Fwiw you still have to pay for all the bandwidth used even when things are cached on vercel. Even fo

Fwiw you still have to pay for all the bandwidth used even when things are cached on vercel. Even for fairly small responses cloudflare can still be cheaper especially depending on the cache hit ratio. (Cloudflare pages also allows you to serve static content completely free with no usage limits)

as an example tho vercel charges $0.40/GB of bandwidth

If you have a response that is 3KB served 1 million times that would add up to 3GB of bandwidth which costs $1.20 on vercel.

That same 3KB response 1 million times on cloudflare workers (lets say with 5ms CPU time to serve) would cost $0.40 on cloudflare ($0.30 for the requests + $0.10 for the CPU time).

Aflina•12/31/23, 9:41 AM

Is there any way to add a WAF rate limiting rule via code (is there any API)? I may want to block some API keys if they miss a payment

Aflina•12/31/23, 11:08 AM

The goal was to prevent additional worker costs. How does a custom rule works? You mean manually adding it?

UUnsmart Fwiw you still have to pay for all the bandwidth used even when things are cache...

Enzo•12/31/23, 11:55 AM

hi there! In fact, I'm not paying on Vercel (and don't want to, since I provide a public API just for hobby)..

so, I was talking mainly about the "Free Plan".. on Vercel, my API works very fine, and after months now I receive an email that I'm using 90% of my Monthly Quota in Execution Time (among other APIs that I have)..

that's why I tried CloudFare, but I reached my quota in 1 day

Enzo•12/31/23, 12:03 PM

I think my API is very small in terms of "Payload Size" and "Bandwidth", but it receives multiple Requests all over the time (that can be cached four hours).. so in Cloudfare I reached my "Daily Requests" so quickly.. but the same API works very fine on Vercel (on Vercel I'm on 90% of my Monthly Quota, but having other APIs on the same account)

Hal Shaw•1/1/24, 4:38 AM

does anyone know how to keep a ws connection with remote server using worker?

Gunnar•1/1/24, 1:47 PM

hi, happy new year everybody!

Gunnar•1/1/24, 1:53 PM

Is there any library that does JWT token validation in Cloudflare Workers?

zegevlier•1/1/24, 1:55 PM

There's jose https://github.com/panva/jose

Gunnar•1/2/24, 9:40 AM

Hi, are you using js/ts/rust in your cf workers? Do you use it for cache/middleware or your main backend?

piffie•1/2/24, 10:03 AM

Monorepo or microservices?
wiht superflare it looks like its much easier to create a monorepo and handle all calls in components in the single codebase.
How does that affect the file limit of cloudflare files over time? will we get an issue when the project grows with a mono architecture and cloudlfare workers? should we directly go for rather having small services where the workers talk to each other?

Tony Leung•1/2/24, 5:05 PM

Hi guys, I’m placing my on-prem service behind cf zero trust and my page function (also placed behind cf zero trust) need to connect to that service, but using fetch normally does not help. I have to pass the cf-auth cookies from the request of page function to my on-prem fetch request header. Is there any more convenient way to authenticate rather than cookie passing?

daniel•1/2/24, 10:33 PM

hi, is there a library that implements the S3 Storage API for Cloudflare Workers?

James•1/2/24, 10:35 PM

https://github.com/mhart/aws4fetch is a pretty common one I've seen people use

GitHub

GitHub - mhart/aws4fetch: A compact AWS client for modern JS enviro...

A compact AWS client for modern JS environments. Contribute to mhart/aws4fetch development by creating an account on GitHub.

James•1/2/24, 10:36 PM

I think the v3 SDK client such as @aws-sdk/client-s3@aws-sdk/client-s3 should work these days too

advaith•1/2/24, 10:40 PM

looks like @cloudflare/workers-types doesnt include headers.getSetCookie(), but it works (in

wrangler dev

wrangler dev

with miniflare 3)

James•1/2/24, 10:43 PM

it does, but you need to set a compat date, sec

James•1/2/24, 10:44 PM

https://www.npmjs.com/package/@cloudflare/workers-types
so rather than @cloudflare/workers-types@cloudflare/workers-types in your

tsconfig.json

tsconfig.json

, you'd want @cloudflare/workers-types/2023-07-01@cloudflare/workers-types/2023-07-01

James•1/2/24, 10:44 PM

(or whatever matches your compat date best)

advaith•1/2/24, 10:45 PM

oh i see, thanks!

advaith•1/2/24, 11:13 PM

i'm trying to use turndown, which supports Node, but I'm getting "Uncaught ReferenceError: window is not defined", even with the nodejs_compat flag. is there a way around that?

James•1/2/24, 11:18 PM

you might be able to inject a

window

window

global, but even if you do. it's then also using DOMParserDOMParser, and

document

document

. This file in particular looks very browser focused: https://github.com/mixmark-io/turndown/blob/master/src/html-parser.js

James•1/2/24, 11:19 PM

you might be able to patch around it, but no guarantees it'll work

advaith•1/2/24, 11:22 PM

you might be able to inject a window global

how would I do that?

advaith•1/2/24, 11:34 PM

ok i think im gonna just make a node or deno server to run turndown, and the worker will just do a http request to that if necessary

James•1/2/24, 11:35 PM

I'm throwing together a quick example to see if I can get this working for you, brb

advaith•1/2/24, 11:36 PM

i tried modifying the lib to get around the

window

window

usage but then i got this

James•1/2/24, 11:47 PM

yeah I got about 3 layers deep in patching and then gave up, sorry. This likely isn't going to work without some serious refactoring of those old dependencies

advaith•1/2/24, 11:50 PM

yeah I figured. thanks for trying though!

dave•1/3/24, 1:52 AM

How do I make a

cloudflared

cloudflared

tunnel only reachable by one of my Workers?

Erisa•1/3/24, 1:55 AM

I believe Access with service tokens in headers and jwt validation on the tunnel is the only truly secure way

EErisa I believe Access with service tokens in headers and jwt validation on the tunnel...

dave•1/3/24, 2:09 AM

so I need to enable this?

Erisa•1/3/24, 2:10 AM

And then setup Access with a service token stored as a secret in the worker, yeah

Erisa•1/3/24, 2:12 AM

Its not ideal but it's the most secure approach available right now to my knowledge, there is no magic button to bind a tunnel to a worker

Ddave how... how did a network connection lost cause a segfault?

Sam•1/3/24, 5:20 AM

Looks similar to what I'm seeing - others have reported it at https://github.com/cloudflare/workerd/issues/1422

GitHub

🐛 wrangler dev error: Received signal #11: Segmentation fault: 11 ·...

Hello! Recently we've been seeing this issue with one of our worker scripts (we have three, the other two seem fine) where after some time being used in dev mode it will segfault and become unr...

dave•1/3/24, 1:29 PM

my theory (untested, I stopped looking into it since I finished my project) is that a thread crashes due to OOM or something, and the network connection lost error is a bit misleading.

dave•1/3/24, 3:41 PM

Is it expected that the cache doesn't work when using

fetch

fetch

locally?

kian•1/3/24, 3:44 PM

Yep.

dave•1/3/24, 3:45 PM

I thought the default caches.defaultcaches.default and the cache that

fetch

fetch

uses were the same?

Erisa•1/3/24, 3:46 PM

When ran in Cloudflare they both access the same CDN cache (sort of, tiered doesnt work with cache api), but when running locally only Cache API is emulated, the

fetch

fetch

doesn't emulate Cloudflare's cache behaviour

dave•1/3/24, 3:47 PM

when you have tiered cache enabled, which cache does

fetch

fetch

use? Always tired cache and never the caches.defaultcaches.default?

UnsmartOP•1/3/24, 3:50 PM

I’d assume it uses both as in if local cache doesn’t have it will go to the upper tier (if it isn’t)

UnsmartOP•1/3/24, 3:52 PM

I’m pretty sure that’s how tiered cache works at least is that every datacenter can still cache the value and not just upper tiers

UUnsmart I’d assume it uses both as in if local cache doesn’t have it will go to the uppe...

dave•1/3/24, 3:54 PM

ahhh, that makes sense! What happens if the local cache is the location of the tiered cache, and the cache API attempts to replace it?

UnsmartOP•1/3/24, 3:57 PM

Would just update its own cache the other tiers would eventually pull from it if needed

dave•1/3/24, 3:58 PM

but then that sounds like the Cache API does work with tiered caching?

UnsmartOP•1/3/24, 3:59 PM

Not really. You would have to use fetch to pull from tiered cache, unless the datacenter you are in already has the cache. At least thats how I understand it

Chaika•1/3/24, 4:03 PM

I assume it depends on the topology as well, if its a location other colos pull from you might update it for tiered cached to use, if its at the bottom/just a local colo it wouldn't ever be used, just a side effect of not storing things twice just for tiered caching/normal cache

Nikita Lapkov•1/3/24, 4:16 PM

Hey! Is there any chance for CloudFlare Workers to support compression algorithms other than just gzip?

We provide a library for developers which has a Rust core compiled to WASM. Currently, gzip is able to compress 3MB WASM module into roughly 1.2MB. We have exhausted all options of compressing the code size on our end (like dependency reduction, unused code elimination, etc). Considering other options, we noticed that LZMA and ZSTD algorithms provide us with 800KB and 770KB WASM modules respectively, which is very promising. Our target is roughly 800KB.

CloudFlare Workers do not allow to run WASM code generated in runtime, so we cannot just unpack the WASM module in runtime ourselves and run it. An ugly alternative to that is including a WASM runtime of our own (like wasm3) and running WASM inside WASM. We would like to avoid that, but this is the only option we have at the moment.

Thus, we are wondering if native support for compressed WASM modules can be added to the platform.

Fwiw you still have to pay for all the bandwidth used even when things are cached on vercel. Even fo

Similar Threads

Similar Threads

Similar Threads