DevOps Examples

Kubernetes applications that use ConfigHub for fleet-wide operations.

The Problem

Argo and Flux deploy configurations. They don't help you:

1. Query a large fleet - "Which of my 500 deployments run image X?"
2. Make fleet-wide changes - "Patch all of them to image Y"
3. Build an API for this - Programmatic access, not file editing

These tools reconcile Git → Cluster. They don't provide an operational layer for querying and mutating configurations at scale.

What ConfigHub Adds

Need	GitOps Tools	ConfigHub
Query fleet	Grep across repos	cub unit list --where "Data CONTAINS 'image:v1'"
Bulk change	Edit files, commit, PR, wait	cub run set-image --space '*-prod-*'
API access	Build your own	ConfigHub API + SDK
See all states	Argo UI (per-app)	Desired vs Live vs Drift (fleet-wide)

Multiple Views of State

View	What it is	Where it lives
Desired	What you declared	ConfigHub unit
Live	What's actually in the cluster	Queried via BridgeWorker
Drift	Are they equal?	Computed

# See what you declared
cub unit get-data trade-service --space prod-eu
# See what's actually running
cub unit livestate trade-service --space prod-eu
# See the difference
cub unit diff trade-service --space prod-eu

Argo shows sync status per-application. ConfigHub shows Desired/Live/Drift across the entire fleet.

The model: Git is the source. CI syncs Git → ConfigHub. ConfigHub provides the query/mutation layer.

Git (source) → CI syncs → ConfigHub (query + mutate) → applies → Kubernetes
↑
These examples use ConfigHub here
│
└── sync back to Git (PR) when needed

Examples

Operational Today

drift-detector

Detects when Kubernetes runtime state differs from ConfigHub units.

Problem it solves: "Something changed my deployment, but I don't know what or when."

How it uses ConfigHub: Queries all units, compares to kubectl output, reports drift, optionally auto-corrects.

cost-optimizer

Analyzes resource usage and suggests right-sizing.

Problem it solves: "My clusters are over-provisioned but I don't know where to cut."

How it uses ConfigHub: Queries units for resource requests/limits, correlates with metrics-server data, suggests patches.

cost-impact-monitor

Pre-deployment cost analysis.

Problem it solves: "I want to know the cost impact before I deploy."

How it uses ConfigHub: Hooks into unit apply events, calculates cost delta, reports before deployment completes.

Planned

cve-responder (SPEC only)

Automated CVE response across infrastructure.

Problem it solves: "A critical CVE dropped. I need to find and patch all affected deployments in minutes, not hours."

How it uses ConfigHub:

1. Query: cub unit list --where "Data CONTAINS 'vulnerable-image:1.0'" (seconds)
2. Patch: cub run set-image --image 'patched:1.1' --where "..." (bulk update)
3. Apply: cub unit apply --where "..." (immediate, no PR wait)
4. Sync: Open PR to Git with the changes (eventual consistency)

See cve-responder/SPEC.md for details.

config-lineage (SPEC only)

Configuration inheritance visualization.

Problem it solves: "Why does prod-eu have replicas=5? Where did that value come from?"

How it uses ConfigHub:

• Walk upstream chain: prod-eu → prod → base
• Show which layer set each value
• Impact analysis: "What downstream units are affected if I change base?"

See config-lineage/SPEC.md for details.

For Argo/Flux Users

ConfigHub doesn't replace your GitOps workflow. It adds what's missing:

1. Git stays the source - your manifests stay in Git
2. CI syncs to ConfigHub - same CI pipeline, additional sync target
3. Query and mutate via ConfigHub - fleet-wide visibility and changes
4. Sync back to Git - PRs keep Git consistent after operational changes

Argo/Flux continue to deploy. ConfigHub provides the operational API they lack.

Quick Start

Each example has setup instructions in its directory. Prerequisites:

# ConfigHub CLI
curl -fsSL https://hub.confighub.com/cub/install.sh | bash
cub auth login
# Kubernetes (local)
kind create cluster
# Verify setup
curl -fsSL https://raw.githubusercontent.com/monadic/devops-sdk/main/test-confighub-k8s | bash

Project Structure

devops-examples/
├── drift-detector/ # Runtime drift detection
├── cost-optimizer/ # Resource right-sizing
├── cost-impact-monitor/ # Pre-deploy cost analysis
├── cve-responder/ # CVE response automation (SPEC)
└── config-lineage/ # Inheritance visualization (SPEC)

All examples use the devops-sdk for ConfigHub operations.

Documentation Standards

Commands in README files are validated before commit:

# Validate cub commands in docs
curl -fsSL https://raw.githubusercontent.com/monadic/devops-sdk/main/cub-command-analyzer.sh | bash -s -- .

License

Proprietary - ConfigHub, Inc.