How does CDN cache invalidation work?

Cache invalidation removes/updates cached content. Methods: TTL-based (automatic expiration), Purge/Invalidation API (immediate removal), Cache tags (group invalidation), Versioned URLs (new URL = new content - most reliable). Challenges: propagation delay across edge nodes, cost per invalidation, eventual consistency. Best practices: use versioned/hashed filenames for static assets, longer TTL + invalidation on publish, cache headers for dynamic content. Balance freshness vs hit rate.