You would do something like ```ts app.use('/input/*', (c, next) => bearerAuth({ c.env.TOKEN })(c, ne

You would do something like

app.use('/input/*', (c, next) => bearerAuth({ c.env.TOKEN })(c, next));

app.use('/input/*', (c, next) => bearerAuth({ c.env.TOKEN })(c, next));

app.use('/input/*', (c, next) => bearerAuth({ c.env.TOKEN })(c, next));

app.use('/input/*', (c, next) => bearerAuth({ c.env.TOKEN })(c, next));

HHello, I’m Allie!You would do something like ```ts app.use('/input/*', (c, next) => bearerAuth({ ...

Bob•1/18/24, 12:04 PM

My hero! What it I just want to access them outside of the "app." context?

BBob My hero! What it I just want to access them outside of the "app." context?

Hello, I’m Allie!OP•1/18/24, 12:04 PM

Like, in the global scope? Is this for initializing a client or something?

Bob•1/18/24, 12:04 PM

yes

BBob yes

Hello, I’m Allie!OP•1/18/24, 12:04 PM

Then you would initialize it within the

app

app

scope, then store it in a global

Hello, I’m Allie!OP•1/18/24, 12:05 PM

You could also use an AsyncLocalStorage to have it accessible globally(but that requires a compat flag)

Hello, I’m Allie!OP•1/18/24, 12:05 PM

But note, env vars are only ever available in the context of handling an incoming event

Hello, I’m Allie!OP•1/18/24, 12:05 PM

For example, you can't just pull a secret on startup

Bob•1/18/24, 12:11 PM

ok, I guess I just have to repeat initialization code for each endpoint. I'm using Langhian and have to initialize the embeddings and vector store for every GET and POST. Is there a best practice?

BBob ok, I guess I just have to repeat initialization code for each endpoint. I'm usi...

Hello, I’m Allie!OP•1/18/24, 12:24 PM

You can have a single global variable to store them in, so they only get initialized once

Hello, I’m Allie!OP•1/18/24, 12:24 PM

Though note too that isolates aren’t reused very often, so it might be over optimization

Hello, I’m Allie!OP•1/18/24, 12:24 PM

Generally, they are only reused by the same client, in a limited period of time afterward

Bob•1/18/24, 12:35 PM

yeah, It's more so about not having to repeat myself. I guess I can write a little wrapper function with all the static pieces and just call them with the envs with each event.

Bob•1/18/24, 12:37 PM

Thanks for your help!

igorpancheski•1/18/24, 1:35 PM

Guys, i want to create a serverless function for converter .pst to .mbox, but the filesize can be heavy like 30gb or more. I need to understand if i can get a super power high processor and RAM for function on Workers of cloudfare. i finded to do it on aws or google but. i need a machine more stronger for my MVP. I don't care at first moment with the price i just want to validate my idea. someone has some option for helpme ?

zegevlier•1/18/24, 1:37 PM

Workers are extremely unsuited for high memory requirement applications. They're not the right fit for your project

conqr ᯅ•1/18/24, 4:31 PM

worker failed to prewarm: Service Temporarily Unavailable

conqr ᯅ•1/18/24, 4:31 PM

captain_enjoyable•1/18/24, 4:35 PM

Hey there!

I'm trying to use the cache api, but running into issues. I have the follow minimal repro. When I ping this worker, I never get a "Cache hit" response. Any idea what might be going on here? I'm using the exact same key for every request.

async fn main(req: Request, env: Env, ctx: Context) -> Result<Response> {
    let cache = Cache::default();
    
    let path = "http://www.google.com";
    let response: Option<Response> = cache.get(path.clone(), false).await?;
    console_log!("here!");
    match response {
        Some(response) => {
            let mut resp: Response = Response::ok("Cache hit")?; 
            Ok(resp)
        }, 
        None => {
            console_log!("{}", path);
            let mut resp = Response::ok("Cache miss")?; 
            Ok(resp)
        }
    }
}

async fn main(req: Request, env: Env, ctx: Context) -> Result<Response> {
    let cache = Cache::default();
    
    let path = "http://www.google.com";
    let response: Option<Response> = cache.get(path.clone(), false).await?;
    console_log!("here!");
    match response {
        Some(response) => {
            let mut resp: Response = Response::ok("Cache hit")?; 
            Ok(resp)
        }, 
        None => {
            console_log!("{}", path);
            let mut resp = Response::ok("Cache miss")?; 
            Ok(resp)
        }
    }
}

Ccaptain_enjoyable Hey there! I'm trying to use the cache api, but running into issues. I have th...

Hello, I’m Allie!OP•1/18/24, 4:37 PM

Maybe try #

rust-on-workers

HHello, I’m Allie!Maybe try #🦀rust-on-workers

captain_enjoyable•1/18/24, 4:37 PM

Ooo ok! Will do!

sathoro•1/18/24, 4:56 PM

I'm getting a 522 when calling my Worker via its public custom domain from inside the same Worker, any ideas?

sathoro•1/18/24, 4:57 PM

its a blank route that just returns a 200. I can call it fine locally

Unsmart•1/18/24, 4:57 PM

Not surprising lots of issues with that I'd use service bindings

Chaika•1/18/24, 4:58 PM

or you could try seting the no miminal subrequests compat flag https://github.com/cloudflare/workerd/issues/787#issuecomment-1597396851

GitHub

🐛 BUG: Worker <-> Worker request over `custom_domain` returns insta...

Which Cloudflare product(s) does this pertain to? Workers/Other What version of Wrangler are you using? 2.9.0 What operating system are you using? Mac Describe the Bug Same zone worker <-> wo...

sathoro•1/18/24, 4:58 PM

why not surprising? it is just like calling any other API right?

sathoro•1/18/24, 4:58 PM

I'm using it to proxy webhook requests to the correct environment

sathoro•1/18/24, 4:58 PM

service bindings not gonna work for this use case

Ssathoro why not surprising? it is just like calling any other API right?

Chaika•1/18/24, 4:58 PM

no, worker to worker is very special lol

Chaika•1/18/24, 4:58 PM

you can't do same zone worker to worker with routes

sathoro•1/18/24, 4:58 PM

but I actually do it already somewhere else

sathoro•1/18/24, 4:59 PM

although I guess that is technically in a DO

Unsmart•1/18/24, 4:59 PM

Service bindings are actually a perfect use case for proxying to the correct env

why would it not work for you?

sathoro•1/18/24, 4:59 PM

ughh this is a big mess

Chaika•1/18/24, 5:00 PM

Worker to worker same-zone over routes > no
Worker to worker same-zone over Custom Domains -> should work, with compat flag?
Worker to worker different zones -> should work, will use worker ipv6
Worker to worker service bindings -> works and optimal, request doesn't go over network, fast

sathoro•1/18/24, 5:00 PM

did not expect this to not work

sathoro•1/18/24, 5:00 PM

does that flag actually work?

Unsmart•1/18/24, 5:00 PM

I'd personally suggest never calling your own worker with an outbound fetch and always using service bindings personally. Works the best and always

sathoro•1/18/24, 5:01 PM

yeah, I understand

sathoro•1/18/24, 5:01 PM

I don't need to code the extra logic of checking if the current worker is calling the same worker and then figuring out the binding when it should just work as is

Chaika•1/18/24, 5:02 PM

well my understanding is some of that complexity is needed because Workers can runs on routes and call the origin behind them

Chaika•1/18/24, 5:02 PM

wouldn't want to infinite loop

Chaika•1/18/24, 5:03 PM

custom domains were supposed to have the ability to act as a real origin and just work, and they did for a bit, but then got broken with the minimal subrequest stuff, so need compat flag

sathoro•1/18/24, 5:04 PM

thought I'd be making it easy on myself by putting this proxy in the worker instead of in a Supabase edge function, I guess I messed up

sathoro•1/18/24, 5:07 PM

no_minimal_subrequestsno_minimal_subrequests did not seem to work

sathoro•1/18/24, 5:12 PM

btw it works locally, of course. so did not spot the issue until now

sathoro•1/18/24, 5:18 PM

maybe I should proxy it through a DO

addison•1/18/24, 10:10 PM

hey, i have a quick question regarding billing

im piping files from Backblaze B2 through CF Workers and have a simple hono app that grabs files and sends them to the user

will the time it takes for this file to go to the user be counted under CPU time?

Jaden•1/19/24, 3:59 AM

Hi, Do Workers use CDN when serving data from R2 storage ? or can it be configured to use a CDN ?

Unsmart•1/19/24, 4:00 AM

If by CDN you are asking mostly about cache then you can use the Cache API in workers. Or you can use an R2 public domain instead which will directly connect to the CDN assuming all files being public is fine.

async fn main(req: Request, env: Env, ctx: Context) -> Result<Response> { let cache = Cache::default(); let path = "http://www.google.com"; let response: Option<Response> = cache.get(path.clone(), false).await?; console_log!("here!"); match response { Some(response) => { let mut resp: Response = Response::ok("Cache hit")?; Ok(resp) }, None => { console_log!("{}", path); let mut resp = Response::ok("Cache miss")?; Ok(resp) } } }

You would do something like ```ts app.use('/input/*', (c, next) => bearerAuth({ c.env.TOKEN })(c, ne

Similar Threads

Similar Threads

Similar Threads