HTTPCache

The HTTPCache service is a managed cache system based on Nginx designed to optimize web traffic and improve web application performance. It combines customized Nginx instances with HAProxy to provide efficient caching and load balancing.

---

Configuration Example

Here is a YAML configuration example to deploy HTTPCache with two replicas for HAProxy and Nginx:

apiVersion: apps.cozystack.io/v1alpha1
kind: HTTPCache
metadata:
  name: httpcache-example
spec:
  external: true
  size: 20Gi
  storageClass: "replicated"
  haproxy:
    replicas: 2
  nginx:
    replicas: 2
  endpoints:
    - url: "https://example-origin.com"
    - url: "https://another-origin.com"

Using the kubeconfig provided by Hikube and this example yaml, saved as a manifest.yaml file, you can easily test the application deployment using the following command:

kubectl apply -f manifest.yaml

---

Main Features

• Nginx Modules and Integrations:

  • VTS module for statistics.
  • Integration with ip2location and ip2proxy for IP geolocation.
  • 51Degrees support for device detection.
  • Cache purge functionality.

• HAProxy Role:

  • HAProxy uses a consistent hash based on URL to direct traffic to appropriate Nginx instances.
  • Active/backup operation for high availability.

• Cache Storage:

  • Each Nginx instance uses a Persistent Volume Claim (PVC) to store cached content, ensuring fast and reliable access to frequently used resources.

---

Deployment Architecture

The deployment architecture is divided into three main layers:

1. Load Balancer: HAProxy for load balancing.
2. Caching Layer: Nginx instances with PVC.
3. Origin Layer: Origin servers for non-cached content.

Here is a diagram illustrating the architecture:

      ┌─────────┐
      │ metallb │ arp announce
      └────┬────┘
           │
           │
   ┌───────▼───────────────────────────┐
   │  kubernetes service               │  node
   │ (externalTrafficPolicy: Local)    │  level
   └──────────┬────────────────────────┘
              │
              │
         ┌────▼────┐  ┌─────────┐
         │ haproxy │  │ haproxy │   loadbalancer
         │ (active)│  │ (backup)│      layer
         └────┬────┘  └─────────┘
              │
              │ balance uri whole
              │ hash-type consistent
       ┌──────┴──────┬──────────────┐
   ┌───▼───┐     ┌───▼───┐      ┌───▼───┐ caching
   │ nginx │     │ nginx │      │ nginx │  layer
   └───┬───┘     └───┬───┘      └───┬───┘
       │             │              │
  ┌────┴───────┬─────┴────┬─────────┴──┐
  │            │          │            │
┌───▼────┐ ┌────▼───┐ ┌───▼────┐  ┌────▼───┐
│ origin │ │ origin │ │ origin │  │ origin │
└────────┘ └────────┘ └────────┘  └────────┘

---

Configurable Parameters

General Parameters

Name	Description	Default Value
external	Allows external access to the HTTPCache service from outside the cluster.	false
size	Size of the persistent volume used for caching.	10Gi
storageClass	Storage class used for data.	"replicated" or "local"
haproxy.replicas	Number of HAProxy replicas.	2
nginx.replicas	Number of Nginx replicas.	2

---

Configuration Parameters

Name	Description	Default Value
endpoints	Endpoints configuration.	[]

---

Known Issues

• Upstream response times in the VTS module: The VTS module displays incorrect response times. This issue is documented here: GitHub Issue - VTS Module

---

Additional Resources

To learn more about configuring and using HTTPCache components, here are some useful resources:

• Official Nginx Documentation Comprehensive guide for configuring and optimizing Nginx, including specific modules like cache management.

• Official HAProxy Documentation Everything you need to know to configure HAProxy as a high-performance load balancer.