SQC-707 Fix Hyperdrive binding slow latency on Windows

[Ltadrian]

Fix Hyperdrive binding slow latency on Windows for npx wrangler dev command

• Update localhost to 127.0.0.1 address to fix issue with Windows defaulting to ipv6 lookup and timing out
• Fix skipped Windows Hyperdrive tests which were failing due to miniflare dispose() causing connection reset errors. It was causing the child process to crash but this is now fixed with calling node spawn and cleanly handling those specific connection errors by error code

Fixes #SQC-707

Customer reached out on discord reporting slow latency of 2000ms when using Hyperdrive binding and localConnectionString pointing at local db. Was able to reproduce on Windows 10 PC and get latency back to 8ms-15ms

---

• Tests
  • Tests included
  • Tests not necessary because:
• Public documentation
  • Cloudflare docs PR(s):
  • Documentation not necessary because: Bug fix
• Wrangler V3 Backport
  • Wrangler PR:
  • Not necessary because: miniflare change

[changeset-bot]

🦋 Changeset detected

Latest commit: 529988a

The changes in this PR will be included in the next version bump.

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[pkg-pr-new]

create-cloudflare

npm i https://pkg.pr.new/create-cloudflare@11443

@cloudflare/kv-asset-handler

npm i https://pkg.pr.new/@cloudflare/kv-asset-handler@11443

miniflare

npm i https://pkg.pr.new/miniflare@11443

@cloudflare/pages-shared

npm i https://pkg.pr.new/@cloudflare/pages-shared@11443

@cloudflare/unenv-preset

npm i https://pkg.pr.new/@cloudflare/unenv-preset@11443

@cloudflare/vite-plugin

npm i https://pkg.pr.new/@cloudflare/vite-plugin@11443

@cloudflare/vitest-pool-workers

npm i https://pkg.pr.new/@cloudflare/vitest-pool-workers@11443

@cloudflare/workers-editor-shared

npm i https://pkg.pr.new/@cloudflare/workers-editor-shared@11443

@cloudflare/workers-utils

npm i https://pkg.pr.new/@cloudflare/workers-utils@11443

wrangler

npm i https://pkg.pr.new/wrangler@11443

commit: 529988a

[jamesopstad]

• Update localhost to 127.0.0.1 address to fix issue with Windows defaulting to ipv6 lookup and timing out

It looks like this is the only implementation change in the PR. Just so I understand, can you explain how this improves the latency? If it was previously timing out wouldn't that be an error rather than slow?

[Ltadrian]

• Update localhost to 127.0.0.1 address to fix issue with Windows defaulting to ipv6 lookup and timing out

It looks like this is the only implementation change in the PR. Just so I understand, can you explain how this improves the latency? If it was previously timing out wouldn't that be an error rather than slow?

Never actually saw a timeout error when running npx wrangler dev. I think Windows would timeout during DNS localhost lookup expecting a response from IPV6 and fallback to IPV4 which would then actually succeed. The end result was slower latency instead of a timeout error from the worker during the connect() step

[jamesopstad]

• Update localhost to 127.0.0.1 address to fix issue with Windows defaulting to ipv6 lookup and timing out

It looks like this is the only implementation change in the PR. Just so I understand, can you explain how this improves the latency? If it was previously timing out wouldn't that be an error rather than slow?

Never actually saw a timeout error when running npx wrangler dev. I think Windows would timeout during DNS localhost lookup expecting a response from IPV6 and fallback to IPV4 which would then actually succeed. The end result was slower latency instead of a timeout error from the worker during the connect() step

I see. Thanks, that makes sense.

[jamesopstad]

Thanks. I've merged a PR to increase the test timeout so please bring this up to date with main before rerunning the tests.