gitlab_evaluate 0.5.0

Scans GitLab instance and ranks projects against a set of criteria. Can be used to identiy projects that may have too much metadata/size to reliably export or import.

Evaluate

Evaluate is a script that can be run to gather information about projects from a gitlab self-managed instance. This information is useful to the GitLab Professional Services (PS) team to accurately scope migration services.

Use Case

GitLab PS plans to share this script with a Customer to run against their self managed instance. Then the customer can send back the output files to enable GitLab engagement managers to scope engagements accurately.

Install

pip install gitlab-evaluate

Usage

System level data gathering

Evaluate is meant to be run by an administrator of a GitLab Self Managed deployment to gather data about every project on the instance.

1. A GitLab system administrator should provision an access token
2. Then, after installing gitlab-evaluate from the Install section above,
3. Run :point_down:

# For evaluating a GitLab instance
evaluate-gitlab -t <access-token-with-api-admin-privileges> -s https://gitlab.example.com

1. This should create a file called evaluate_output.csv
2. If you're coordinating a GitLab Professional Services engageemnt, email this file to the GitLab account team.

To gather CI data from a single repo

# For evaluating a single git repo's CI readiness
evaluate-ci-readiness -r|--repo <git-repo-url>

Command help screen

usage: evaluate-gitlab [-h] [-t TOKEN] [-s SOURCE] [-f FILENAME] [-o] [-i] [-p PROCESSES]

optional arguments:
  -h, --help            show this help message and exit
  -t TOKEN, --token TOKEN
                        Personal Access Token: REQ'd
  -s SOURCE, --source SOURCE
                        Source URL: REQ'd
  -f FILENAME, --filename FILENAME
                        CSV Output File Name. If not set, will default to 'evaluate_output.csv'
  -o, --output          Output Per Project Stats to screen
  -i, --insecure        Set to ignore SSL warnings.
  -p PROCESSES, --processes PROCESSES
                        Number of processes. Defaults to number of CPU cores

usage: evaluate-ci-readiness [-h] [-r REPO]

optional arguments:
  -h, --help            show this help message and exit
  -r REPO, --repo REPO  Git Repository To Clone (ex: https://username:[e-mail]

Using a docker container

Docker containers with evaluate installed are also available to use.

Local usage

# Spin up container
docker run --name evaluate -it registry.gitlab.com/gitlab-org/professional-services-automation/tools/utilities/evaluate:latest /bin/bash

# In docker shell
evaluate-ci-readiness -r|--repo <git-repo-url>
evaluate-gitlab -t <access-token-with-api-admin-privileges> -s https://gitlab.example.com

Example GitLab CI job using evaluate ci readiness script

evaluate node-js:
  stage: test
  script:
    - evaluate-ci-readiness --repo=https://github.com/nodejs/node.git
  artifacts:
    paths:
      - node.csv

To test, consider standing up local docker container of gitlab. Provision a personal access token of a user who has system admin priviledges. Create multiple projects with varying number of commits, pipelines, merge requests, issues. Consider importing an open source repo or using GPT to add projects to the system.

Design

Design for the script can be found here

Project Thresholds

Below are the thresholds we will use to determine whether a project can be considered for normal migration or needs to have special steps taken in order to migrate

Project Data

• Pipelines - 1,500 max
• Issues - 1,500 total (not just open)
• Merge Requests - 1,500 total (not just merged)
• Container images - 20GB per project
• Packages - Any packages present

Repo Data

• commits - 20K
• branches - 1K
• tags - 1K
• Disk Size - 10GB