Cloudflare enhances Rust Workers reliability

Cloudflare has improved the reliability of its Rust Workers by implementing advanced panic and abort recovery mechanisms within wasm-bindgen. Previously, Rust panics or unexpected aborts could leave a WebAssembly instance in an unstable state, potentially causing failures for other requests or even rendering the Worker temporarily unusable. The new system, leveraging WebAssembly Exception Handling and panic=unwind support, ensures that individual request failures do not poison the entire Worker instance. This allows for stateful workloads to recover without losing data, significantly enhancing the robustness and predictability of Rust-based applications on the Cloudflare Workers platform.