README.md.gotmpl

# CTFd Helm Chart
{{ template "chart.badgesSection" . }}
[![Lint and Server-side Dryrun Chart](https://github.com/ScribblerCoder/CTFd-Helm/actions/workflows/helm-lint-test.yaml/badge.svg)](https://github.com/ScribblerCoder/CTFd-Helm/actions/workflows/helm-lint-test.yaml)

This is a Helm chart for deploying CTFd on Kubernetes. It deploys the official [CTFd Docker image](ghcr.io/ctfd/ctfd). HA and Autoscaling + MariaDB + Redis + S3 Storage. Also supports using external MariaDB/Redis/S3.

## Add the helm repo
```bash
helm repo add ctfd https://scribblercoder.github.io/CTFd-Helm
```
## Install
```bash
helm install ctfd ctfd/ctfd
# OR
helm install ctfd ctfd/ctfd -f values.yaml
```

## Install from source

Build helm dependencies (MariaDB/Redis/SeaweedFS) before installing the chart.

```bash
helm dependency update
```

Set the values in `values.yaml` to your desired configuration. Then install

```bash
helm install release-name . -f values.yaml --create-namespace --namespace ctfd
```

## Uninstall
```bash
helm uninstall release-name --namespace ctfd
```

## Info

- CTFd `SECRET_KEY` is automatically generated during installation/upgrade. You can find it in the secret `release-name-ctfd-secret-key`. This secret is injected as environment variable in all CTFd pods.
- Redis in this chart uses single master with multiple workers.
- This chart deploys SeaweedFS S3 as an uploadprovider. You can use AWS S3 or any other external S3 compatible storage as an upload provider. Just set `seaweedfs.enabled` to `false` and configure the external S3 provider in `ctfd.uploadprovider.s3`.
- This chart intentionally refrains from supporting `filesystem` uploadprovider. This needs `ReadWriteMany` PVCs which are expensive in cloud providers and not recommended for production use. S3 is fast and cheap.

## Values examples

### Deploy Bitnami MariaDB/Redis and SeaweedFS S3
```yaml
ctfd:
  image:
    tag: "latest"
  autoscaling:
    enabled: true
    minReplicas: 2
    maxReplicas: 10
mariadb:
  enabled: true
  architecture: standalone
  primary:
    size: 2Gi
redis:
  enabled: true
seaweedfs:
  enabled: true
  s3:
    enabled: true
```

### Configure your own external DB/Redis/S3
```yaml
ctfd:
  image:
    tag: "latest"
  autoscaling:
    enabled: true
    minReplicas: 2
    maxReplicas: 10
  uploadprovider:
    s3:
      bucket: ""
      endpoint_url: ""
      secret_access_key: ""
      access_key_id: ""
mariadb:
  enabled: false
  external:
    port: 3306
    host: ""
    username: ""
    password: ""
    database: ""
redis:
  enabled: false
  external:
    port: 6379
    host: ""
    username: ""
    password: ""
seaweedfs:
  enabled: false
```

### Disable CTFd autoscaling
```yaml
ctfd:
  image:
    tag: "latest"
  replicas: 2
  autoscaling:
    enabled: false
  resources:
    limits:
      cpu: "2"
      memory: 2Gi
    requests:
      cpu: "1"
      memory: 1Gi
```

## Features

- [x] HA and horizontal autoscaling with CPU and memory metrics
- [x] Configurable CPU/memory requests and limits
- [x] Deploys bitnami Redis, bitnami MariaDB and SeaweedFS S3 as Helm dependencies
- [X] Option to use AWS S3 or any other external S3 compatible storage as an upload provider
- [x] Option to use external Redis and MariaDB (e.g., AWS RDS, ElastiCache)
- [x] Customizable CTFd configuration
- [x] Adjustable configurations for Redis and MariaDB
- [x] Integration with external storage as upload provider (AWS S3 or SeaweedFS or any S3 compatible storage)
- [x] Liveness and Readiness checks
- [x] Affinity/Toleration/nodeSelector rules
- [x] Automatically rolls out config updates to CTFd pods (Using checksum annotation)
- [ ] Deploys self-hosted mail server for CTFd email notifications as a helm dependency
- [ ] Automated backups (CTFd export. This could be done with batch/v1 CronJob)
- [ ] Deploys postgres db as a helm dependency (ctfd.io doesn't actively support it so this is a low priority)
- [ ] Support for custom CTFd themes/plugin (using initContainers?)

## To Do

- [ ] Performance testing to verify autoscaling capabilities + e2e testing for verification
- [x] Fine tune cpu/mem requests and limits
- [ ] Chaos testing to verify HA capabilities
- [x] Add Pod Disruption budget and rolling strategy
- [ ] Security testing to verify deployment security
- [x] Helm linting and testing with GitHub Actions
- [ ] Publish Helm chart to Artifact Hub or to Github Pages
- [x] Custom NOTES.txt (post-installation message)
- [ ] Support custom metrics for autoscaling
- [x] README.md with badges and detailed information
- [x] Add Chart Values table to README.md
- [ ] Support custom CTFd themes/plugin

{{ template "chart.maintainersSection" . }}

{{ template "chart.sourcesSection" . }}

{{ template "chart.requirementsSection" . }}

{{ template "chart.valuesSection" . }}

Autogenerated from chart metadata using [helm-docs](https://github.com/norwoodj/helm-docs)