Is wrangler.toml ignored for next-on-pages
Is wrangler.toml ignored for next-on-pages
HOn deploy, yes
CI have
compatibility_flags = ["nodejs_compat"]
CSo all bindings and config must be done from dashboard?
HYes
CI'm getting this error when trying to call
getRequestContext
CI have this in next.config.mjs:
CI'm on @cloudflare/next-on-pages version 1.9.0
CDo I not need to use getRequestContext?
CThe example included with the S3 next on pages uses it, but on the Cloudflare Docs, it just uses process.env https://developers.cloudflare.com/pages/framework-guides/nextjs/deploy-a-nextjs-site
Cloudflare Docs
Deploy a full-stack Next.js site (recommended).
Dare you calling the
getRequestContext
CI switched to using
process.envAnd yeah, I'm calling from the edge runtime
Dcould you provide a reproduction of it? if 
process.envgetRequestContextI found the issue, I didn't add
export const runtime = "edge";process.envah! 
Dweird.... what were you trying to access?
Can env variable
Dcould it be that in 
process.envI think what happened is the nextjs dev server loaded
.envprocess.envyeah, it does that 
CI had it in
.envok yeah that makes sense then 
XD@conqr ᯅ I'll improve our error message to remind developers to use the edge runtime , hopefully this can help avoid developers confusion/frustration: https://github.com/cloudflare/next-on-pages/issues/674
, hopefully this can help avoid developers confusion/frustration: https://github.com/cloudflare/next-on-pages/issues/674GitHub
We've recently made the getRequestContext local error more helpful, reminding people to set up the dev platform: next-on-pages/packages/next-on-pages/src/api/getRequestContext.ts Lines 52 to 61...
CStill learning next, so I forget this quite often
Dyeah.... but it's not your fault, it's the problem of Next.js always defaulting to node.js 
Dsome versions back they did allow you to globally specify the edge runtime and have it applied to all your routes!
DI so wish they didn't remove that
CYou don't cache pages when you do SSR? That doesn't make sense to me.
Take for example a site with 10s of thousands of pages...it wouldn't be efficient to statically generate them all, so they would be generated server side and the response would be cached at the edge. This way the next request comes from cache vs generating it again
Take for example a site with 10s of thousands of pages...it wouldn't be efficient to statically generate them all, so they would be generated server side and the response would be cached at the edge. This way the next request comes from cache vs generating it again
CWhat would this be caused by?
CSeems like NPM package htmlrewriter caused the issue :(
Bthe site is deployed, but i am getting 404 error. this is the url: https://testing.client-6ek.pages.dev/
B
Bok, so i deleted the project and did it again, i got the worker limit error
Bbut when i ran this command
it showed me this size. not sure what the issue here is
❯ yarn wrangler deploy .vercel/output/static/_worker.js/index.js --dry-run --compatibility-date 2024-02-22 --name testit showed me this size. not sure what the issue here is
BDid you set the nodjs_compat flag?
Byes
Dyou should use
wrangler pages deploywrangler deployactually i was following your guide. but ok
Bhttps://github.com/cloudflare/next-on-pages/issues/523 this link took me to this page. https://developers.cloudflare.com/workers/platform/limits/#worker-size hence the confusion
Bgot this error.
Dno sorry.... you are going to face the worker-size limit no matter what you use
I was just saying that regarding the last error (the
I was just saying that regarding the last error (the
node:async_hooksdo we tell users to run 
if so please let me know as that needs amending! 
wrangler deploycheck the second link. it tells user to run wrangler deploy with dryrun to check the size
B
Dyes... but that's
but yeah I guess we should have the same information in the
WorkersPagesbut yeah I guess we should have the same information in the
Pagesyes. worker limit for pages is an extremely common issue. i have pinged this chat 100s of times with same request coz it is very arbritary.
Bbut the issue at hand right now, how do i check my worker size for the nextjs site?
Dcan you remove the cli arguments? (since they are not supported in
wrangler pages deploy