I have an npm dependency in my worker that has a dependency on the "url" lib/module that is built into the nodejs runtime. Excerpt:

const url_1 = require("url");
...
...
...
const url = new url_1.URL(input);//throws undefined

When I run wrangler dev I receive an undefined error I am guessing because the worker env doesn't know how to resolve the require("url") which is built into the nodejs runtime environment. Is there a way to bundle some kind of replacement for nodejs's url lib/module in order to get the require("url") to resolve correctly inside of the worker runtime. I was looking at the Configuration > Bundling documentation in the Wrangler docs but no lightbulbs have gone off on how to leverage it to provide a solution. Any experience anyone can share would be greatly appreciated!

I would like to run a script on my local machine which would hook into the context of my worker which is already configured on Cloudflare and then perform any operation in that context, e.g. using D1, KV directly without having to actually run a worker with a fetch handler and then do a an HTTP request to achieve essentially the same. You could think about it as simply running the

node

command with a file in your project which has access to the whole context of the project (e.g. databases, env variables etc). A similar approach from other language/frameworks are Django management commands. The way I imagine this is simply running

wrangler dev

and instead of launching a worker service (which essentially runs as an HTTP service) it would execute a worker "script" which runs once and then exits. Is this possible in workers and if so how?

Hi everyone. I use Remix, CF Workers, and PlanetScale. When I connect a sample db it works. However, when I connect the original db, the error appears: "No fetch handler defined and no upstream to proxy to specified. Make sure you're calling addEventListener("fetch", ...)." How to fix this?

New project, ran

pnpm create cloudflare workers worker-rust

and I'm trying to run

pnpm wrangler dev

but I get an error

Running custom build: cargo install -q worker-build && worker-build --release
error[E0658]: use of unstable library feature 'bool_to_option'

I'll add my toml to a separate post

Hi all - I'm hitting the analytics api graphql endpoint (https://api.cloudflare.com/client/v4/graphql/) to get metric information on my workers scripts which are running. I haven't yet found a way to get information about the error rates, specifically number of 4XX vs 5XX (and 3XXs would be ideal, too). Is this possible with this api? If not, are there any api calls which I can use which can get me this information? Cheers

using the workers playground testing site to just make a simple http request to a GQL API, but the request just stays stuck on pending, works with other websites, but this call just hangs and i have no idea why.

HI, I've setup 2 worker according to docs, But it's not working. plz check this https://github.com/AggressivelyMeows/Horseman/discussions/8

Hi, first of all thank you for the great products , pages is just amazing ! Ive been trying to figure out how to enable logs on pages for some time and can't figure it. I'm using svelte kit with cloudflare adapter, and deploying onto cloudflare pages. I would like to store logs somehwre and I have read about logpush. Ive created a logpush job and connected it to R2, how do I connect what I deployed on pages to that job ? Is there another way to store the logs for later viewing for pages ?

I keep getting 1006 errors in my websocket connections on the frontend. what is a good way to resolve this? Extending timeout?

i want to make a redirection for a user if he is in a specific country, and the redirection is being triggered if i enter from the cloudflare generated url for testing in the worker but if i enter the website directly the worker is not being triggered, please any help would be appreciated

Howdy! What is the best way for a worker script to know its current origin? The use-case is a router library that can match request objects. In a service worker environment, that worker may intercept multiple origins, so SW routers are "origin aware." I want to re-use that library in a CF Worker context. The

WorkerGlobalScope

has a

location

property that seems unavailable (at least on local workers.). Thanks for your help!

As workers is using esbuild to compile typesript code but its skips experimentalDecorators functionalities.Is there any solution for this?

Hi, I noticed that index page (

/

) requests that contain query parameters (

?anything

) are served by the origin instead of one of workers. I'm not sure why this happens. I would like all requests to

/

to be served by CF. I've set up the example to illustrate this:

$ curl https://test.0x0f.ru/
served from a worker

$ curl https://test.0x0f.ru/?a=b
served by the origin

I have recently setup a Cloudflare worker with a custom domain. I am trying to redirect users to specific domains based on the current path (domain/code),then by searching for that code in Upstash Redis. The issue I am facing is that even after updating the Redis key’s value, I keep getting redirected to the previously set value. I believe this is due to caching, but I am unable to figure out how to disable or bypass the caching. My worker function is: export default { async fetch(request, env) { return await handleRequest(request) } } async function handleRequest(request) { const url = new URL(request.url) const path = url.pathname const code = path.substring(1) const redirect_url = await get_url(code) return Response.redirect(redirect_url, 301) } async function get_url(code){ const response = await fetch(

https://upstash_redis.url

, { headers: { Authorization: "Bearer Token" } }) const data = await response.json() console.log(data) if(data.result != null ){ return data.result } else { return "https://www.example_deafult.com" } }

I'd like to build a simple "fan out" service and wonder if this is conceptually possible using Worker's support for WebSockets. I'd like the service to have an HTTP endpoint that an external system can post to when an event is created and then the worker would relay this event to all connected clients via WebSockets. Is such a use-case feasible using workers? I've read most of the docs on Workers integration with WebSockets but it's still unclear to me how all the pieces fit together. Would I need a durable object for this use case? Thanks in advance for any and all help!

Im trying to get the MTA-STS worker to work from https://365labs.cloud/blog/mta-sts-using-cloudflare-workers but whenever I follow the steps my worker will show; xxxxx.workers.dev is not defined in the mta-sts worker And the URL for my own domain https://mta-sts.xxx.well-known/mta-sts.txt Will show nothing, does anyone know how to solve this ?

Hi guys, I have this data in my statistic in workers page: Unbound requests 9.2k Billable duration 277 GB-sec Bundled requests 221.18M Durable Objects storage 0 B However, In my billing panel, I see Cloudflare Workers request: Cumulative total this month: 11k requests What does it mean? Why 221.18M of bundled requests doesn’t count in the billing panel?

Is it possible to extract the original headers from a worker fetch request to an end-point which returns a 302? e.g. const res = await fetch("http://localhost:3000/", { "redirect": "manual" }) res = {"body":{},"bodyUsed":false,"headers":{},"url":"http://localhost:3000/","redirected":false,"ok":false,"statusText":"Found","status":302} If I run the same request via curl I get the original headers (e.g. location + set-cookie): curl -v http://localhost:3000/ < HTTP/2 302 < cache-control: private < location: https://www.google.com < set-cookie: SIGNIN.REQUEST=1683699853; path=/; secure; HttpOnly < access-control-expose-headers: Request-Context < date: Tue, 09 May 2023 14:49:59 GMT < content-length: 0 Is it possible to extract the original headers using Cloudflare workers (or any other Cloudflare service)? The main problem is that the set-cookie header is lost.

Hey, I'm tryna make a file uploading system. My full code is https://pastebin.com/dUCApxp4, but this should be the important part:```ts export default { async fetch(request, env) { const url = new URL(request.url); const key = url.pathname.slice(1); const { pathname } = new URL(request.url); if(pathname == "/") { return new Response("CDN is working! :D") } switch (request.method) { case 'PUT': if(request.headers["Authorization"] != "nop!") return "no, go away" await env.data.put(key, request.body, { customMetadata: Record({"uploader":request.headers["Author"]}) }); return "helo mate" ...``` However, when I sent a

PUT

request with the valid headers and data, it returns 500, and the logs show

Uncaught (in promise) TypeError: Incorrect type for Promise: the Promise did not resolve to 'Response'.

. I've been trying to fix this for a very long time, so any help would be appreciated!

Has anyone had trouble with differences in how WebSocket clients work when running locally with `wrangler dev --local `vs plain

wrangler dev

or when deployed? I am building a WebSocket client for Discord's Gateway API using DOs. Everything works when running locally with

--local

, but when I remove

--local

or when I deploy to CF, my WebSocket client (created via

new WebSocket()

) immediately gets an error event that says

Failed to establish the WebSocket connection: expected server to reply with HTTP status code 101 (switching protocols), but received 401 instead.

. Discord's Gateway WS protocol doesn't require any authorization to open the WS connection normally (a token gets passed in a message later), and this exact same code works with

wrangler dev --local

. Could there be some difference in how

new WebSocket()

when deployed to Cloudflare vs when running locally?

Hello, I'm trying to create a worker that will recieve forwarded emails, extract some info from them, and contact my api using that info. I already have the code written to do that, however I am confused on how to get cloudflare to accept those emails and point them to the worker. The documentation for email workers seems to be essentially non-existent. If anyone has any link to something I just couldn't find, please send that. The "Using email triggers" button on the dashboard just goes to a page that talk about normal http routes (). "mail" is not mentioned a single time on that page. How can I configure it? Do I just put an email in?

I have a cloudflare worker i would like to use to fetch from a cloudflare page that is behind cloudflare access/zero trust ACCESS->WORKER->ACCESS->PAGE I need the worker to be able to authenticate with cloudlflare access. I have tried using client id and secret, but because the workers are short lived they must constantly reauthenticate which is very slow and breaks some requests. Is there any other way for me to avoid this authentication redirect behaviour? PS: I need this because I'm using a worker to route traffic from a subroute of our main domain to a cloudflare page. Essentially we're serving many cloudflare pages under the same domain under subroutes

throw new Error('Multipart: Boundary not found');
            ^
Unhandled Promise Rejection: Error: Multipart: Boundary not found

Was missing this error

I am following these instructions but it is not working: https://developers.cloudflare.com/workers/platform/bindings/about-service-bindings/#wranglers

1. Created a npx wrangler init . 2. Then "npm install pg" and added the query code into index.js 3. then, npx wrangler dev, got compatibility errors, and got suggestions to add "node_compat = true", and I did, those errors disappeared 4. Then again ran "npx wrangler dev", got error ' **could not resolve "pg-native**" '. And this is where I'm struck. I ran same code in normal js and it worked, but I don't know why is it asking "pg-native", and also I tried to install it locally and globally, and it won't get installed. 5. What to do?

Where can I find TypeScript type definitions of

@cloudflare/workers-types

pkg in Github?

return new ReadableStream({ start(controller) { } , cancel() { } }) Cancel is not called however its working on nodejs , deno and browser .Why workers are so inconsistent with others.

I have a lot of JSON data that I need to transform to a common format. Is there anything like HTMLRewriter, but for JSON? The JSON payload is getting too big to be stored in mem in my worker

I've set up the pay-as-you-go plan, updated the

workers-chat-demo

account_id, and run

wrangler login

. Running

wrangler publish

gives me an error: > In order to use Durable Objects, you must agree to pricing at > https://dash.cloudflare.com/aac1a529c1da5d1cbd4c1c1cb12f9675/workers/overview?enable-durable-objects This link sends me to the workers overview page, with no CTA to enable durable objects. Going to the docs, https://developers.cloudflare.com/workers/runtime-apis/durable-objects/#durable-objects > ... Navigate to “Workers” and opt in to pricing. I expected that I should be able to set up my pay-as-you-go plan and re-login via wrangler but it's not working. What can I try next?

I have two works, A & B, A binds to B. When A does the

env.BINDING.fetch(...)

call it leads to a 403 response. That response is NOT coming from our code. Why is CF doing this? The body response is: error code: 1003