Git Basics

Why Git matters in DevOps

Everything starts with a commit: application code, pipeline definitions, Dockerfiles, Terraform and Kubernetes manifests all live in Git, so every change is reviewed, versioned and can trigger automation.

The core workflow

git clone https://github.com/acme/api.git
git status                  # what changed?
git add src/app.py          # stage a change
git commit -m "Add health endpoint"
git push                    # share it; CI starts

git init -q -b main api && cd api
echo 'print("hello")' > app.py
git add app.py
git commit -qm "Add app"
echo 'print("health: ok")' >> app.py
git status --short
git commit -qam "Add health output"
git log --oneline

 M app.py
cb43f0c Add health output
1a5901f Add app

Looking at history

git log --oneline, git show <commit>, git diff (unstaged), git diff --staged and git blame file answer "what changed, when and why".

Undoing things

Never rewrite history that others have pulled (no force-push to main). git revert is the safe way to undo a bad change in production code.

git init -q -b main api && cd api
echo "timeout=30" > config.ini && git add . && git commit -qm "Add config"
echo "timeout=1" > config.ini && git commit -qam "Lower timeout"        # a bad change
git revert --no-edit HEAD > /dev/null
git log --oneline
cat config.ini

c29b8f4 Revert "Lower timeout"
ed26f90 Lower timeout
323418b Add config
timeout=30