flaky / intermittent `Response.clone: Body has already been consumed` errors

[thescientist13]

What happened?

As observed most impactfully in #1457 (comment) was this issue of getting this error in development (and even as recent during a greenwood build during v0.33.0-alpha.8 testing)

****************************************************************************
TypeError: Response.clone: Body has already been consumed.
    at webidl.errors.exception (node:internal/deps/undici/undici:3610:14)
    at _Response.clone (node:internal/deps/undici/undici:9192:31)
    at file:///C:/Users/owenbuckley/Workspace/greenwood/packages/cli/src/lifecycles/serve.js:70:45
    at async file:///C:/Users/owenbuckley/Workspace/greenwood/packages/cli/src/lifecycles/serve.js:49:5

It's flaky in the sense that if you ran it five times again, it will probably just work.

Steps to reproduce

It happens intermittently but was definitely consistent in #1457 but as mentioned, still seems to flake out every once in a while

Environment

Any recent versions of Greenwood across all OSs.

Additional Context

There was even #1565 trying to fix it that was a bit of an exercise. In general, I think we may need to rethink how we do response cloning in lifecycle/develop.js as well as other areas of the project

• https://github.com/ProjectEvergreen/greenwood/blob/master/packages/cli/src/lifecycles/bundle.js#L21
• https://github.com/ProjectEvergreen/greenwood/blob/master/packages/cli/src/loader.js#L22
• https://github.com/ProjectEvergreen/greenwood/blob/master/packages/cli/src/lifecycles/prerender.js#L33

---

Will tag this in #970

[thescientist13]

Seeing this issue more prominently now when trying to integrate the monaco editor in this project - ProjectEvergreen/playground.wcc.dev#9

I wonder if we are having an issue with unconsumed response bodies when cloning? 🤔
https://blog.cloudflare.com/a-better-web-streams-api/#exhausting-resources-with-unconsumed-bodies

The problem is compounded by APIs that implicitly create stream branches. Request.clone() and Response.clone() perform implicit tee() operations on the body stream – a detail that's easy to miss. Code that clones a request for logging or retry logic may unknowingly create branched streams that need independent consumption, multiplying the resource management burden.

Also, from the MDN docs
https://developer.mozilla.org/en-US/docs/Web/API/Response/clone

If only one cloned branch is consumed, then the entire body will be buffered in memory. Therefore, clone() is one way to read a response twice in sequence, but you should not use it to read very large bodies in parallel at different speeds.