Connect a New GitLab Repository or Group

How to get a new GitLab repository or group building and deploying through SVIEW, what to send in your request, and how to move projects safely.

This guide explains how to get a new GitLab repository, or repositories in a
different GitLab group, building and deploying through SVIEW.

When you need this

  • You have a new application in GitLab and want it deployed to one or more
    Environments.
  • You are moving existing projects to a different GitLab group or namespace,
    or copying them into a new one, and want them to keep deploying.
  • You want the same pipeline setup you already have in one group to work in a
    second group.

In every case the pipeline connection has to be set up for the new location.
It does not follow a project automatically.

Prerequisites

  • GitLab access for CTO2B on the new project or group. Grant the
    cto2bgeneric account Maintainer on each project. If the pipeline will
    use CI/CD variables defined on the group — the usual case when several
    projects share the same registry credentials and SVIEW endpoint — grant
    Owner on that group instead, because GitLab does not let a Maintainer
    read or set group-level variables. See
    GitLab Access for CTO2B for the details.
  • A Dockerfile in the repository, so the pipeline can build an image.

What to send in your request

Raise a request in your dedicated Slack channel with:

ItemExample
GitLab group or project pathexample-group/new-api, or the whole group example-group
Which Environments it should deploy todevelopment and production
Which branch or tag deploys to which Environmentdev → development, main → production
The container image repository for each Environmentone repository per Environment, or "please create them"
Whether it is new or moved"moved from old-group on 1 October"

Use a separate image repository per Environment. Registry retention rules
apply to every image in a repository, so sharing one repository between
development and production lets routine development pushes push a production
image out of retention.

What happens next

flowchart LR
  A[Access granted] --> B[Pipeline and CI/CD variables set up]
  B --> C[Application Component added in SVIEW]
  C --> D[First push builds and deploys]
  1. Pipeline setup. We add or adapt the .gitlab-ci.yml so it builds the
    image, scans it, pushes it to the image repository, and tells SVIEW to
    deploy the new version. The CI/CD variables it needs — registry
    credentials, the SVIEW endpoint, and the deploy token — are set on the
    project or on the group.
  2. Application Component. Each target Environment gets an application
    Component for the new app, if it does not have one already. See
    Components and Deployments for how Components are organised.
  3. First deployment. The first push to the configured branch runs the full
    pipeline and deploys to the matching Environment.

Moving projects between GitLab groups

A GitLab project transfer carries the project's history and its
project-level CI/CD variables, but never the group-level variables of
the group it left. A pipeline that relied on them can keep working for a while
and then start failing with empty or invalid registry credentials, with no
change to the pipeline file.

To avoid that:

  1. Tell us before you transfer, and name the destination group.
  2. Grant CTO2B the access described above on the destination group.
  3. After the transfer, run a pipeline on each moved project and confirm it
    deploys.

A newly created top-level group starts with no group-level variables at all,
so every project moved into it needs the full setup.

If your team signs in to SVIEW with GitLab accounts, mention in the request
whether people are moving to a new group too, since SVIEW access can be tied
to GitLab group membership.

Verify it worked

  • The pipeline in GitLab completes all stages, including the deploy stage.
  • In SVIEW, the application Component in the target Environment shows the new
    image tag and a healthy status.

If the deploy stage passes but the Component shows Degraded, the problem is
in the application rather than the pipeline — see
Components and Deployments.

Related articles


Did this page help you?