Do we know if theres a size limit for json file in the bulk kv put command?
Do we know if theres a size limit for json file in the bulk kv put command?
DFor a single file, the size limit is always 25 MB. For the combined limit, I believe it is 100MB total.
Dso the limit is 10k key value pairs or 25mb?
Doh each pair is at max 25mb, and the overall request can't be above 100mb
Whave you added anything to your astro project (was it the starter files or blog?) - i'm unable to reproduce with either, but we're seeing this error a bit and would like to get to the bottom of it womp :(
AI don't think it was the blog one, but the other one that they recommend with some starter files. I didn't change anything
WAh yep it’s broken - thanks for the report, I’ll go have a look 
ADoes anyone have best practices for handling wrangler.toml files in a monorepo? We have them all over, and with apps/packages/workers all needing them to some extent, it can get pretty confusing to make sure we have the right bindings in the right places without massive duplication everywhere.
�AiError
KHey folks, I had a pretty rough dev experience with using durable objects, and it would be nice to get some guidance on best practices.
In a dev branch, i set up my Toml file to have a durable object binding without an env (as per docs), but it looked like when i ran wrangler dev —remote, it created a durable object in prod.
That caused a lot of problems, and a separate branch that wasn’t related was unable to deploy bc it didn’t have the migration to delete the durable object.
Also, it looked like deploying to remote dev would update the durable object in prod.
As an example, I don’t really have a good idea of the dev lifecycle. Like if I create a DO for the first time, what should my toml file look like to test & deploy? And after several iterations, what would it look like (e.g. do I need to delete migrations after every deploy?)
In a dev branch, i set up my Toml file to have a durable object binding without an env (as per docs), but it looked like when i ran wrangler dev —remote, it created a durable object in prod.
That caused a lot of problems, and a separate branch that wasn’t related was unable to deploy bc it didn’t have the migration to delete the durable object.
Also, it looked like deploying to remote dev would update the durable object in prod.
As an example, I don’t really have a good idea of the dev lifecycle. Like if I create a DO for the first time, what should my toml file look like to test & deploy? And after several iterations, what would it look like (e.g. do I need to delete migrations after every deploy?)
HMigrations are only for if you create a new class, rename a class, or delete a class. They don't need to be added for each deploy
Kthanks for responding.
is it expected that doing dev —remote would deploy a production asset?
My unrelated branch was failing to deploy because the toml file didn’t have a reference to a durable object that was only defined in a test branch. I can’t remember the exact error, but I had to add a migration in my unrelated branch to delete the existing durable object and then in a follow up, I needed to remove the migration because now there was no durable object to delete.
It was a headache because I didn’t even deploy the durable object intentionally
is it expected that doing dev —remote would deploy a production asset?
My unrelated branch was failing to deploy because the toml file didn’t have a reference to a durable object that was only defined in a test branch. I can’t remember the exact error, but I had to add a migration in my unrelated branch to delete the existing durable object and then in a follow up, I needed to remove the migration because now there was no durable object to delete.
It was a headache because I didn’t even deploy the durable object intentionally
HMaybe? I don't know the particulars for DOs specifically, but
--remoteI would go local, if possible, to minimize issues like that
Kok, in general, do you know if you have to clean up the migrations in the toml files?
Even if i developed locally, it seems like there’s a step where you need to remove migrations after classes are deleted etc, and that seems annoying if you have version control and work with multiple people
Even if i developed locally, it seems like there’s a step where you need to remove migrations after classes are deleted etc, and that seems annoying if you have version control and work with multiple people
CHow to solve this issue, when using a custom port with HTTPS
known issue with fetch() requests to custom HTTPS ports in published Workers: the custom port will be ignored when the Worker is published using the "wrangler deploy" command.
NAnyone knows how to login to wrangler in project IDX?
XHas been stuck on
Waiting for DNS to propagate (36s) ...Welp, I ctrl + c it, changed some stuff and run
npm run deploy- now it worked fine ¯\_(ツ)_/¯
Hhello, can i load the code deployed from the worker dashboard in wrangler?
Rhttps://github.com/cloudflare/workers-sdk/pull/7031/#issuecomment-2432587096
how do I build something like this on my local?
how do I build something like this on my local?
JI’d start in the contributing guide: https://github.com/cloudflare/workers-sdk/blob/main/CONTRIBUTING.md
Rsomehow, it's not beginner friendly enough, and I had to use private NPM registry to test the package I edit
JJIf you can create an issue with specifically which steps during build and execute failed for you and how exactly they failed, I’m sure the team would love to improve the docs
Cdoes anyone know if there's a ticket that tracks adding workerd support in vitejs powered apps via vitejs's environment api? or the work hasn't been started yet? thx
main use case is to have our remix app in local development mode being able to access r2/d1/ai/queue/etc. without relying on limited support like getPlatformProxy()
main use case is to have our remix app in local development mode being able to access r2/d1/ai/queue/etc. without relying on limited support like getPlatformProxy()
FHas anybody run into the problem where no matter what workaround they try (eg those mentioned in https://github.com/cloudflare/workers-sdk/issues/2874#issuecomment-2439725814 and in the docs at https://developers.cloudflare.com/workers/wrangler/commands/#login) they cannot successfully login to wrangler on a headless machine? IMO, that github issue should be reopened since the problems with the headless login impl were never properly addressed, and none of those workarounds seem generalizable / stable
FI added a little explanation at the end of my comment for what could be changed to implement headless logins like gcp's gcloud, which I've never had problems with. I'd go so far as to try to implement it myself if I could, but I think it necessitates coordinated serverside changes, because it doesn't seem possible to change redirect URLs away from localhost clientside
KI think most people just use API tokens
FLet me try setting that up again, I tried a couple months ago (if by API tokens you mean but I remember running into some issue where wrangler treated me as logged in in some cases and unauthenticated in others. To me that looked like an authn issue on wrangler's end rather than authz.
I started poking through wrangler's internal impl (IIRC the problem was that wrangler login set more login configs than just token so manually setting the token didn't work when those other configs were need). Then I checked out the cf HTTP apis to see if I could do it myself over curl, but then I discovered at least a couple things I wanted to do had no http API documentation / what did exist didn't have enough info to know what to pass as args without studying the wrangler impl too much. Tried anyway for a bit and then called it quits after a couple hours total
I started poking through wrangler's internal impl (IIRC the problem was that wrangler login set more login configs than just token so manually setting the token didn't work when those other configs were need). Then I checked out the cf HTTP apis to see if I could do it myself over curl, but then I discovered at least a couple things I wanted to do had no http API documentation / what did exist didn't have enough info to know what to pass as args without studying the wrangler impl too much. Tried anyway for a bit and then called it quits after a couple hours total
Fanyway it's possible it was user error on my side but if this is a typical user experience (trying to login and it failing with a weird error, having to track down the official docs at https://developers.cloudflare.com/workers/wrangler/commands/#login, realize they don't work, then track down https://github.com/cloudflare/workers-sdk/issues/2874 and try one or more things, then maybe have to track down the cf discord) I humbly request that it be treated as a bug even if "it's always been this way" or doesn't end up blocking most users
AIs there a way to have
wrangler typesQueue
DI don't think there is one today, but this would be a good feature request. Could you open an issue for it? https://github.com/cloudflare/workers-sdk/issues/new/choose
A
Aguys quick question what are the flag parameters for wrangler to create a angular app without ssr.
AHey, I'm having trouble with wrangler and I don't know if it's me or not. I'm currently trying save a secret for the first time using
npx wranger secret put foo
M:Wave_Doggo: Is
--test-scheduledGitHub
Which Cloudflare product(s) does this pertain to? Wrangler What version(s) of the tool(s) are you using? 3.80.0 What version of Node are you using? 20.9.0 What operating system and version are you ...
Wout of interest does
(it shouldn't but i can't see any other changes that would haha)
--x-dev-env=false(it shouldn't but i can't see any other changes that would haha)
MYessir, it does
Whuh..
CHow to disable the spinner on the CLI?
CBasic dev qeustion here. I've patched create-cloudflare to disable the spinner on git commit. This is needed to prevent corruping the gpg password prompt. However, even after I run
pnpm run build --filter create-cloudflarepnpm create cloudflare hello-world --type hello-world --ts --git
TAnyone have advice on getting devtools to open on linux? I have chromium and chrom installed but when I press d I just get "Inspector depends on having a Chromium-based browser installed, maybe you need to install one?"
VVHeyo. Here's the recording of the Wrangler Office Hours
THey, is it possible to connect to the logs when doing integration testing via
unstable_dev?Tah nvm found
logLevel
FWith wrangler 3, is it possible to do development while using remote D1 database?
pnpm devpnpm dev --remote like there was using wrangler dev