The primary user persona for Distribution is the system administrator responsible for managing a GitLab instance. The team goals are to make it as easy as possible to deploy, scale, upgrade, and fine tune a GitLab instance on a range of on-prem and cloud platforms.
Target deployments are anything from a single node deployment for evaluating GitLab all the way through to the 50K user reference architecture and beyond. The ultimate goal is to ensure the end user has a high-speed experience when using GitLab with no downtime or disruptions.
The Omnibus package and Helm Chart are the primary deployment methods that Distribution focuses on. In FY21, the team will:
Distribution team vision is shown below, and not limited to:
Distribution ensures the experience of installing and maintaining GitLab is easy and safe for everyone. It is Distribution team's task to think about the widest variety of installation/update use-cases and provide a solution that will satisfy most needs. Distribution is there to provide the best possible experience for a user that is a novice but also a veteran when it comes to installing and maintaining software.
Increase # of active installations
Reduce average days behind latest version
The following people are members of the Distribution Team:
|Steven Wilson||Backend Engineering Manager, Distribution|
|David 'DJ' Mountney||Staff Distribution Engineer|
|Jason Plum||Staff Distribution Engineer|
|Balasankar "Balu" C||Distribution Engineer|
|Robert Marshall||Distribution Engineer|
|Hossein Pursultani||Distribution Engineer|
|Gerard Hickey||Senior Distribution Engineer|
|Dustin Collins||Distribution Engineer|
|Mitch Nielsen||Distribution Engineer|
|Dmitry Makovey||Senior Distribution Engineer|
The following members of other functional teams are our stable counterparts:
Distribution team is responsible for:
Based on team responsibilities, the following objectives apply:
Exception Ratio: 2 Staff Eng:Team
Justification: The Distribution team is tasked with maintaining the Omnibus package and Helm Chart deployment methods. Expertise is required in both traditional software packaging and cloud native technologies. Having two Staff engineers allows the team to move both deployment models forward while minimizng drift between them. One Staff engineer has a more broad focus on commmunity contributions, reviews, maintenance and some feature work (tech lead) while the other is tightly aligned to cloud native projects inlcuding OpenShift, product maturity and large deployments (architect).
Future Growth or Anticipated Change: It is expected that no significant changes during FY22-FY23 will influence the gearing ratio.
omnibus-gitlab - This project creates platform-specific, self-contained GitLab packages and images for self-managed consumption in cloud environments and on-premisis hosting.
GitLab's cloud native solutions span a collection of image, Helm chart and Operator projects to deploy and manage GitLab on container platforms including Kubernetes and OpenShift.
|Omnibus GitLab||gitlab-org/omnibus-gitlab||Build Omnibus packages with HA support for LTS versions of all major Linux operating systems such as Ubuntu, Debian, CentOS/RHEL, OpenSUSE, SLES|
|Docker All in one GitLab image||gitlab-org/omnibus-gitlab/docker||Build Docker images for GitLab CE/EE based on the omnibus-gitlab package|
|GitLab Helm Chart||gitlab-org/charts/gitlab||Cloud Native GitLab Helm Charts|
|Docker images for GitLab Helm Chart||gitlab-org/build/CNG||Individual images used by GitLab Helm Charts|
|GitLab OpenShift operator||gitlab-org/gl-openshift/gitlab-operator||The GitLab operator creates and manages GitLab instances/deployments in a container platform such as Openshift or Kubernetes. It will run an any environment that provides native Kubernetes resources including deployments, statefulsets, services, ingress/openshift routes, persistent volume claims, persistent volumes, etc.|
|Kubernetes Helm Charts index||charts/charts.gitlab.io||Helm charts repository index|
|AWS images||AWS marketplace||AWS image based on the omnibus-gitlab package|
|GitLab provisioner||gitlab-org/distribution/gitlab-provisioner||Creates a GitLab HA installation using Ansible and Terraform|
|Reference Architecture Tester||gitlab-org/distribution/reference-architecture-tester||Spins up reference architecture based GitLab deployments using GET and runs QA against them|
|GitLab Terraform configuration||gitlab-terraform||Terraform configuration for various cloud providers|
|Omnibus GitLab Builder||GitLab Omnibus Builder||Create environment containing build dependencies for the omnibus-gitlab package|
|Upgrade time metrics||Upgrade time metrics page on GL Pages||Stores the result of calculation of time needed for upgrade between versions in chart form. Backed by Google sheets. Hosted using GL Pages|
|Licenses of bundled dependencies||Licenses page on GL Pages||Webpage listing the bundled dependencies in each package along with their license.|
All work carried out by the Distribution team is public. Some exceptions apply:
Some of the team work is carried out on our development server at
Infrastructure overview document lists the reasons.
Unless your work is related to the security, all other work is carried out in projects on
If you need to submit a sensitive issue, please use confidential issues.
If you are unsure whether something needs to remain private, check with the team Engineering Manager.
The team regularly publishes demos, discussions and meetings to these playlists:
In addition to general company on-boarding and off-boarding, Distribution team has its own process to get new team members up to speed more quickly.
If you are starting with your onboarding, open an issue in Distribution team issue tracker, select
Team-onboarding template and assign the issue to yourself.
Going through the steps noted in the issue should be your top priority, higher than the general company on-boarding issue. This is because items in team on-boarding are specific to your role and it will allow you to get up-to-speed quicker.
Off-boarding should be carried out by the Engineering Manager of the team, using the appropriate issue template in the same issue tracker.
General resources available to developers are listed in the Engineering handbook.
In the Distribution team specifically, everyone should have access to the following resources:
If you don't have access to any of these resources, create an Access Request and assign it to your manager for approval.
As part of the team responsibilities, team owns maintenance of infrastructure used for day to day work. For list of nodes and description of the maintenance tasks, see the infastructure and maintenance page.
General engineering workflow applies to the Distribution team. Since Distribution is working across multiple projects, our team workflow is further explained and summarised on the Distribution workflow page.
Everything that is done in GitLab will end up in the supported installation methods that are distributed to users. While that sounds like the last link in the chain, it is one of the most important ones. This means that informing the Distribution team of a change in an early stage is crucial for releasing your feature. While last minute changes are inevitable and can happen, we should strive to avoid them.
We expect every team to reach out to the Distribution team before scheduling a feature in an upcoming release in the following cases:
To sum up the above list:
If you need to do
chmod, do compilation or configuration change in any part of GitLab stack, you
need to reach out to the Distribution team for opinion as early as possible.
This will allow us to schedule appropriately the changes (if any) we have to make to the packages.
If a change is reported late in the release cycle or not reported at all, your feature/change might not be shipped within the release.
If you have any doubt whether your change will have an impact on the Distribution team, don't hesitate to ping us in your issue and we'll gladly help.
There are occasions where the experise of the Distribution team may be needed in
support of a customer issue. When this does occur, the appropriate method of requesting
our engagement is by opening an issue on the Distribution team tracker
Support Request template. This process allows us to track time involved
and ensure that the right parties are involved at the correct time.
Requests should be opened two or more business days before action is needed to ensure the team has time to prepare.
How did Distribution get its name? We iterated, as always. "Distribution" was chosen as better than "Install" when renaming the "Build" team, live on an AMA with our CEO Sid.