so if we hit an API on the backend, and it took a while to respond it counts in that execution time?

Jjjjrmy CPU time counts as like any backend processing SSR right?

C

Chaika•12/14/23, 10:30 PM

If Astro uses SSR it would include time to generate the page yea

Jjjjrmy so if we hit an API on the backend, and it took a while to respond it counts in ...

C

Chaika•12/14/23, 10:30 PM

not if it's just waiting on the fetch request/API to respond

C

Chaika•12/14/23, 10:31 PM

CPU Time is time the CPU is actively doing something, generating html, deserializing a json response, etc. Waiting on i/o like fetch requests, etc, doesn't consume any cpu time

Oom I have a question.. i was using remix on pages.. i was trying to put the html ge...

C

Chaika•12/15/23, 2:29 PM

Functions even on Custom Domains run on the pages.dev, you essentially have two layers of cache
CustomDomain.com -> Cache -> Pages Functions -> Cache
You know you're getting hits based on log/output (you should also be able to see the age header), you can ignore the header being overriden by the custom domain cache, or you could use a cache rule on your zone to use the custom domain cache instead. Just be careful with what you're caching

T

t•12/15/23, 5:45 PM

Hi I am having issues with ctx.waitUntil. It seems to be blocking the rest of the request

K

kian•12/15/23, 5:52 PM

How are you using it? Do you have a code example?

T

t•12/15/23, 5:53 PM

export default {
 async fetch(
    request: Request,
    env: Env,
    ctx: ExecutionContext
  ): Promise<Response> {

    ... do stuff

    ctx.waitUntil(asyncFunction())

     return new Response(...)
}

export default {
 async fetch(
    request: Request,
    env: Env,
    ctx: ExecutionContext
  ): Promise<Response> {

    ... do stuff

    ctx.waitUntil(asyncFunction())

     return new Response(...)
}

export default {
 async fetch(
    request: Request,
    env: Env,
    ctx: ExecutionContext
  ): Promise<Response> {

    ... do stuff

    ctx.waitUntil(asyncFunction())

     return new Response(...)
}

export default {
 async fetch(
    request: Request,
    env: Env,
    ctx: ExecutionContext
  ): Promise<Response> {

    ... do stuff

    ctx.waitUntil(asyncFunction())

     return new Response(...)
}

T

t•12/15/23, 5:53 PM

If i remove the waitUntil line it completes immediately. If i add it back in, it seems to wait for my long running async function to complete before returning a response. And this is for a deployed worker, not running locally

K

kian•12/15/23, 5:54 PM

Is this your actual code? I can’t reproduce it with that layout.

T

t•12/15/23, 6:08 PM

can't share full code but have narrowed it down to the function i am passing to waitUntil

K

kian•12/15/23, 6:08 PM

If you’re awaiting what you pass to waitUntil, that would break

T

t•12/15/23, 6:08 PM

This function works as expected:

async function wait(ms: number) {
  await new Promise((resolve) => setTimeout(resolve, ms));
  console.log("hello");
}

async function wait(ms: number) {
  await new Promise((resolve) => setTimeout(resolve, ms));
  console.log("hello");
}

async function wait(ms: number) {
  await new Promise((resolve) => setTimeout(resolve, ms));
  console.log("hello");
}

async function wait(ms: number) {
  await new Promise((resolve) => setTimeout(resolve, ms));
  console.log("hello");
}

and this waits until completion before returning:

await init((imports) => WebAssembly.instantiate(wasm, imports));
const encoder = new Tiktoken(
  model.bpe_ranks,
  model.special_tokens,
  model.pat_str
);

async function handleCountTokens(data: string) {
  return new Promise((resolve, reject) => {
    const tokens = encoder.encode(data);
    // encoder.free();

    resolve({
      tokens_count: tokens.length,
    });
  });
}

await init((imports) => WebAssembly.instantiate(wasm, imports));
const encoder = new Tiktoken(
  model.bpe_ranks,
  model.special_tokens,
  model.pat_str
);

async function handleCountTokens(data: string) {
  return new Promise((resolve, reject) => {
    const tokens = encoder.encode(data);
    // encoder.free();

    resolve({
      tokens_count: tokens.length,
    });
  });
}

await init((imports) => WebAssembly.instantiate(wasm, imports));
const encoder = new Tiktoken(
  model.bpe_ranks,
  model.special_tokens,
  model.pat_str
);

async function handleCountTokens(data: string) {
  return new Promise((resolve, reject) => {
    const tokens = encoder.encode(data);
    // encoder.free();

    resolve({
      tokens_count: tokens.length,
    });
  });
}

await init((imports) => WebAssembly.instantiate(wasm, imports));
const encoder = new Tiktoken(
  model.bpe_ranks,
  model.special_tokens,
  model.pat_str
);

async function handleCountTokens(data: string) {
  return new Promise((resolve, reject) => {
    const tokens = encoder.encode(data);
    // encoder.free();

    resolve({
      tokens_count: tokens.length,
    });
  });
}

T

t•12/15/23, 6:09 PM

usage is:

ctx.waitUntil(handleCountTokens(text));

ctx.waitUntil(handleCountTokens(text));

ctx.waitUntil(handleCountTokens(text));

ctx.waitUntil(handleCountTokens(text));

Tt This function works as expected: ```async function wait(ms: number) { await ne...

I

Isaac McFadyen•12/15/23, 6:09 PM

You are returning a new promise from an async function.

I

Isaac McFadyen•12/15/23, 6:09 PM

So you're effectively returning a Promise<Promise<>>

I

Isaac McFadyen•12/15/23, 6:09 PM

Get rid of the

async

async

async

async there.

T

t•12/15/23, 6:09 PM

trying that now, I originally wasn't returning a promise and had the same issue

T

t•12/15/23, 6:11 PM

still same issue

IIsaac McFadyen So you're effectively returning a Promise<Promise<>>

H

Hello, I’m Allie!•12/15/23, 6:12 PM

JS should automatically flatten promises when awaiting

HHello, I’m Allie!JS should automatically flatten promises when awaiting

I

Isaac McFadyen•12/15/23, 6:18 PM

Yes but it's being sent into waitUntil which calls into

workerd

workerd

workerd

workerd

I

Isaac McFadyen•12/15/23, 6:18 PM

And I'm uncertain if they follow that behavior.

T

t•12/15/23, 6:58 PM

going to split this waitUntil into a separate worker i think

T

t•12/15/23, 6:59 PM

wouldn't have anything to do with it using wasm would it?

J

jparrish88•12/15/23, 10:19 PM

Hi All, I'm curious if I am doing something wrong. I'm trying to follow these instructions for D1, but I don't see D1 database bindings in that menu "Settings > Functions > D1 Database. https://developers.cloudflare.com/pages/functions/bindings/ It seems to be the only one that is missing in my pages settings>functions.

Bindings · Cloudflare Pages docs

A binding enables your Pages Functions to interact with resources on the Cloudflare developer platform. Use bindings to integrate your Pages Functions …

K

kian•12/16/23, 12:24 AM

K

kian•12/16/23, 12:24 AM

What do you see in the middle of R2 and Service Bindings, nothing?

Kkian Click to see attachment

J

jparrish88•12/17/23, 2:17 AM

I should Of taken a picture but now it's totally there. I wonder if it was being updated, or temporarily offline. Either way, thank you so much for your help.

A

Angius•12/17/23, 10:58 PM

Is there any way to run a specific function on a CRON schedule, or is scheduling exclusive to Workers?

A

Angius•12/17/23, 11:01 PM

Huh, looks like I'm not the only one with this question... and not the only one that could not find an answer: https://community.cloudflare.com/t/cron-queues-support-in-cloudflare-pages/577004

A

Angius•12/17/23, 11:03 PM

There's Advanced Mode, I guess, but OOF

no support for

.ts

.ts

.ts

.ts

K

kian•12/17/23, 11:04 PM

Advanced Mode doesn’t support scheduled handlers either

K

kian•12/17/23, 11:04 PM

You need Workers if you want queue, scheduled, tail, etc handlers

A

Angius•12/17/23, 11:05 PM

So multiple projects just to run some check on a schedule, just what I needed KEKW

A

Angius•12/17/23, 11:06 PM

Thanks for the confirmation, though

A

Angius•12/17/23, 11:09 PM

Oh, cool, Netlify has schedules functions in beta, I won't need multiple projects after all

J

jjjrmyOP•12/18/23, 11:07 PM

oh no, is there any way to go over teh 1MB limit?

K

kian•12/18/23, 11:11 PM

Paid plan has a 10 MiB limit.

J

jjjrmyOP•12/19/23, 12:05 AM

I'm trying to get as far as I can on the free plan for now

K

kian•12/19/23, 12:07 AM

You can’t go over it, you can only try and reduce the size of your Functions bundle.

J

jjjrmyOP•12/19/23, 12:44 AM

yeah I understand, it's better I reduce anyways until I can't

J

jjjrmyOP•12/20/23, 7:29 PM

is there a list of or IP ranges for Cloudflare? My data provider requires a whitelist for incoming POST requests... any solution?

Jjjjrmy is there a list of or IP ranges for Cloudflare? My data provider requires a whit...

D

DarkDeviL•12/20/23, 7:31 PM

https://www.cloudflare.com/ips/

J

jjjrmyOP•12/20/23, 7:31 PM

R

RenderCoder•12/21/23, 8:18 AM

Is there any way to debug the page function? I can not find a solution

S

savtrip•12/21/23, 10:33 AM

Hey guys, I am migrating websites to cloudflare that utilise the same API keys, for example for a transactional email service. Is there a way to set an API key across multiple projects just like Netlify does with global environment variables? Without global environment variables if a key expires for any reason I need to go and set it on up to 50 separate websites.

RRenderCoder Is there any way to debug the page function? I can not find a solution

S

savtrip•12/21/23, 10:36 AM

I am utilising wrangler to test and debug my functions locally:

npx wrangler pages dev -- SERVER COMMAND

npx wrangler pages dev -- SERVER COMMAND

npx wrangler pages dev -- SERVER COMMAND

npx wrangler pages dev -- SERVER COMMAND . You are also able to get the logs of your functions by pressing "d".

Ssavtrip Hey guys, I am migrating websites to cloudflare that utilise the same API keys, ...

A

Altrius•12/21/23, 11:06 AM

you could maybe use KV, but maybe encrypt it first, just to be safe

so if we hit an API on the backend, and it took a while to respond it counts in that execution time?

Similar Threads

Similar Threads

Similar Threads