Conversation
…ing duri…" This reverts commit 7d81a83.
🦋 Changeset detectedLatest commit: e750b57 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 |
|
Codeowners approval required for this PR:
Show detailed file reviewers |
|
I've now thoroughly reviewed all the changes in this PR. The code is well-structured, the proactive refresh logic is correct, the timer cleanup is properly handled in all lifecycle events (bundleStart, teardown), the error handling changes are sound, and the lazy auth resolution is a good improvement. The tests are comprehensive and cover the key scenarios (proactive refresh, timer cancellation on bundle start, timer cancellation on teardown). LGTM |
|
✅ All changesets look good |
create-cloudflare
@cloudflare/kv-asset-handler
miniflare
@cloudflare/pages-shared
@cloudflare/unenv-preset
@cloudflare/vite-plugin
@cloudflare/vitest-pool-workers
@cloudflare/workers-editor-shared
wrangler
commit: |
|
/bonk can you fix the E2E test that is failing on this pr? |
The PR changed LocalRuntimeController to pass the auth hook lazily to maybeStartOrUpdateRemoteProxySession. This caused the user's auth callback to be called twice during initial startup: once during the nested startWorker's resolveDevConfig zone validation (line 121), and again in RemoteRuntimeController.#onBundleComplete when unwrapping config.dev.auth. Revert to eagerly resolving auth in LocalRuntimeController before passing it, so the resolved CfAccount object flows through without triggering additional auth calls in the nested startWorker chain.
E2E Test FixThe failing E2E test Root CauseThe PR changed This caused the user's auth callback to be invoked twice during initial startup of the remote proxy session:
Previously, when the resolved FixReverted the Other Failing Tests
|
|
The fix has been committed and pushed. Here's a summary: Problem: The E2E test Root cause: The PR changed Fix: Restored eager auth resolution in |
|
/bonk that's absolutely not the right fix. eagerly resolving auth is a bug, not a solution. resolving auth twice is fine, since the second time it'll be cached and the lookup will be fast. update the e2e test assertion instead |
|
UnknownError: ProviderInitError |
|
@penalosa Bonk workflow failed. Check the logs for details. View workflow run · To retry, trigger Bonk again. |
|
/bonk retry the previous task |
…ion" This reverts commit 587471f.
The lazy auth passing causes auth to be resolved twice during startup (once in resolveDevConfig for zone validation, once in RemoteRuntimeController#onBundleComplete). The second call is cached and fast, so this is expected behavior. Update assertions from toHaveBeenCalledOnce() to toHaveBeenCalled().
|
Done. Here's what I did:
|
workers-devprod
left a comment
There was a problem hiding this comment.
Codeowners reviews satisfied
Reverts #13464