Kubernetes (Helm)

recon-web ships with a Helm chart at helm/recon-web/ that deploys the API server, web frontend, and optional scheduled scanner as Kubernetes workloads.

Prerequisites

Kubernetes 1.25+
Helm 3.10+
kubectl configured to talk to your cluster

Quick install

Clone the repository (or add the chart from a Helm registry if published):
Terminal window
```
git clone https://github.com/brunoafk/recon-web.git
cd recon-web
```
Install with default values:
Terminal window
```
helm install recon-web ./helm/recon-web
```
Port-forward to access the UI locally:
Terminal window
```
kubectl port-forward svc/recon-web-web 8080:80
```
Open http://localhost:8080.

Custom values

Create a values-override.yaml to configure the deployment for your environment:

ingress:
  enabled: true
  className: nginx
  host: recon.example.com
  tls: true
  annotations:
    cert-manager.io/cluster-issuer: letsencrypt-prod

auth:
  enabled: true
  token: "your-secret-token-at-least-32-chars-long"

api:
  env:
    GOOGLE_CLOUD_API_KEY: "AIza..."
    VIRUSTOTAL_API_KEY: "abcdef..."
    MAX_CONCURRENCY: "4"

scanner:
  enabled: true
  schedule: "0 */6 * * *"
  urls:
    - https://example.com
    - https://another-site.com

notifications:
  telegram:
    enabled: true
    botToken: "123456789:ABCdefGHI..."
    chatId: "123456789"
  email:
    enabled: true
    smtpHost: smtp.gmail.com
    smtpPort: "587"
    smtpUser: alerts@example.com
    smtpPass: "abcd efgh ijkl mnop"
    notifyEmail: team@example.com

persistence:
  enabled: true
  size: 5Gi
  storageClass: gp3

Install with the override file:

helm install recon-web ./helm/recon-web -f values-override.yaml

Resource defaults

Component	CPU request	CPU limit	Memory request	Memory limit
API	100m	1000m	256Mi	1Gi
Web	50m	200m	64Mi	128Mi

Override these under api.resources and web.resources in your values file.

Custom images

To use images from a private registry:

api:
  image:
    repository: registry.example.com/recon-web/api
    tag: "1.0.0"
    pullPolicy: IfNotPresent

web:
  image:
    repository: registry.example.com/recon-web/web
    tag: "1.0.0"
    pullPolicy: IfNotPresent

Persistence

The API server stores scan history in a SQLite database backed by a PersistentVolumeClaim.

Setting	Default	Description
`persistence.enabled`	`true`	Create a PVC for scan data
`persistence.size`	`1Gi`	Volume size
`persistence.storageClass`	(cluster default)	Storage class to use
`persistence.accessMode`	`ReadWriteOnce`	PVC access mode

Scheduled scanner

The Helm chart can deploy a Kubernetes CronJob that runs the CLI image on a schedule:

scanner:
  enabled: true
  schedule: "0 0 * * *"
  urls:
    - https://example.com
  image:
    repository: ghcr.io/brunoafk/recon-web/cli
    tag: latest

This is an alternative to the built-in scheduler (SCHEDULE_ENABLED). The Kubernetes CronJob approach gives you native retry policies, resource limits, and pod-level observability.

Upgrading

helm upgrade recon-web ./helm/recon-web -f values-override.yaml

Uninstalling

helm uninstall recon-web