Is there any trick(s) I can do to prevent a 100 second timeout on a Worker subrequest? I do not cont

Is there any trick(s) I can do to prevent a 100 second timeout on a Worker subrequest? I do not control the endpoint that I'm fetching from sadly..

Ddave Is there any trick(s) I can do to prevent a 100 second timeout on a Worker subre...

James•5/5/25, 2:52 PM

You could a signalsignal to your outgoing

fetch

fetch

with signal: AbortSignal.timeout(5000)signal: AbortSignal.timeout(5000) or whatever and then catch the abort signal timeout?

JJames You could a `signal` to your outgoing `fetch` with `signal: AbortSignal.timeout(...

daveOP•5/5/25, 2:54 PM

ah, I was more hoping to keep the connection alive somehow.

James•5/5/25, 2:54 PM

Ah, not really then

daveOP•5/5/25, 2:54 PM

I find it weird subrequests don't obey my zone's timeout.

James•5/5/25, 2:54 PM

If you do it inside a DO, it might not timeout? I’ve not tested in a while

Shrey•5/5/25, 2:55 PM

Hey, we're launching a white-label enabled app and urgently need Custom Hostnames enabled for client domain mapping. Our clients are setting up CNAMEs now, so speed is key. Does any know where i can enable it.

PS- i am already on enterprise plan

Chaika•5/5/25, 3:03 PM

?crossposting -> Custom hostname

SuperHelpflare•5/5/25, 3:03 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.

cj•5/5/25, 3:39 PM

I have a workers app built with Hono. I've defined a few routes - one uses wasm, while the others are pretty lightweight and just talk to DO or D1. Will the wasm make this worker slow to init?
ref: binary-size of wasm in worker

cj•5/5/25, 3:40 PM

Should I be having multiple workers in this situation? (But how do I share the DO?)

soos•5/5/25, 7:02 PM

Could someone help me in #Help with HTML Rewriter please

kaelys.js•5/6/25, 1:50 AM

Is it possible to enable an access policy on both a pages.dev subdomain project as well as a custom domain tied to it?

oscar•5/6/25, 9:17 AM

hey! what's the best channel to ask questions regarding a bug with Cloudflare Images?

Kkaelys.js Is it possible to enable an access policy on both a pages.dev subdomain project ...

Hard@Work•5/6/25, 4:49 PM

No. Access Policies on the

pages.dev

pages.dev

usually break the Custom Domain(unless you just entirely bypass Access, which basically means it does nothing)

HHard@Work No. Access Policies on the `pages.dev` usually break the Custom Domain(unless yo...

daveOP•5/6/25, 4:53 PM

@Cole It depends, are you trying to only allow some people access, or do you just want to block access? The latter is easier.

daveOP•5/6/25, 4:54 PM

This rule works if you want to prevent all pages.dev access from the interwebs.

kaelys.js•5/6/25, 4:58 PM

@dave

Thanks for the reply! Want to protect the custom domain with access and disable/block direct access to the pages.dev domain

kaelys.js•5/6/25, 4:58 PM

But only for that specific pages project

Kkaelys.js @dave Thanks for the reply! Want to protect the custom domain with access and ...

daveOP•5/6/25, 5:00 PM

Probably adding the subdomain instead of just pages.dev like my screenshot might work

predaytor•5/6/25, 5:38 PM

what is the actual cpu limit for workers free plan? I have strange time guessing why some requests works even above ~250ms, while others fail on 10-20ms

olafg•5/6/25, 5:38 PM

When using auxillaryWorkers in the Vite Plugin to point to a worker that is bound with a service binding, does it not emulate the bound resources (d1) in the bound worker's directory? I.e.

Worker A (frontend) has a service binding to Worker B (Hono API).
Worker B has a d1 bound to it.

Hakemo•5/6/25, 5:39 PM

yo i want to build a static website with hono api for few parts of the website

will each request be counted toward the request limit?

currently in cloudflare pages it doesn't unless i hit the api

will the same thing work for workers?

Ppredaytor what is the actual cpu limit for workers free plan? I have strange time guessing...

Peps•5/6/25, 5:42 PM

?cpu-time-overage

SuperHelpflare•5/6/25, 5:42 PM

See this message from the Workers lead on why you can sometimes consume more CPU time than your limit: https://canary.discord.com/channels/595317990191398933/1204401951026974780/1204805369658867752

PPeps ?cpu-time-overage

predaytor•5/6/25, 5:44 PM

thx! As I understand correctly, 10ms of CPU time on a Free tier, right?

Ppredaytor thx! As I understand correctly, 10ms of CPU time on a Free tier, right?

Peps•5/6/25, 5:45 PM

correct

daveOP•5/6/25, 7:13 PM

so I'm having a pretty big unexpected issue with the cache API

daveOP•5/6/25, 7:21 PM

actually maybe not.. I am confused. One sec lol

daveOP•5/6/25, 7:33 PM

ahhh yep, it was a bad cache config

rufi•5/6/25, 8:06 PM

For gradual deployments, there is a way to enforce a version with a Cloudflare-Workers-Version-Key header.
But how does that work with service bindings (can we pass a "header")? does it supports gradual deployment?

daveOP•5/7/25, 12:16 AM

Is there a way to tell which region my email Worker is running in?

Ddave Is there a way to tell which region my email Worker is running in?

Hello, I’m Allie!•5/7/25, 12:16 AM

Hit /cdn-cgi/trace?

HHello, I’m Allie!Hit /cdn-cgi/trace?

daveOP•5/7/25, 12:16 AM

from an email Worker?

daveOP•5/7/25, 12:17 AM

oh I see what you mean, like to another site?

Ddave from an email Worker?

Hello, I’m Allie!•5/7/25, 12:17 AM

Yeah

Ddave oh I see what you mean, like to another site?

Hello, I’m Allie!•5/7/25, 12:17 AM

Yeah, I usually use Quad1

daveOP•5/7/25, 12:17 AM

hmm, was hoping to avoid opening another connection

HHello, I’m Allie!Yeah, I usually use Quad1

daveOP•5/7/25, 12:17 AM

why not cloudflare.com?

Ddave why not cloudflare.com?

Hello, I’m Allie!•5/7/25, 12:17 AM

Shorter

Hello, I’m Allie!•5/7/25, 12:17 AM

Anything on Ent should work

daveOP•5/7/25, 12:17 AM

1.1.1.1 doesn't work, right?

HHello, I’m Allie!Anything on Ent should work

daveOP•5/7/25, 12:17 AM

why would non-ent not work?

Ddave why would non-ent not work?

Hard@Work•5/7/25, 12:18 AM

Ent just ensures that you get the same colo. I don't remember off the top of my head, but Email Workers might run in a colo that wouldn't be served by a lower-tier plan

Ddave 1.1.1.1 doesn't work, right?

Hard@Work•5/7/25, 12:19 AM

$ curl https://1.1.1.1/cdn-cgi/trace
fl=4f603
h=1.1.1.1
ip=111.111.111.111
ts=1746577084.432
visit_scheme=https
uag=curl/8.7.1
colo=SJC
sliver=001-tier4
http=http/2
loc=US
tls=TLSv1.3
sni=off
warp=on
gateway=on
rbi=off
kex=XREALKEY

$ curl https://1.1.1.1/cdn-cgi/trace
fl=4f603
h=1.1.1.1
ip=111.111.111.111
ts=1746577084.432
visit_scheme=https
uag=curl/8.7.1
colo=SJC
sliver=001-tier4
http=http/2
loc=US
tls=TLSv1.3
sni=off
warp=on
gateway=on
rbi=off
kex=XREALKEY

HHard@Work ```sh $ curl https://1.1.1.1/cdn-cgi/trace fl=4f603 h=1.1.1.1 ip=111.111.111.111...

daveOP•5/7/25, 12:36 AM

Oh I mean from Workers.

I remember it used to require a hostname.

Ddave Oh I mean from Workers. 🙂 I remember it used to require a hostname.

Hard@Work•5/7/25, 12:38 AM

IIRC Quad1 is special

Ddave Oh I mean from Workers. 🙂 I remember it used to require a hostname.

Hard@Work•5/7/25, 12:41 AM

export default {
    async fetch() {
        return new Response(`You are hitting ${(
            await (await fetch("https://1.1.1.1/cdn-cgi/trace")).text()
        ).match(/^colo=(.+)/m)[1]} via Quad1.`)
    }
}

export default {
    async fetch() {
        return new Response(`You are hitting ${(
            await (await fetch("https://1.1.1.1/cdn-cgi/trace")).text()
        ).match(/^colo=(.+)/m)[1]} via Quad1.`)
    }
}

works on https://workers.cloudflare.com/playground?name=quad-fun

Is there any trick(s) I can do to prevent a 100 second timeout on a Worker subrequest? I do not cont

Similar Threads

Similar Threads

Similar Threads