Wrangler just uses esbuild, shouldn't do anything that crazy. Your code you showed is equal to the o

Wrangler just uses esbuild, shouldn't do anything that crazy. Your code you showed is equal to the older version of the package: https://github.com/subhendukundu/worker-auth-providers/blob/7548cda3470072a9a276fd32637b079140728172/dist/providers/google/redirect.js so probably just not updating properly/looking at code from an old push or something else

CChaika Wrangler just uses esbuild, shouldn't do anything that crazy. Your code you show...

Jadugar_Jaggu•10/16/24, 3:06 PM

You can see here in VS CODE, I have the latest one

Jadugar_Jaggu•10/16/24, 3:09 PM

and one more thing this issue arises recently, may be this month, previously it was working fine. Thats why I checked the uploaded code in cloudflare to find some traces

Jadugar_Jaggu•10/16/24, 3:17 PM

Never Mind, its fixed now, after removing and adding the pakage again

Shados•10/16/24, 4:19 PM

Hi everyone. I've searched on the discord and the docs, and there's discussion around worker size limit, but couldn't figure out exactly what "counts" or how to calculate the exact size of my app.

We are on an enterprise plan, and I know right now we're nowhere close to the limit with our prod app, but I want to plan for the future. I know its gzipped size, 10mb limit, but what "counts"? Is it everything thats uploaded to the worker? Do the new static asset support count or not? I upload sourcemaps with the code, do the sourcemaps count?

Any patterns people use to reduce size that I should know from day 1? (eg: obviously large static assets should be hosted outside of the Worker. Anything else in the captain obvious category)?

SShados Hi everyone. I've searched on the discord and the docs, and there's discussion a...

ChaikaOP•10/16/24, 4:21 PM

gzipped size is what matters. Static uploads are separate. Source maps are seperate too and have a pretty strict size limit of their own (I think 1mb or something? I can't find at a glance in docs)

Shados•10/16/24, 4:21 PM

Is there a way in Wrangler or the dashboard to see the size of my worker as it matters to that limit?

ChaikaOP•10/16/24, 4:21 PM

When you upload a worker in a recentish wrangler version it'll throw out your gzipped size

ChaikaOP•10/16/24, 4:23 PM

Total Upload: 10.17 KiB / gzip: 2.57 KiB

Shados•10/16/24, 4:23 PM

Oh yeah, lookie at that!

Shados•10/16/24, 4:23 PM

Excellent, thank you, once again.

Constantin•10/16/24, 4:35 PM

howdy, does CF API can be binded? i mean is it possible to interact with custom domains api in a RPC way?

CChaika > Route pattern must include zone name: example.org [code: 10022] afaik that err...

nikita•10/16/24, 4:37 PM

Thank you, you are right. It looks like it got switched off at some point when our billing failed.

CConstantin howdy, does CF API can be binded? i mean is it possible to interact with custom ...

ChaikaOP•10/16/24, 4:39 PM

not currently, have to go through the normal rest api

123generic•10/16/24, 11:05 PM

Hi all, is there a way to provide "deletion protection" to a worker? One of our developers accidentally deleted a worker we were using in production. It's easy to redeploy, but I'd prefer a better solution.

Constantin•10/17/24, 12:43 AM

hey there how can i pass github env vars to a worker/astro?

when using astro as a Page (not a worker) i can use the locals.runtime.env.CF_PAGES_BRANCH

Constantin•10/17/24, 12:43 AM

however, with workers, there's a github pipeline in place

Constantin•10/17/24, 12:45 AM

i added them in the global scope.. still they are not available not on locals.runtime.env or import.meta.env

Walshy•10/17/24, 12:48 AM

You'd need to inject them into the env in some way

Walshy•10/17/24, 12:48 AM

You could use --var or replace contents in wrangler.toml

Walshy•10/17/24, 12:49 AM

E.g wrangler deploy --var GH_SHA=$GITHUB_SHAwrangler deploy --var GH_SHA=$GITHUB_SHA

Constantin•10/17/24, 1:03 AM

ty @Walshy | Out of office

Constantin•10/17/24, 2:21 AM

does any of you manage to render assets in astro in a different way than setting assets = { directory = "./dist" } in wrangler.toml .. by setting in wrangler you can't use placement smart, plus some other limitations.

Adi•10/17/24, 7:00 AM

So I upgraded to the latest version of wrangler and deployed a website using

wrangler deploy

wrangler deploy

  Total Upload: 3324.79 KiB / gzip: 1022.32 KiB
  Worker Startup Time: 235 ms
  Your worker has access to the following bindings:
  - KV Namespaces:
    - CACHIFIED_CACHE: 191f201005724b038b6facd2ae37e1d9
  - Services:
    - PUBLIC_DATA_SERVICE: 🔴 naamdeo-courses-api#PublicDataService
    - AUTHENTICATED_USER_SERVICE: 🔴 naamdeo-courses-api#AuthenticatedUserService
    - RESTRICTED_SERVICE: 🔴 naamdeo-courses-api#RestrictedService

  Total Upload: 3324.79 KiB / gzip: 1022.32 KiB
  Worker Startup Time: 235 ms
  Your worker has access to the following bindings:
  - KV Namespaces:
    - CACHIFIED_CACHE: 191f201005724b038b6facd2ae37e1d9
  - Services:
    - PUBLIC_DATA_SERVICE: 🔴 naamdeo-courses-api#PublicDataService
    - AUTHENTICATED_USER_SERVICE: 🔴 naamdeo-courses-api#AuthenticatedUserService
    - RESTRICTED_SERVICE: 🔴 naamdeo-courses-api#RestrictedService

This must be a recent change, can I just say, the red dot is kind of scary

I thought something was broken for a second, but I think it's just a choice someone made haha.

Elliot Hesp•10/17/24, 8:13 AM

If I have a worker which proxies a websocket request, am I charged for messages to the proxied endpoint or just the initial request

BBrett Willis Needing some help, is there is no mechanism for handling the event that a client...

Brett Willis•10/17/24, 8:50 AM

[Feature request] The Request.signalRequest.signal API (https://developer.mozilla.org/en-US/docs/Web/API/Request/signal) would be a convenient API to utilise for signalling when the client has disconnected. Currently we have no what of detecting when the request has been cancelled. Particularly important for SSE/streaming scenarios.
Another potential API (new) could be something like ctx.addEventListener('end', () => {})ctx.addEventListener('end', () => {}).

Constantin•10/17/24, 9:10 AM

hey, i'm using astro with cloudflare. currently having develop and main envs. i also created an extra env.local with [env.local.kv_namespaces] aswell. however when running npm run dev --env local .. the KV is not binded, (actually wrangler is not switching/reading from local env)

Original message was deleted

Elliot Hesp•10/17/24, 9:27 AM

Was that in response to my question; so just the initial request?

Elliot Hesp•10/17/24, 9:28 AM

Great thank you

BBrett Willis [Feature request] The `Request.signal` API (https://developer.mozilla.org/en-US/...

kian•10/17/24, 9:59 AM

AbortSignal and AbortController are already a thing in Workers, are you saying they don’t work for this use-case?

Kkian AbortSignal and AbortController are already a thing in Workers, are you saying t...

Brett Willis•10/17/24, 10:24 AM

Yeah I mean we don't have a way of being able to detect when a client has disconnected while the request is in flight.
I think normally the "thread"/context just aborts in-flight? If we use ctx.waitUntil()ctx.waitUntil() then it stays alive but we have no way of knowing when the client has disconnected.
The use case in my mind is when we are generating a perpetual stream (SSE) and need to clean up or do "stuff" when the client disconnects.

Brett Willis•10/17/24, 10:27 AM

Request.signalRequest.signal is an existing API which currently does nothing. The Workers runtime could "abort" that signal when the client disconnects, as a convenient signal if scripts want to know.

kian•10/17/24, 10:32 AM

fwiw it does do something, but it isn’t for the incoming request - it’s for your own fetch requests. Fair suggestion though!

Jürgen B•10/17/24, 1:29 PM

whats going on in the Workers DevTools environment? extremely laggy and weird stuff going on since you updated to instead SEND or SUBMIT button now only shows a "REDO"-icon..the placement of the tooltip isn't that great covering the path field?!

Jürgen B•10/17/24, 1:37 PM

Walshy•10/17/24, 1:42 PM

what kinda lag/buggy stuff you seeing?

Walshy•10/17/24, 1:43 PM

the icon change was because the behaviour has changed, instead of auto-updating the preview you now manually "reload" it

Walshy•10/17/24, 1:44 PM

the tooltip placement, we could move that below - cc: @Vero

(another for Onboarding)

Jürgen B•10/17/24, 1:51 PM

it feels the reload now takes way way longer time than the behavoiur in the past with the auto-updating.. and irrational behaviour if its loading the previous or the actual current code.. it freezes etc

Jürgen B•10/17/24, 1:52 PM

it also seems to auto execute on any changes to the url or headers

Jürgen B•10/17/24, 2:12 PM

so when i hit reload it executes the old script.. then after many seconds its executing the new script..

AAdi So I upgraded to the latest version of wrangler and deployed a website using `wr...

Isaac McFadyen•10/17/24, 2:20 PM

https://github.com/cloudflare/workers-sdk/issues/7009 TY for the feedback, filed issue.

GitHub

FEEDBACK: red or green dot on DO services · Issue #7009 · cloudflar...

In PR #6893, a red or green dot was added beside each service when deploying a DO. I'm not too familiar with Wrangler internals, but it looks like the red vs. green is decided based on whether ...

L3v_•10/17/24, 4:02 PM

Hello,

I am integrating our logs from the Cloudflare WAF into the SIEM but i have a problem which is that in the rawlog we only receive the ruleid and we want to have the rulename in order to map everything into siem rules and for it to be easier for our analysts to look at.

The problem is that I can't find a list (Managed RuleSet and OWASP Ruleset) that contains both the number of the rule names and the rule IDs.

Does anyboday have a suggestion for what I should do?
Thank you!

Flare•10/17/24, 4:05 PM

Please do not post your question in multiple channels/post it multiple times per the rules at ⁠#

welcome-and-rules. It creates confusion for people trying to help you and doesn't get your issue or question solved any faster.

CCodex This was a legendary callout, thank you for this, I completely forgot signals we...

Codex•10/17/24, 9:12 PM

one caveat to this is that it appears that the fetchMock provided by miniflare / vitest-pool-workers doesn't respect this signal when testing / mocking which makes behaviour a bit tricky to test

JJürgen B so when i hit reload it executes the old script.. then after many seconds its ex...

Jürgen B•10/18/24, 7:48 AM

It seems you rolled it back to the original SEND button? But now it doesn't work at all?

JJürgen B It seems you rolled it back to the original SEND button? But now it doesn't work...

Jürgen B•10/18/24, 9:01 AM

seems you're working on it, seeing both SEND and refresh icon.. still it seems when you hit it it first run the one in memory then reloads or re-initiate the one in your code view and executes again.. so you get 2 runs .. confusing as hell

swizzlr•10/18/24, 1:19 PM

Hey all, I’m trying to deploy workers with a custom domain (fully managed by CF) and no origin. The docs call out that I should use custom domains and not routes. However I’d really like to use API gateway, which seems to require routes? I’m currently on the free plan but I intend to upgrade to pro CF and pro Workers when going live. I know gateway is mostly enterprise level, so let me know if I’m barking up the wrong tree here, but I figured gateway would be cheaper for deflecting the vast majority of those annoying inbound attack scans (or maybe I need WAF?). Very new to CF and the ecosystem so any help would be appreciated.

dave•10/18/24, 2:06 PM

Did

wrangler dev

wrangler dev

recently change the way scheduledscheduled needs to be exported?

dave•10/18/24, 2:07 PM

__scheduled__scheduled requests are ending up being sent to my

fetch

fetch

handler

dave•10/18/24, 2:07 PM

oh wait