What is Hydra?¶

Hydra is a tool that automates deploying and managing applications on Kubernetes clusters. Think of it as an autopilot for your servers — instead of manually installing software on each server one by one, you describe what you want in files, and Hydra makes it happen.

The Problem Hydra Solves¶

Imagine you have 10 Kubernetes clusters, each running 20+ applications. Without Hydra, you would need to:

SSH into (or connect to) each cluster
Run helm install for each application, one by one
Remember the right configuration values for each cluster
Manually track what's installed where
Repeat everything when a new cluster is needed

This is a full-day job just for one cluster. For 10 clusters, it's a nightmare.

How Hydra Solves It¶

Hydra replaces all of that with:

A directory structure where you define "cluster X should run applications A, B, C with these settings"
A single command to install everything: hydra gitops apply prod.**
Automatic value merging — shared settings are defined once, cluster-specific overrides are defined per cluster
A web UI to visualize what's deployed, what depends on what, and what the configuration looks like

Hydra currently exposes these command modes:

hydra local for working only with the local Hydra definitions.
hydra gitops for working with those local definitions plus the live Kubernetes cluster.
hydra cluster as the reserved future command surface for cluster-only workflows.

Hydra in One Sentence¶

Hydra is a CLI tool and directory structure that lets you define, deploy, and manage all applications across multiple Kubernetes clusters from a single Git repository.

The Three Parts of Hydra¶

1. The Directory Structure¶

Two directories in this repository define everything:

gitops-repository/ — defines which clusters exist and their configuration (values, secrets)
charts-repository/ — contains the shared Helm charts (application packages) that clusters can install

2. The CLI (`hydra-go`)¶

A command-line tool written in Go. Key commands:

What you want to do	Command
See what an app's config looks like	`hydra local values prod.infra.cert-manager`
Preview what would be deployed	`hydra local template prod.infra.cert-manager`
See what's different on the live cluster	`hydra gitops diff prod.infra.*`
Deploy everything	`hydra gitops apply prod.**`
Set up a brand-new cluster from scratch	`hydra gitops apply prod.** --bootstrap`

3. The Web UI (`hydra-ui`)¶

A static web page (no backend server needed) that loads an exported file and shows you:

Which applications are deployed
How they depend on each other (e.g., "this app needs that database secret")
What configuration values each app receives
Who has permissions to do what (RBAC)

How It Fits Into the Bigger Picture¶

┌─────────────────────────────────────────────────────────────────┐
│                        This Repository                          │
│                                                                 │
│  ┌──────────────────┐  ┌────────────────────┐  ┌─────────────┐  │
│  │ gitops-repository│  │ charts-repository  │  │   hydra-go/ │  │
│  │                  │  │                    │  │             │  │
│  │ "Which clusters  │  │ "What apps are     │  │ "The tool   │  │
│  │  exist and what  │  │  available and how │  │  that puts  │  │
│  │  they should run"│  │  they're packaged" │  │  it all     │  │
│  │                  │  │                    │  │  together"  │  │
│  └──────────────────┘  └────────────────────┘  └─────────────┘  │
└─────────────────────────────────────────────────────────────────┘

The Goal¶

The goal of Hydra is to turn cluster setup from a full-day manual process into a single command. And to make it so that Git is the only source of truth — if you want to know what's running on a cluster, you look at the Git repository, not at the cluster itself. This approach is called GitOps.

Next Steps¶

Key Concepts — learn the technical terms used throughout this documentation
Quickstart — the shortest path to running your first Hydra command