cloudnative wiki

cdn

3 min read

CDN

Content Delivery Networks (CDNs) help deliver web content quickly by caching it at various points globally. There are two primary types of CDNs: push-based and pull-based. Here’s a detailed explanation and examples of each type:

Pull-Based CDNs

How They Work:

  1. Initial Request: When a user requests content, the CDN checks if the content is already cached.
  2. Cache Miss: If the content is not cached (cache miss), the CDN fetches it from the origin server.
  3. Caching: The content is then cached at the CDN edge server.
  4. Subsequent Requests: Future requests for the same content are served from the cache until the cache expires.

Pros:

  • Automatic Updates: Content is fetched automatically when requested.
  • Less Management: Easier to manage as content is only cached when needed.
  • Cost-Effective: Efficient for large amounts of content with varying request patterns.

Cons:

  • Initial Latency: First request might be slower due to cache miss.
  • Cache Expiration: Must manage cache expiration and invalidation.

Examples:

  • Amazon CloudFront
  • Akamai
  • Google Cloud CDN

Push-Based CDNs

How They Work:

  1. Content Upload: Content is manually pushed to the CDN edge servers by the origin server or another mechanism.
  2. Caching: The content is stored on the CDN edge servers.
  3. User Request: When users request the content, it’s served directly from the CDN edge servers.

Pros:

  • Immediate Availability: Content is available immediately on edge servers.
  • Control: More control over what is cached and when it is updated.

Cons:

  • Management Overhead: Requires manual management and updates.
  • Storage Costs: Higher storage costs since content is pushed regardless of demand.

Examples:

  • MaxCDN (now StackPath)
  • Rackspace CDN
  • KeyCDN

Detailed Differences:

  1. Content Deployment:
    • Pull-Based: Content is deployed automatically on demand.
    • Push-Based: Content must be manually pushed to edge servers.
  2. Cache Miss Handling:
    • Pull-Based: Cache miss results in fetching content from the origin.
    • Push-Based: Content is preloaded, so there are no cache misses.
  3. Management:
    • Pull-Based: Less management required; CDN handles fetching and caching.
    • Push-Based: Requires manual updates and management.
  4. Initial Latency:
    • Pull-Based: Initial request might be slower due to fetching from origin.
    • Push-Based: No initial latency as content is preloaded.
  5. Use Cases:
    • Pull-Based: Suitable for dynamic or frequently updated content.
    • Push-Based: Ideal for static content that doesn’t change often.

Use Cases in Detail:

  • Pull-Based CDN Use Case:
    • E-commerce Websites: Dynamic content such as product pages, where the content changes frequently based on inventory and user interactions.
    • News Websites: Articles and media that are updated frequently.
  • Push-Based CDN Use Case:
    • Software Distribution: Large static files like software updates, which are less likely to change often.
    • Media Streaming: Pre-recorded videos or music files that need to be available immediately without latency.

Understanding the differences between pull-based and push-based CDNs helps in selecting the appropriate solution based on the specific needs of content delivery, frequency of content updates, and management preferences.

Graph View