Cloudflare Developers•8mo ago

Interesting issue -- Log stream causes NextJS Edge runtime app to fail

This is kind of an intriguing one..... 1. App deploys fine via wrangler pages deploy .vercel/output/static 2. I can hit the side all is dandy... 3. Go to the deployment UI, go to the "Real-time logs" section, click "Begin Log Stream" 4. Log flows begin to appear -- but now every request is an HTTP 500 5. Even after ending the log stream, the worker appears to be in a broken state and continues to throw HTTP 500s 6. Wrangler tail does not show any useful error messages:

❯ pnpm wrangler pages deployment tail
 ⛅️ wrangler 3.48.0
-------------------
Select a project:
❯ nameplace
  nameplace-trpc
No deployment specified. Using latest deployment for production environment.
Connected to deployment c1ae8c2d-324c-452c-a680-e2aa14b63721, waiting for logs...
GET https://c1ae8c2d.nameplace.pages.dev/places - Ok @ 4/9/2024, 6:04:20 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/favicon.ico - Ok @ 4/9/2024, 6:04:20 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/places - Ok @ 4/9/2024, 6:04:23 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/favicon.ico - Ok @ 4/9/2024, 6:04:24 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/places - Ok @ 4/9/2024, 6:04:24 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/favicon.ico - Ok @ 4/9/2024, 6:04:24 PM
  (error) Error: internal error

❯ pnpm wrangler pages deployment tail
 ⛅️ wrangler 3.48.0
-------------------
Select a project:
❯ nameplace
  nameplace-trpc
No deployment specified. Using latest deployment for production environment.
Connected to deployment c1ae8c2d-324c-452c-a680-e2aa14b63721, waiting for logs...
GET https://c1ae8c2d.nameplace.pages.dev/places - Ok @ 4/9/2024, 6:04:20 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/favicon.ico - Ok @ 4/9/2024, 6:04:20 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/places - Ok @ 4/9/2024, 6:04:23 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/favicon.ico - Ok @ 4/9/2024, 6:04:24 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/places - Ok @ 4/9/2024, 6:04:24 PM
  (error) Error: internal error
GET https://c1ae8c2d.nameplace.pages.dev/favicon.ico - Ok @ 4/9/2024, 6:04:24 PM
  (error) Error: internal error

The annoying bit is that without any code change a re-deployment ends up fixing the issue -- but I then have to wait for a full build/upload cycle (which is not that bad because of caching but still, my app is broken until I do this) dev: https://nameplace.pages.dev/ deployment id: 0877b232-f464-44fb-bd88-fa620eb21341 account id: f1087e918dcedfd9575e5e595a817813

4 Replies

Chaika•8mo ago

Do you have smart placement on? If so, sadly known issue we're hoping they address soon

jacobdrOP•8mo ago

I do have smart placement on, Ill turn that off now. Just saw that being discussed on another thread in a quite different context Thanks for the quick response I need to re-deploy from the CLI to have that take effect, right?

Chaika•8mo ago

Yea changing smart placement would only take effect on the next deployment (cli, git, etc)

jacobdrOP•8mo ago

Thanks a lot ill mark this as solved Ok that definitely resolved my issue

Gaming

Programming

Interesting issue -- Log stream causes NextJS Edge runtime app to fail