[BUG] memory-queue response always STALE

[minibox24]

Describe the bug

In a wrangler dev --remote or Workers deployment production environment, memory-queue does not work properly and keeps showing as STALE.

Steps to reproduce

https://github.com/minibox24/opennext-cloudflare-isr

https://opennext-cloudflare-isr.minibox.workers.dev/

Expected behavior

ISR every 10 seconds are working normally

@opennextjs/cloudflare version

0.5.8

Wrangler version

3.109.2

next info output

(wsl)

Operating System:
  Platform: linux
  Arch: x64
  Version: #1 SMP Tue Nov 5 00:21:55 UTC 2024
  Available memory (MB): 15818
  Available CPU cores: 12
Binaries:
  Node: 20.18.0
  npm: 10.8.2
  Yarn: 1.22.22
  pnpm: N/A
Relevant Packages:
  next: 14.2.24 // An outdated version detected (latest is 15.2.1), upgrade is highly recommended!
  eslint-config-next: 14.2.24
  react: 18.3.1
  react-dom: 18.3.1
  typescript: 5.8.2
Next.js Config:
  output: N/A

Additional context

fixed this issue with a nextjs patch
https://gist.github.com/minibox24/b2dc210086d4ff8d1b1f9984835354a8

In direct mode, it works fine in production.

[conico974]

For reference : https://discord.com/channels/1283128968140161065/1346540443495825418
I'm wondering if it might be an issue with clearTimeout used in the Memory Queue

[vicb]

@james-elicx could you please take a look?

[james-elicx]

Sorry, this one slipped through my notifications. The problem is that we're invoking the worker in a way that actual deployments don't support, so it's unable to actually fire the request to trigger the revalidation, so we need to switch to a service binding for it instead.