since I forgot to mention in the original response, this seems to be the issue (other routes removed)

json
{
  "routes": {
    "GET /api/stop/search/near/:latitude]/[longitude": {
      "module": [
        "api/stop/search/near/[latitude]/[longitude].ts:onRequestGet"
      ]
    }
  },
  "baseURL": "/"
}

what is the best way to log what is happening in a function right now?

There is no log feature like that of Workers. Do we just write our own log system to store it somewhere? Would love to store it in R2 but don't have access to that yet 😦

Yeah for now the best way is to pipe off the logs to some external system via

waitUntil

or similar. I believe there will be a better way to get Functions logs in the future, similar to normal Workers, once Functions hits GA.

yes, renaming my firebase /functions directory has prevented my cloudflare workers requests from exceeding 105% of max every day!

asking this here as well as #773219443911819284 in case this is a better spot for it... I'm trying to figure out from various different repos / docs references exactly how I should best go about creating a new Durable Object for a new Pages Functions project / use case. Ideally, I'd like them to live in the same repo and have the least deployment separation / overhead possible (i.e. I'd prefer to deploy new Durable Object code alongside new Pages / Functions code, if possible). I get the loose sense that maybe I need to create the Durable Object in its own folder / deploy it completely independently, but it's not 100% clear to me if that's correct and even if so what the "best way" to manage that is to keep them in sync. Also not clear whether I should be using wrangler@beta or 1.19+ still, etc... anyone here who can advise what is possible and what is best practice ATM? Much appreciated in advance!

Wondering about the same thing, would be cool to have an update! 🙂

Easiest way is to deploy as a separate Worker.

Woops, didn't see the discussion took place in #773219443911819284! Looking forward to the time when we'll be able to create Durable Objects from Pages Functions !

in functions on this page: https://developers.cloudflare.com/pages/platform/functions/ for utilizing path segments is says

When naming your files:
[name] is a placeholder for a single path segment.
[[name]] matches any depth of route below this point.

there are examples provided saying

/api/todos/* => ./functions/api/todos/[id].ts

In turn, the URL segment(s) that match the placeholder will be available under the context.params object using the filename placeholder as the key. Are there any code examples that show how to use that context.params object and how to properly reference the parts / pieces?

Has anyone had the

Only absolute URLs are supported

error when using fetch and gatsby? I can't seem to find how to get the URL of a preview build to fix this error on production (using relative URLs gives this error too).

Copying from #789155108529111069 as it seems relevant for here: Question about setting an environment value based on the subdomain of a Pages preview deployment - I have a Nuxt site hitting CORS issues trying to hit the API endpoints I've set up with the beta Functions feature. Only occuring on the preview builds from PRs. I have the following axios configuration in my nuxt config:

axios: {
    baseURL: 'http://localhost:3000/',
    browserBaseURL: 'http://localhost:3000/',
  },

  publicRuntimeConfig: {
    axios: {
      browserBaseURL: process.env.BROWSER_BASE_URL,
    },
  },

  privateRuntimeConfig: {
    axios: {
      baseURL: process.env.BASE_URL,
    },
  },

I set the env vars to the production URL in the Pages settings. So the API functions work on the production deployment, but all testing from PRs is broken. I don't see env vars (https://developers.cloudflare.com/pages/platform/build-configuration/#environment-variables) available today that would help me resolve this. Is there a way?

Also: I have an

onRequestPost()

function set up and I'm logging the full context I receive. I don't see the POST body parameters anywhere. Shouldn't they be under

context.request.body

? I POST an object containing the fields "page" and "token" and

context.request.body

is the object

{ locked: false }

Any logs I can collect to help identify what is going on? POSTing like this from a Nuxt site:

await $axios.$post('/api/checkToken', payload)

using the provided example, if we have a file at

/functions/api/todos/[id].ts

, if we were to visit the URL

mydomain.dev/api/todos/133

, then the

context.params

object will contain an

id

property that's set to

"133"

. Effectively,

context.params.id = "133"

Any way to get logs from a _worker.js deployed to pages?

Seeing some behaviour that might be totally normal but struggling to explain it when using pages functions. We're using an

_worker.js

to server render a page, and it's stable and consistently fine. But if we smash the refresh button a few times it will

1102 worker resource limits exceed

. It recovers very quickly and then is consistently fine until the you aggressively refresh again. It also doesn't seem to affect other users in the same location, so if one user pushes the worker to

1102

, it seems to be isolated to that user. Is this expected/some kind of protection against users doing that? Or are we likely over/too close to the limits and actually pushing them over? It's difficult to tell whats going on with just the daily usage insights email.

Sounds like you're hitting memory limits. A lot of requests from the same person in a short time will usually end up in the same isolate. Could you share your code?

Not yet, planned to be possible once Functions is out of beta but currently you'll need to send to an external log service like Honeycomb or Sentry.

Any thoughts on why I’m not seeing my post body parameters?

Hi All, I am facing a weird issue. I have written functions in pages to add response header for a route which looks something like this (see configuration below) "/nl/fundraising/:slug": { "middleware": [ "nl/fundraising/[slug]/_middleware.ts:onRequest" ] } Now, I have also written Transform rules which will fallback all request to /nl/index.html so the Deep links will work. The weird issue is that due to the URL rewrite rules the request is not going through the function's middleware. And if i turn off the URL rewrite rules then Deep links will not work. Does any one have solution to this issue? Is this an issue with cloudflare or I am doing something wrong? Please note that, my project is based on angular v12 framework and it follows i18n build as per angular docs.

Tricky to share as the code as it's full react app, but memory limits would kinda makes sense if they hit the same isolate.

Hey folks! I'm running into an issue deploying- I previously deployed with a

functions

dir with no issue, then removed the

functions

dir in a later build. I added the

functions

dir back in my most recent build, but now I'm running into

Failed: an internal error occurred

when deploying. Has anyone run into this before / quick fixes? Thanks so much!

_worker.js file for above

Don't see any

process

or `global`s in there. What's the deployment ID?

@User waiting for comments on this.

You're saying if you enable a Transform Rule rewriting to

/nl/index.html

your Function isn't firing?

Where can I find the deployment ID? I can't find it after doing some poking

It's the UUID in the URL (we shouldn't really add this to the panel...)

When it fails to deploy I don't have a URL to look at 😦

The URL of your page when looking at the deployment