- The new runner registration workflow
- Estimated time frame for planned changes
- Prevent your runner registration workflow from breaking
- Using registration tokens after GitLab 17.0
- Impact on existing runners
-
Changes to the
gitlab-runner register
command syntax - Impact on autoscaling
- Creating runners programmatically
- Installing GitLab Runner with Helm chart
- Known issues
Migrating to the new runner registration workflow
In GitLab 16.0, we introduced a new runner creation workflow that uses runner authentication tokens to register runners. The legacy workflow that uses registration tokens is deprecated and will be removed in GitLab 18.0.
For information about the current development status of the new workflow, see epic 7663.
For information about the technical design and reasons for the new architecture, see Next GitLab Runner Token Architecture.
If you experience problems or have concerns about the new runner registration workflow, or if the following information is not sufficient, you can let us know in the feedback issue.
The new runner registration workflow
For the new runner registration workflow, you:
- Create a runner directly in the GitLab UI or programmatically.
- Receive a runner authentication token.
- Use the runner authentication token instead of the registration token when you register a runner with this configuration. Runner managers registered in multiple hosts appear under the same runner in the GitLab UI, but with an identifying system ID.
The new runner registration workflow has the following benefits:
- Preserved ownership records for runners, and minimized impact on users.
- The addition of a unique system ID ensures that you can reuse the same authentication token across multiple runners. For more information, see Reusing a GitLab Runner configuration.
Estimated time frame for planned changes
- In GitLab 15.10 and later, you can use the new runner registration workflow.
- In GitLab 17.0, we plan to disable runner registration tokens.
- In GitLab 18.0, we plan to completely remove support for runner registration tokens.
Prevent your runner registration workflow from breaking
Until GitLab 17.0, you can still use the legacy runner registration workflow.
In GitLab 17.0, the legacy runner registration workflow will be disabled automatically. You will be able to manually re-enable the legacy runner registration workflow for a limited time. For more information, see Using registration tokens after GitLab 17.0.
If no action is taken before your GitLab instance is upgraded to GitLab 17.0, then your runner registration
workflow will break, and the gitlab-runner register
command will receive a 410 Gone - runner registration disallowed
error.
To avoid a broken workflow, you must:
- Create a runner and obtain the authentication token.
- Replace the registration token in your runner registration workflow with the authentication token.
Using registration tokens after GitLab 17.0
To continue using registration tokens after GitLab 17.0:
- On GitLab.com, you can manually enable the legacy runner registration process in the top-level group settings until GitLab 18.0.
- On GitLab self-managed, you can manually enable the legacy runner registration process in the Admin area settings until GitLab 18.0.
Impact on existing runners
Existing runners will continue to work as usual even after 18.0. This change only affects registration of new runners.
The GitLab Runner Helm chart generates new runner pods every time a job is executed. For these runners, enable legacy runner registration to use registration tokens. In GitLab 18.0 and later, you must migrate to the new runner registration workflow.
Changes to the gitlab-runner register
command syntax
The gitlab-runner register
command will stop accepting registration tokens and instead accept new runner
authentication tokens generated in the GitLab runners administration page.
The runner authentication tokens are recognizable by their glrt-
prefix.
When you create a runner in the GitLab UI, you specify configuration values that were previously command-line options
prompted by the gitlab-runner register
command.
These command-line options have been deprecated.
If you specify a runner authentication token with:
- the
--token
command-line option, thegitlab-runner register
command does not accept the configuration values. - the
--registration-token
command-line option, thegitlab-runner register
command ignores the configuration values.
Token | Registration command |
---|---|
Runner authentication token |
gitlab-runner register --token $RUNNER_AUTHENTICATION_TOKEN
|
Runner registration token (deprecated) |
gitlab-runner register --registration-token $RUNNER_REGISTRATION_TOKEN <runner configuration arguments>
|
Authentication tokens have the prefix, glrt-
.
To ensure minimal disruption to your automation workflow,
legacy-compatible registration processing
triggers if a runner authentication token is specified in the legacy parameter --registration-token
.
Example command for GitLab 15.9:
gitlab-runner register \
--non-interactive \
--executor "shell" \
--url "https://gitlab.com/" \
--tag-list "shell,mac,gdk,test" \
--run-untagged "false" \
--locked "false" \
--access-level "not_protected" \
--registration-token "REDACTED"
In GitLab 15.10 and later, you create the runner and some of the attributes in the UI, like the
tag list, locked status, and access level.
In GitLab 15.11 and later, these attributes are no longer accepted as arguments to register
when a runner authentication token with the glrt-
prefix is specified.
The following example shows the new command:
gitlab-runner register \
--non-interactive \
--executor "shell" \
--url "https://gitlab.com/" \
--token "REDACTED"
Impact on autoscaling
In autoscaling scenarios such as GitLab Runner Operator or GitLab Runner Helm Chart, the registration token is replaced with the runner authentication token generated from the UI. This means that the same runner configuration is reused across jobs, instead of creating a runner for each job. The specific runner can be identified by the unique system ID that is generated when the runner process is started.
Creating runners programmatically
In GitLab 15.11 and later, you can use the POST /user/runners REST API to create a runner as an authenticated user. This should only be used if the runner configuration is dynamic or not reusable. If the runner configuration is static, you should reuse the runner authentication token of an existing runner.
For instructions about how to automate runner creation and registration, see the tutorial, Automate runner creation and registration.
Installing GitLab Runner with Helm chart
Several runner configuration options cannot be set during runner registration. These options can only be configured:
- When you create a runner in the UI.
- With the
user/runners
REST API endpoint.
The following configuration options are no longer supported in values.yaml
:
## All these fields are DEPRECATED and the runner WILL FAIL TO START with GitLab Runner 18.0 and later if you specify them.
## If a runner authentication token is specified in runnerRegistrationToken, the registration will succeed, however the
## other values will be ignored.
runnerRegistrationToken: ""
locked: true
tags: ""
maximumTimeout: ""
runUntagged: true
protected: true
The replacement field for the invalid runnerRegistrationToken
field is the runnerToken
field. In the context of the GitLab Runner on Kubernetes, Helm deploy passes the runner authentication token
to the runner worker pod and the runner configuration is created. If you continue to use the runnerRegistrationToken
token field on Kubernetes hosted runners attached to GitLab.com, then the runner worker pod tries, on creation, to use the Registration API method that is no longer supported as of GitLab 17.0.
If you store the runner authentication token in secrets
, you must also modify them.
In the legacy runner registration workflow, fields were specified with:
apiVersion: v1
kind: Secret
metadata:
name: gitlab-runner-secret
type: Opaque
data:
runner-registration-token: "REDACTED" # DEPRECATED, set to ""
runner-token: ""
In the new runner registration workflow, you must use runner-token
instead:
apiVersion: v1
kind: Secret
metadata:
name: gitlab-runner-secret
type: Opaque
data:
runner-registration-token: "" # need to leave as an empty string for compatibility reasons
runner-token: "REDACTED"
runner-registration-token
,
you can set it to any string - it will be ignored when runner-token
is present.Known issues
Pod name is not visible in runner details page
When you use the new registration workflow to register your runners with the Helm chart, the pod name is not visible in the runner details page. For more information, see issue 423523.
Runner authentication token does not update when rotated
Token rotation with the same runner registered in multiple runner managers
When you use the new workflow to register your runners on multiple host machines and the runner authentication token rotates automatically, only the first runner manager to handle the token renewal request receives the new token. The remaining runner managers continue to use the invalid token and become disconnected. You must update these managers manually to use the new token.
Token rotation in GitLab Operator
When you use the new registration workflow to register your runners with the GitLab Operator, the runner authentication token referenced by the Custom Resource Definition does not update when the token is rotated. This occurs when:
- You’re using a runner authentication token (prefixed with
glrt-
) in a secret referenced by a Custom Resource Definition. - The runner authentication token is due to expire. For more information about runner authentication token expiration, see Authentication token security.
For more information, see issue 186.