Question 1

How does Neutrino compare to HAProxy for custom routing?

Accepted Answer

Neutrino excels in extensibility, allowing custom L7 rules based on HTTP headers or auth values, while HAProxy is faster but less flexible. For complex, programmable routing, Neutrino is better, but for raw performance, HAProxy is preferred.

Question 2

How to add a custom load balancing algorithm in Neutrino?

Accepted Answer

Neutrino's extensible pipeline lets developers write Scala or Java modules to implement custom algorithms. You need to hook into the pipeline architecture, though specific implementation details may require diving into the codebase and documentation.

Question 3

Does Neutrino support SSL termination?

Accepted Answer

The README doesn't explicitly mention SSL termination, focusing on L7 switching. Given its Netty foundation, it likely can handle SSL, but you may need to implement or configure it separately, unlike some load balancers with built-in support.

Question 4

Can Neutrino be deployed in Kubernetes?

Accepted Answer

Yes, Neutrino supports container deployment including Docker, so it can run in Kubernetes. However, you'll need to manage configuration and JVM resources, as it's not natively optimized for K8s like cloud-native solutions.

Question 5

Is Neutrino suitable for production use outside eBay?

Accepted Answer

While built for eBay's test infrastructure, Neutrino offers robust features like zero traffic loss. However, its limited community and eBay-specific origins might mean fewer production examples and support compared to widely-used tools.

Question 6

What are Neutrino's performance benchmarks vs Nginx?

Accepted Answer

Neutrino's JVM-based architecture may introduce higher latency and memory usage than Nginx, which is C-based and highly optimized. For high-performance needs, Nginx might be better, but Neutrino wins in customizability.

Neutrino

What is Neutrino?

Overview

Use Cases

Best For

Related Projects

Found a gem we're missing?

Not Ideal For

Pros & Cons

Pros

Cons

Open Source Alternative To

Frequently Asked Questions