hmm so it seems to be persistent when i already have the service running and i startup my backend (b

hmm so it seems to be persistent when i already have the service running and i startup my backend (backend is dependent on the service)

ddts86•4/27/24, 11:41 AM

what does this error mean: [wrangler:err] InferenceUpstreamError: {"success":false,"errors":[{"code":10000,"message":"Authentication error"}]}

ddts86•4/27/24, 11:43 AM

some config issue? I have confirmed that the AI api works fine.

Jackenmen•4/28/24, 12:30 AM

Is it possible to define secrets in

wrangler.toml

wrangler.toml

so that they get added to types that are auto-generated by

wrangler types

wrangler types

?
I can specify regular env vars in

[vars]

[vars]

but trying to set such var's value with

wrangler secret put

wrangler secret put

after deployment results in an error saying that the variable is already set and I have to delete all of those encrypted variables first which doesn't really make it seem like the right way to do it.

originally posted in #workers-discussions but I deleted it after finding out that there's a channel specifically for wrangler rather than just workers in general

Dddts86 what does this error mean: [wrangler:err] InferenceUpstreamError: {"success":fal...

ddts86•4/28/24, 2:40 AM

The deployed worker works but not the local dev env. But it worked days ago.

ddts86•4/28/24, 3:53 AM

I think I got it. I searched a lot and it always tells. that a customed api token configured
when I try to deploy and I found that what does this mean, I have a env var CLOUDFLARE_AUTH_TOKEN and api var configured.
when I remove these two and it works.

Sinon•4/28/24, 6:12 AM

Hi, how can I deploy workers to different accounts if I have multiple accounts?

ddts86•4/28/24, 7:08 AM

you should login with different account I think. wrangler login.

Dddts86 you should login with different account I think. wrangler login.

Sinon•4/28/24, 7:20 AM

I don't see any documents about that

SSinon I don't see any documents about that

ddts86•4/28/24, 7:59 AM

just my guess.

SSinon Hi, how can I deploy workers to different accounts if I have multiple accounts?

Hello, I’m Allie!•4/28/24, 8:00 AM

Have a single account that has perms for the others. Or, pass a

CLOUDFLARE_API_TOKEN

CLOUDFLARE_API_TOKEN

for the account you want to use

HHello, I’m Allie!Have a single account that has perms for the others. Or, pass a `CLOUDFLARE_API_...

Sinon•4/28/24, 8:03 AM

yes, I have a main account that has permission for both of them but after the first deploy it asks me where to deploy, after that it auto select that account for me

Sinon•4/28/24, 8:03 AM

anyway I can switch it?

SSinon yes, I have a main account that has permission for both of them but after the fi...

Hello, I’m Allie!•4/28/24, 8:17 AM

I believe it stores the account selected in the

node_modules/.wrangler

node_modules/.wrangler

folder

Hello, I’m Allie!•4/28/24, 8:17 AM

So if you nuke that folder, it should allow you to choose it again

Sinon•4/28/24, 1:55 PM

Ok let me try, ty

bun•4/28/24, 9:09 PM

im getting

  cause: ConnectTimeoutError: Connect Timeout Error

  cause: ConnectTimeoutError: Connect Timeout Error

for wrangler dev, both on

next-on-pages

next-on-pages

(setupDevPlatform) and regular worker project

bun•4/28/24, 9:09 PM

whats going on

bun•4/28/24, 9:24 PM

fixed after restarting my mac

bun•4/28/24, 9:24 PM

this is the second time i came across this error, my environment doesnt contain any proxies or anything. so its a weird bug

JJackenmen Is it possible to define secrets in `wrangler.toml` so that they get added to ty...

AndyJessop•4/29/24, 12:59 PM

For local development, secrets should be defined in a

.dev.vars

.dev.vars

file (remember to add to

.gitignore

.gitignore

), just as you would in a

.env

.env

. And then remotely they would either be set via the UI, or using the

wrangler secret put

wrangler secret put

command. You can find all the details here: https://developers.cloudflare.com/workers/configuration/secrets/

Secrets should not be added to your

wrangler.toml

wrangler.toml

because that file is committed.

Jackenmen•4/29/24, 1:00 PM

Wrangler toml is meant to be source of truth so what I was wondering is if I can define that a secret is needed in it

JJackenmen Wrangler toml is meant to be source of truth so what I was wondering is if I can...

Hello, I’m Allie!•4/29/24, 1:01 PM

You cannot, no

Jackenmen•4/29/24, 1:01 PM

Or well, the point is that wrangler generates typescript definitions

Jackenmen•4/29/24, 1:01 PM

And if I can't somehow let wrangler know that those will exist, it won't include them

Isaac McFadyen•4/29/24, 1:01 PM

Wrangler does not currently generate TS definitions for env vars.

Isaac McFadyen•4/29/24, 1:01 PM

The interfaces are generic though so you can define them yourself.

IIsaac McFadyen Wrangler does not currently generate TS definitions for env vars.

Jackenmen•4/29/24, 1:02 PM

It does though?

Isaac McFadyen•4/29/24, 1:02 PM

And the reason secrets can't go in wrangler.toml is that wrangler.toml can sometimes be shared publicly whereas secrets need to stay secret.

Jackenmen•4/29/24, 1:02 PM

wrangler cf-types command

Isaac McFadyen•4/29/24, 1:02 PM

Oh, interesting - wonder when they added that.

Isaac McFadyen•4/29/24, 1:02 PM

Thanks for the info.

Jackenmen•4/29/24, 1:03 PM

Sorry, the command is

wrangler types

wrangler types

, got that wrong

Jackenmen•4/29/24, 1:04 PM

It generates a file such as this one:
https://github.com/Jackenmen/cf-email-worker-transcribe-audio-attachments/blob/main/worker-configuration.d.ts

Jackenmen•4/29/24, 1:04 PM

But it doesn't seem like there's a way to let wrangler know about secrets so that can include it in the automatically generated type definitions

Jackenmen•4/29/24, 1:26 PM

Sure!

Jackenmen•4/29/24, 1:40 PM

I actually have another question about wrangler's automatically generated type definitions. Is it possible for AI binding to have a proper type instead of unknown? For comparison, send email binding uses correct type but it is defined somewhat differently from the AI binding (

type = "send_email"

type = "send_email"

binding = "AI"

binding = "AI"

). I know that

Ai

Ai

type exists but it isn't used automatically so not sure if it's a missing feature too or if perhaps I just need to set something to make it work like send_email does

Jackenmen•4/29/24, 1:42 PM

I can make an issue that as well then, thanks

Evil Morty•4/29/24, 7:29 PM

Hello,
i use this command to deploy my sveltekit app to Clouddflare pages

npm run build && npx wrangler pages deploy .svelte-kit/cloudflare --commit-dirty=true --env production

npm run build && npx wrangler pages deploy .svelte-kit/cloudflare --commit-dirty=true --env production

this is my wrangler.toml file

name = "resume"
send_metrics = false
pages_build_output_dir = ".svelte-kit/cloudflare"

[[d1_databases]]
binding = "DB" # i.e. available in your Worker on env.DB
database_name = "resume_db"
database_id = "2ed32f58-b02a-454f-a14d-6eaf9dc7bc4s"
migrations_dir = "migrations"

[[r2_buckets]]
binding = 'BUCKET' # <~ valid JavaScript variable name
bucket_name = 'resume'

[vars]

name = "resume"
send_metrics = false
pages_build_output_dir = ".svelte-kit/cloudflare"

[[d1_databases]]
binding = "DB" # i.e. available in your Worker on env.DB
database_name = "resume_db"
database_id = "2ed32f58-b02a-454f-a14d-6eaf9dc7bc4s"
migrations_dir = "migrations"

[[r2_buckets]]
binding = 'BUCKET' # <~ valid JavaScript variable name
bucket_name = 'resume'

[vars]

it always deploy to the preview environment
what i can i make it deploy to production
thank you

EEvil Morty Hello, i use this command to deploy my sveltekit app to Clouddflare pages ```bas...

Evil Morty•4/29/24, 7:42 PM

adding

--branch=main

--branch=main

did the trick for me

Vincent•5/2/24, 8:17 AM

Good morning! We're using a custom build, but would also love to use source maps for better stack traces. Is it possible to use

upload_source_maps = true

upload_source_maps = true

together with a custom esbuild? We are already generating a sourcemap. We just need to find a way to have Wrangler to pick it up, I think? Any tips welcome!

Erez•5/2/24, 3:03 PM

I ran into a very weird issue today.
I deploy my worker as usual, and it seem to be successful in the terminal and inside the cloudflare dashboard. However, nothing changes. I have also tried to run wrangler locally and I am experiencing the same thing.

Once in every 10+~ deploys, the changes do take affect and work.
What's going on? I have been using the same setup for a while now.

AGuyThatCodes•5/4/24, 10:31 PM

I am writing a load simulation tester for my worker using wrangler. Curious about what kinds of tools and instrumentation are available (I.e.; avg request time, request counts, simulated billing etc.)

kian•5/4/24, 10:37 PM

Wrangler doesn't have any of that as far as I'm aware

kian•5/4/24, 10:38 PM

Deployed Workers have avg CPU time/request count metrics, no simulated billing

AGuyThatCodes•5/4/24, 11:09 PM

No worries, I could always implement my own instrumentation (although somewhat painfully)

AGuyThatCodes•5/4/24, 11:16 PM

I wonder if there is some kind of interception capabilities for internal fetches

johtso•5/6/24, 12:19 PM

Is there any chance this might get some eyes? Such a big difference in behaviour between testing and production environment for something as important as

fetch

fetch

probably wants fixing https://github.com/cloudflare/workers-sdk/issues/5571

GitHub

🐛 BUG: Fetching a URL that doesn't exist triggers an "internal erro...

Which Cloudflare product(s) does this pertain to? Workers Runtime What version(s) of the tool(s) are you using? 3.48.0 [wrangler] What version of Node are you using? 21.7.1 What operating system an...

nathanl•5/8/24, 2:39 AM

hi,Does anyone know how to debug in vscode?

chronark•5/10/24, 8:58 AM

a flag to get the latest versionId would be so helpful, right now it's incredibly annoying and error prone to automate rollouts

wrangler deployments list --current --json
{ "versionId": "...", ... }

wrangler deployments list --current --json
{ "versionId": "...", ... }

name = "resume" send_metrics = false pages_build_output_dir = ".svelte-kit/cloudflare" [[d1_databases]] binding = "DB" # i.e. available in your Worker on env.DB database_name = "resume_db" database_id = "2ed32f58-b02a-454f-a14d-6eaf9dc7bc4s" migrations_dir = "migrations" [[r2_buckets]] binding = 'BUCKET' # <~ valid JavaScript variable name bucket_name = 'resume' [vars]

hmm so it seems to be persistent when i already have the service running and i startup my backend (b

Similar Threads