Load balancing and failover inside Bunny CDN
joe-at-bunny
20 points
2 comments
September 30, 2026
Related Discussions
Found 5 related stories in 129.5ms across 12,728 title embeddings via pgvector HNSW
- Bunny CDN Security Incident SondreDag · 212 pts · October 30, 2026 · 63% similar
- Scriptable GeoDNS: Leaving AWS for Bunny sshah03 · 58 pts · November 06, 2026 · 53% similar
- HTTP/3 is faster with some work and help from Cloudflare saikatsg · 13 pts · October 31, 2026 · 43% similar
- AWS Lambda Web Functions squiwardStan5 · 13 pts · October 12, 2026 · 41% similar
- AWS Lambda Web Functions HatchedLake721 · 112 pts · October 13, 2026 · 41% similar
Discussion Highlights (1 comments)
joe-at-bunny
Hi HN, I’m Joe, and I lead content delivery and security at bunny.net. We’ve just added a load balancer to Bunny CDN in public preview. Previously, a Pull Zone pointed at a single origin. Customers who needed multiple origins had to handle balancing separately. Now they can configure a pool of origins and have the CDN distribute traffic and handle failover. For example, you can have two regions sharing traffic with a third on standby. The edge builds the fallback order before contacting the first origin, so those priorities still apply if it needs to try another. Retries take request safety into account, and the failover chain has a time limit. It’s HTTP(S) origin balancing within the CDN. Backends can also be Bunny Storage, Edge Scripts, or Magic Containers, so a static copy of your site could serve as the standby. I’m around to answer questions about the implementation. I’d be interested in how you’re handling multiple origins today, especially across different providers.