

Resolved in: 5.2.0

A module included with Puppet Enterprise 2023.4 or later is not compatible with Ruby 3.2. While this does not affect Puppet Enterprise users, it can affect users integrating Continuous Delivery with these versions of PE. If you are integrating Continuous Delivery with Puppet Enterprise 2023.4 or later, install a version of the module compatible with newer Ruby versions using the following command:

puppet module upgrade puppetlabs-cd4pe_jobs

Continuous Delivery fails when used with Puppet 8 (Puppet Enterprise 2023.4 or later)

Affects version: 5.1.2 and earlier

Resolved in: 5.2.0

Puppet 8 (included with Puppet Enterprise 2023.4 or later) has removed legacy facts by default. Continuous Delivery relies on the legacy fact operatingsystem in order to develop a list of systems that are available to be used as job hardware. If you are running Continuous Delivery with Puppet Enterprise 2023.4 or later, you need to add the following to puppet.conf:

include_legacy_facts=true

Continuous Delivery fails to restart after rebooting host

Affects version: 5.1.2 and earlier

Resolved in: 5.2.0

If you are running Continuous Delivery with the Podman runtime, Continuous Delivery may not restart after a host reboot. Run the following (using puppetlabs-cd4peadm) to get Continuous Delivery back up:

bolt plan run cd4peadm::ctl action=stop
bolt command run "rm /etc/puppetlabs/cd4pe/*.ip.txt" --targets cd4pe-5-host
bolt plan run cd4peadm::ctl action=start

Continuous Delivery fails to install on localhost

Affects version: 5.1.2 and earlier

Resolved in: 5.2.0

Users installing Continuous Delivery using Bolt's local transport may encounter an error if /etc/puppetlabs has not been created. To work around this issue, do one of the following:

  • Create the directory /etc/puppetlabs before installing
  • If you are using Bolt 3.30.0 or later, remove bundled-ruby: false from your inventory.yaml

Jobs fail in 5.x due to improperly configured CA certificate

Affects version: 5.1.2 and earlier

Resolved in: 5.2.0

Users running Continuous Delivery may encounter an SSL error with the message certificate verify failed (certificate signature failure) when running jobs. To work around this, restart the UI container on the Continuous Delivery host.