30-50ms, seems to be the same as my endpoint without cache headers added

Chaika•4/26/24, 6:51 PM

are you close enough to ORD that 7ms is physically possible or it could just be local cache of whatever utility you are using/something else?
could try pinging their hostname and seeing icmp latency

WebDev0323•4/26/24, 6:51 PM

Hello, How are you? everyone!
I created the worker, do you know how to integrate with my project?

DawsonOP•4/26/24, 6:51 PM

I am pretty close to a datacenter, not sure if its 7ms close

Chaika•4/26/24, 6:52 PM

what's the endpoint of unkey that you are testing? Curious if I can see the diff at all

DawsonOP•4/26/24, 6:54 PM

I was testing last night, let me find the url real quick

DDawson 30-50ms, seems to be the same as my endpoint without cache headers added

Chaika•4/26/24, 6:55 PM

oh it may be helpful to explain as well: No matter what cache headers you return from a worker you'll never be able to use edge cache unless you specifically use cache api in the worker (which is of course an invocation even just to get a cached response)

Chaika•4/26/24, 6:55 PM

Workers are a layer in front of cache

CChaika oh it may be helpful to explain as well: No matter what cache headers you return...

DawsonOP•4/26/24, 6:56 PM

I used the hono/cachehono/cache package inside of my middleware, they covered everything else

DawsonOP•4/26/24, 6:56 PM

I just specified a cache name and maxage

Chaika•4/26/24, 6:56 PM

ahh ok sounds handy, probably cache api under the hood then

Chaika•4/26/24, 6:57 PM

The Cf-Cache-Status header is an easy way to tell the cache status of an object. If it's missing, didn't even try cache

DawsonOP•4/26/24, 6:57 PM

curl -XPOST 'https://api.unkey.dev/v1/keys.verifyKey' \
  -H 'Content-Type: application/json' \
  -d '{
    "key": "3ZgvH8KKWrP8W8J25xU1KBeN"
  }'

curl -XPOST 'https://api.unkey.dev/v1/keys.verifyKey' \
  -H 'Content-Type: application/json' \
  -d '{
    "key": "3ZgvH8KKWrP8W8J25xU1KBeN"
  }'

This was what I was using, 150 first hit, then 30-40 after that

DawsonOP•4/26/24, 6:57 PM

I set a limited use on the key too, I dont mind if people have it

CChaika The Cf-Cache-Status header is an easy way to tell the cache status of an object....

DawsonOP•4/26/24, 6:58 PM

I do not see this inside of the headers for unkey, im assuming Hoppscotch wouldn't hide it somewhere

CChaika The Cf-Cache-Status header is an easy way to tell the cache status of an object....

DawsonOP•4/26/24, 7:00 PM

I do not have this header

DawsonOP•4/26/24, 7:01 PM

"The Cache middleware uses the Web Standard's Cache API. It caches a given response according to the Cache-Control headers." Could this be something with the browser instead of CF?

DDawson I do not have this header

Chaika•4/26/24, 7:01 PM

with the cache api the usual implementation is such that it would be missing if the cache missed, try hitting it a few times?

Chaika•4/26/24, 7:01 PM

it's also possible unkey doesn't use cf cache or uses it indirectly, like with kv

DawsonOP•4/26/24, 7:01 PM

Yea stil no header, let me see with postman

DDawson ``` curl -XPOST 'https://api.unkey.dev/v1/keys.verifyKey' \ -H 'Content-Type: ...

Chaika•4/26/24, 7:03 PM

fwiw for me (to ORD) that endpoint is slower or comparable to a direct worker

Chaika•4/26/24, 7:03 PM

ex:

DawsonOP•4/26/24, 7:03 PM

ah, may just be hoppscotch. I am getting more realistic numbers with postman

Chaika•4/26/24, 7:04 PM

Chaika•4/26/24, 7:04 PM

direct worker

DawsonOP•4/26/24, 7:04 PM

I would assume it wouldn't do something like that due to 60k stars on gh

DawsonOP•4/26/24, 7:04 PM

but postman is consistent around the 45ms mark

Chaika•4/26/24, 7:04 PM

I mean if you type ping api.unkey.devping api.unkey.dev what's the latency you see icmp?

DDawson I would assume it wouldn't do something like that due to 60k stars on gh

Chaika•4/26/24, 7:04 PM

well a more popular worker might be slightly faster because it's already in memory, but it should be within a few ms. Additionally higher plans get more locations/better routing, but you're hitting the same

CChaika I mean if you type `ping api.unkey.dev` what's the latency you see icmp?

DawsonOP•4/26/24, 7:05 PM

18ms avg, practically the exact same as me

DawsonOP•4/26/24, 7:05 PM

Possibly they just have some amazing db routing. They advertised 22ms response times somewhere too

DDawson Possibly they just have some amazing db routing. They advertised 22ms response t...

Chaika•4/26/24, 7:08 PM

https://github.com/unkeyed/unkey/blob/main/apps/api/wrangler.toml looksl ike durable objects are involved, ord is a dc supporting DOs

GitHub

unkey/apps/api/wrangler.toml at main · unkeyed/unkey

Open source API management platform. Contribute to unkeyed/unkey development by creating an account on GitHub.

DawsonOP•4/26/24, 7:09 PM

Okay, ill go look into this a good bit and see what I can do. Thank you for the help

Chaika•4/26/24, 7:10 PM

can see they use cache as well https://github.com/unkeyed/unkey/blob/ae36a6aa5e4d0d55cacc812d14ebc044f6e3ec91/apps/api/src/pkg/keys/service.ts#L171

Chaika•4/26/24, 7:11 PM

and it looks like their DB is just planetscale w/ drizzle
https://github.com/unkeyed/unkey/blob/ae36a6aa5e4d0d55cacc812d14ebc044f6e3ec91/apps/api/src/pkg/db.ts

Chaika•4/26/24, 7:11 PM

they have a readonly db as well so I assume they've just got a bunch of planetscale replicas: https://planetscale.com/docs/concepts/regions

Chaika•4/26/24, 7:12 PM

that only really matters for the first uncached hit, then it's all Cache API and Durable Objects it seems

DawsonOP•4/26/24, 7:12 PM

ah didn't know ps had replicas, that will be useful.

DawsonOP•4/26/24, 7:12 PM

Thats the whole reason I used turso haha

DawsonOP•4/26/24, 7:13 PM

Yea this helps a ton, experience on workers is great compared to hosting on a dedi

Chaika•4/26/24, 7:15 PM

yea shouldn't be too much you couldn't replicate and get close numbers to them, I assume that 7ms was just that tool being drunk or caching somehow, you get ~18ms to the dc