Is this still the case? - Cloudflare Developers

© 2026 Hedgehog Software, LLC

More

Communities Docs About Terms Privacy

dave

Ddave Is this still the case?

Isaac McFadyen

I

Isaac McFadyen•12/28/23, 2:13 AM

As far as I know, yes.

dave

D

daveOP•12/28/23, 2:13 AM

oh.

Isaac McFadyen

I

Isaac McFadyen•12/28/23, 2:15 AM

Yeah, see https://developers.cloudflare.com/cache/how-to/cache-keys/

Isaac McFadyen

I

Isaac McFadyen•12/28/23, 2:15 AM

Screenshot_2023-12-27_at_21.15.09.png

Will

W

Will•12/28/23, 10:02 AM

what happened if I need more than 10ms for my operation?

Will

W

Will•12/28/23, 10:03 AM

let's say 12ms

Will

W

Will•12/28/23, 10:03 AM

I see that my median worker cpu time is 12ms

Will

W

Will•12/28/23, 10:03 AM

but not sure why it's still running, cause I have exceeded the limit

Ashley

A

Ashley•12/28/23, 11:34 AM

It’s not exact in my experience, I had a worker that would sometimes exceed CPU usage (on the old plans) by 20-30 ms and still “succeed” and some that were 5ms over and fail - the paid plan ($5/month) has a much higher CPU limit per invocation though if you do hit it

Erisa

E

Erisa•12/28/23, 11:37 AM

The limit is applied quite loosely and doesnt stop execution unless you go over it by a lot or several times. But as a general rule if your median cpu time is above the limit then even if it didn't trigger this time it might in future so its better to get usage down or switch to a paid plan to avoid interruption

Erisa

E

Erisa•12/28/23, 11:38 AM

Paid plans allow for up to 30 seconds of CPU usage per request, where you pay more if you use more (above the amount you get for $5/mo)

Bram

BBram that is good to know, I understand that the binding is identical to `Fetcher` th...

schkovich

S

schkovich•12/28/23, 1:17 PM

a bit confusing? it's utterly confusing NotLikeThis

NotLikeThis

kian

Kkian Probably for the best so `new URL(req.url)` doesn't break in Service Bindings/Du...

schkovich

S

schkovich•12/28/23, 1:23 PM

Thank you for the tip. I think i will opt for cloudflare

cloudflare

documentation URL in this case

valkyrie_pilot

V

valkyrie_pilot•12/28/23, 4:41 PM

does cloudflare have a nice solution for OAuth CSRF tokens? I don’t want to use kv, as it’s not really designed for that… can i keep the worker waiting somehow?

Unsmart

U

Unsmart•12/28/23, 4:46 PM

KV is perfectly fine for CSRF tokens

can also use durable objects probably cheaper and faster also

Unsmart

UUnsmart KV is perfectly fine for CSRF tokens 🤷 can also use durable objects probably ch...

valkyrie_pilot

V

valkyrie_pilot•12/28/23, 5:01 PM

the main problem is the write pricing

valkyrie_pilot

V

valkyrie_pilot•12/28/23, 5:05 PM

Though i was clearly misremembering the pricing, as i thought the free tier was 1000 a month and its 1000 a day

valkyrie_pilot

Vvalkyrie_pilot the main problem is the write pricing

Unsmart

U

Unsmart•12/28/23, 5:24 PM

Durable objects are a lot cheaper. $1/million writes + 0.15/million for the invokes. D1 can also be a choice (for non production critical apps currently)

valkyrie_pilot

Vvalkyrie_pilot does cloudflare have a nice solution for OAuth CSRF tokens? I don’t want to use ...

dave

D

daveOP•12/28/23, 5:34 PM

why not D1?

dave

D

daveOP•12/28/23, 5:44 PM

@kian Is there any downside to storing S3 objects with gzip compression?

Unsmart

UUnsmart Durable objects are a lot cheaper. $1/million writes + 0.15/million for the invo...

valkyrie_pilot

V

valkyrie_pilot•12/28/23, 5:47 PM

nah, for my volumes 1000 writes/day is plenty

dave

Ddave Is the Cache API shared among all Workers in a zone?

dave

D

daveOP•12/28/23, 7:42 PM

Is the Cache API shared with

fetch

fetch

's cache?

Original message was deleted

dave

D

daveOP•12/28/23, 9:27 PM

if you use aws4fetch, you can get free caching without needing to fiddle with the Cache API manually.

const aws_client = new AwsClient({
  accessKeyId: c.env.S3_ACCESS_KEY_ID,
  secretAccessKey: c.env.S3_SECRET_ACCESS_KEY,
  retries: 0,
  service: "s3",
})

const s3_url = new URL(`${c.env.S3_ENDPOINT}/${sha256}.json`);

const s3_result = await aws_client.fetch(s3_url.toString(), {
  method: "GET",
  cf: {
    cacheEverything: true,
    cacheTtl: 31536000, // 1 year
  }
});

const aws_client = new AwsClient({
  accessKeyId: c.env.S3_ACCESS_KEY_ID,
  secretAccessKey: c.env.S3_SECRET_ACCESS_KEY,
  retries: 0,
  service: "s3",
})

const s3_url = new URL(`${c.env.S3_ENDPOINT}/${sha256}.json`);

const s3_result = await aws_client.fetch(s3_url.toString(), {
  method: "GET",
  cf: {
    cacheEverything: true,
    cacheTtl: 31536000, // 1 year
  }
});

Original message was deleted

dave

D

daveOP•12/28/23, 9:28 PM

Why?

dave

D

daveOP•12/28/23, 11:04 PM

Do I pay for the idle time on the setTimeout?

for(let i = 0; i < 10; i++) {
    await new Promise(r => setTimeout(r, 5000));
    const result = await fetch(loc);
// ...
}

for(let i = 0; i < 10; i++) {
    await new Promise(r => setTimeout(r, 5000));
    const result = await fetch(loc);
// ...
}

James

J

James•12/28/23, 11:05 PM

In Unbound, you would pay for that wall time, yes. With Bundled, or the new Standard, no you will not

dave

D

daveOP•12/28/23, 11:32 PM

thanks!

dave

D

daveOP•12/28/23, 11:32 PM

I switched to standard just now

dave

D

daveOP•12/28/23, 11:34 PM

How do I not have enough subrequests for smart placement..?

James

J

James•12/28/23, 11:50 PM

I'm not sure how accurate those analytics are for smart placement. Are you making multiple subrequests per request?

Unsmart

U

Unsmart•12/28/23, 11:54 PM

Exhibit a of why I hate smart placement instead of manual YEP

YEP

James

JJames I'm not sure how accurate those analytics are for smart placement. Are you makin...

dave

D

daveOP•12/28/23, 11:57 PM

yes. 3 subrequests if there's no cache, or 0 if we hit the cache.

Unsmart

U

Unsmart•12/28/23, 11:58 PM

I wonder if the cache messes with it

dave

D

daveOP•12/28/23, 11:58 PM

Will cacheEverything: truecacheEverything: true cache non-200 responses?

Unsmart

U

Unsmart•12/28/23, 11:59 PM

It should afaik you can set specific codes to limit the cache

dave

D

daveOP•12/28/23, 11:59 PM

oh crap

dave

Ddave @kian Is there any downside to storing S3 objects with gzip compression?

kian

K

kian•12/29/23, 12:30 AM

Nope

Unsmart

UUnsmart It should afaik you can set specific codes to limit the cache

dave

D

daveOP•12/29/23, 12:31 AM

sanity check, as long as my remote doesn't have a cache-control header, non-200's won't be cached, right?

dave

Ddave sanity check, as long as my remote doesn't have a cache-control header, non-200'...

Erisa

E

Erisa•12/29/23, 12:38 AM

No, there are defaults https://developers.cloudflare.com/cache/how-to/configure-cache-status-code/#edge-ttl

Erisa

EErisa No, there are defaults <https://developers.cloudflare.com/cache/how-to/configure...

dave

D

daveOP•12/29/23, 12:39 AM

ahh.. does a cache-control header override those?

Erisa

E

Erisa•12/29/23, 12:39 AM

It does

Erisa

E

Erisa•12/29/23, 12:40 AM

cache-control: no-storecache-control: no-store should do the trick

Erisa

dave

D

daveOP•12/29/23, 12:48 AM

oh sweet, so cacheTtlcacheTtl is probably a bad idea, ty!

I've removed it from my code.

Unsmart

UUnsmart Exhibit a of why I hate smart placement instead of manual <:YEP:8072727784654111...

dave

D

daveOP•12/29/23, 12:50 AM

manual placement isn't a thing, is it?

dave

Ddave manual placement isn't a thing, is it? 😛

Unsmart

U

Unsmart•12/29/23, 12:50 AM

Unfortunately it’s only “smart” placement currently

Erisa

EErisa No, there are defaults <https://developers.cloudflare.com/cache/how-to/configure...

dave

D

daveOP•12/29/23, 12:52 AM

How does

fetch

fetch

's cache behave with query strings?

dave

Ddave How does `fetch`'s cache behave with query strings?

Erisa

E

Erisa•12/29/23, 12:53 AM

It respects whatever cache setting you have for the zone the worker runs on

Erisa

EErisa It respects whatever cache setting you have for the zone the worker runs on

dave

D

daveOP•12/29/23, 12:53 AM

is there a way to configure that per-fetch?

Erisa

E

Erisa•12/29/23, 12:55 AM

I don't believe so, no

const aws_client = new AwsClient({
  accessKeyId: c.env.S3_ACCESS_KEY_ID,
  secretAccessKey: c.env.S3_SECRET_ACCESS_KEY,
  retries: 0,
  service: "s3",
})

const s3_url = new URL(`${c.env.S3_ENDPOINT}/${sha256}.json`);

const s3_result = await aws_client.fetch(s3_url.toString(), {
  method: "GET",
  cf: {
    cacheEverything: true,
    cacheTtl: 31536000, // 1 year
  }
});

const aws_client = new AwsClient({
  accessKeyId: c.env.S3_ACCESS_KEY_ID,
  secretAccessKey: c.env.S3_SECRET_ACCESS_KEY,
  retries: 0,
  service: "s3",
})

const s3_url = new URL(`${c.env.S3_ENDPOINT}/${sha256}.json`);

const s3_result = await aws_client.fetch(s3_url.toString(), {
  method: "GET",
  cf: {
    cacheEverything: true,
    cacheTtl: 31536000, // 1 year
  }
});