fix: deployed worker unable to invoke itself in memory queue

[james-elicx]

Service bindings are required for fetching the same worker in production.

A bit of a shame that our local e2es were unable to catch this as wrangler doesn't fail when doing fetches in local mode.

Docs PR - opennextjs/docs#89

[changeset-bot]

🦋 Changeset detected

Latest commit: b7f7b12

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

This PR includes changesets to release 1 package

Name	Type
@opennextjs/cloudflare	Patch

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]

Open in Stackblitz

pnpm add https://pkg.pr.new/@opennextjs/cloudflare@445

commit: b7f7b12

[conico974]

That's really weird, in the issue they say that it works fine with direct in production, which should do the exact same thing

[vicb]

Thanks for the fix @james-elicx, LGTM.

Maybe add a little more details to the changeset for the changelog

[vicb]

That's really weird, in the issue they say that it works fine with direct in production, which should do the exact same thing

Any chance that direct goes through the path @dario-piotrowicz modified (removing an internal header).
I might be off, still trying to catch up with everything...

[james-elicx]

That's really weird, in the issue they say that it works fine with direct in production, which should do the exact same thing

Just tried the direct queue and that doesn't work with their repro when deployed either for me.