Not every fetch call of course, but the whole function that calls fetch() must run inside a dressed
Not every fetch call of course, but the whole function that calls fetch() must run inside a dressed up worker, that counts as one.
Kthank you, I didn't know about that
RI am getting an error with dynamic routes that generateStaticParams… I believe it was working before next 15 … So you’re saying that generateStaticParams should not throw an error?
Specifically starts with next@15.0.0-canary.175 and up.
Edge + generateStaticParams not allowed.
Specifically starts with next@15.0.0-canary.175 and up.
Edge + generateStaticParams not allowed.
BWill there be a migration guide from next on pages to opennext?
JNext.js 15 hasn't been tested with it so may have issues
BThanks!
NHi everyone. So I have a NextJS app (<15.0.0) and was trying to deploy to pages. My has both server and client functions and is still relatively small. While trying to deploy I see the files in /.next/cache are larger than the 25mb limit. Tried to troubleshoot to optimize where I can (don't have many static assets and they're small). Build direct was set to /.next and npm run build as the command. Am I doing something wrong or is it plainly a matter of refactoring everything to use Cloudflare workers?
NI'm still in early stages of Dev and while I have been deploying onto my local server and serving it up that way, thought it would be convenient to test out Pages instead
BSubject: Build Failure Due to Bundle Size Limit Exceeded
Hi everyone,
I'm experiencing an issue while deploying my Next.js app on Cloudflare Pages. The build fails with this error:
Earlier in the logs, it mentions:
It seems my bundle size exceeds the 25 MiB limit. Here's a snippet from the build log:
I believe the large number of modules and functions is causing the issue. Does anyone have advice on how to reduce the bundle size or handle this situation? Or how to check bundle size?
Thanks in advance!
Hi everyone,
I'm experiencing an issue while deploying my Next.js app on Cloudflare Pages. The build fails with this error:
Earlier in the logs, it mentions:
It seems my bundle size exceeds the 25 MiB limit. Here's a snippet from the build log:
I believe the large number of modules and functions is causing the issue. Does anyone have advice on how to reduce the bundle size or handle this situation? Or how to check bundle size?
Thanks in advance!
MDoes the new worker's framework platform have support for cron ? If i have a nextjs app deployed can i add a scheduled trigger to the function ?
Does anyone know how to run a cronjob with open-next cloudflare, nextjs 15 ?
Does anyone know how to run a cronjob with open-next cloudflare, nextjs 15 ?
HWorkers with Assets are also able to export a Scheduled Handler for Crons. As for whether
open-next
CHey guys
Was working on nextjs on CF pages and I always use binding on the API side now I wanted to use on server actions and it is saying
Does anyone knows how to use since I kinda need the bindings to work.
Was working on nextjs on CF pages and I always use binding on the API side now I wanted to use on server actions and it is saying
only async function are allowed to be used in use server. File Does anyone knows how to use since I kinda need the bindings to work.
BGenerated Pages Functions bundle size (26282201) is over the limit of 25.0 MiB
Failed: generating Pages Functions failed. Check the logs above for more information. If this continues for an unknown reason, contact support: https://cfl.re/3WgEyrH
How to check bundle side? How to understand which file is causing this?
Failed: generating Pages Functions failed. Check the logs above for more information. If this continues for an unknown reason, contact support: https://cfl.re/3WgEyrH
How to check bundle side? How to understand which file is causing this?
DRemove the
export const runtime = "edge"
CIt is needed for any server side code. The issue is with the context which is a sync function and all server actions functions must be async
DNo it's not, the error is clearly highlighting that line. I use server actions too and can confirm that the runtime option is unavailable/not needed. Do ensure that you have it in your root layout or in single pages, tho
DAlso, about the context, you're calling
getRequestContext()Heya, can you open an issue on the opennext-cloudflare repo for cron support
JThe problem is that you're exporting the runtime from an action file. Actions don't need that. You only need to export the runtime from either a layout or a page file
MSure 
DHi everyone, my CF builds for next.js just stopped with a message
and nothing has changed in the config, I am not using the not found page it is auto-generated. Has anyone experienced similar?
UPDATE: Just fixed by itself after some time, really strange
and nothing has changed in the config, I am not using the not found page it is auto-generated. Has anyone experienced similar?
UPDATE: Just fixed by itself after some time, really strange
JSounds like a Next.js issue
PI'm aware that the edge runtime builds as required/requested or something like that but I was wondering request times should by in the 6-8s realm?
JSounds odd. Maybe it's network requests inside the app or very large component trees?
PI do have an API it calls that does roughly take 6s but I have a skeleton UI that should display whilst it's doing so. It functions as intended with the skeleton UI when I'm testing locally.
Pnew to next on cloudflare. I just pushed and built my prototype project without adding
export const runtime = "edge"I also just saw in my logs that all pages have been built as edge function routes uncluding those, where I didn't delcare edge runtime at the top. I'm confused 
DYou need to add it only to the layout.tsx file
PNormally if there’s issues with the runtime declarations, it’ll show the files/routes in the build log then fail.
Pthx! I didn't come to a clear conclusion with the current documentation, especially when I reached the point in Trouble Shooting that said one has to make a custom error page with runtime = edge
Pok so I can expect to run into problems immediately . Thanks!
TGetting this error when I try to deploy a static nextjs application
I have output: "export", on next.config.mjs file
I have output: "export", on next.config.mjs file
JIt looks like you're not using next-on-pages so have the wrong output directory or wrong build command set
PIs there a way to access bindings in a client component? I have a hono worker, and I want to use hono's client rpc features inside a client component
HIf you mean a client-side component, then no
KI'm developing a Nextjs application that I basically will needs the seo features because all my backend is hosted on a external api, so sould I host this project on cloudflare pages or workers?
Kthis is a personal project so I would like to stay in free tier, so what shoud I chose ?
MI'm using nextjs with workers on static assets and I'm noticing that even an emptyish project is causing worker CPU usage to spike to several hundred MS
From my standpoint, every page that I'm deploying has "use client" and I'm not doing any API calls. Frankly, everything should be static and so Idk what is even calling workers.
Is there a way to know what endpoint causes the CPU spikes? Or even know what endpoint is considered "static" and what endpoint is actually a worker running
From my standpoint, every page that I'm deploying has "use client" and I'm not doing any API calls. Frankly, everything should be static and so Idk what is even calling workers.
Is there a way to know what endpoint causes the CPU spikes? Or even know what endpoint is considered "static" and what endpoint is actually a worker running
Nif you do
next build
MThanks for the reply. So all my pages are static and the only dynamic API call is never made. The API call is hello world anyways
Any ideas what could be hitting my workers and causing the CPU? IDK why worker requests are even being made in the first place. Tails shows requests as I navigate between the static pages i.e. "/dashboard" and it's not clear why
Any ideas what could be hitting my workers and causing the CPU? IDK why worker requests are even being made in the first place. Tails shows requests as I navigate between the static pages i.e. "/dashboard" and it's not clear why
HJThat's because your requests go through the routing system
MI could do this, but I want to eventually use the nextjs api routes, and so doing that would require me to have two seperate deployments.
MI swear I've read https://developers.cloudflare.com/workers/static-assets/routing/ several times and also thrown it into perplexity to see if it'll help explain my situation but I feel like the actual experience is different from the documentation.
From the documentation, it seems like requests to static pages (i.e. "dashboard" should not be hitting a worker but tails and baselime show that it does). In addition i'm suspecting that for whatever reason that this sometimes causes several hundred ms of CPU usage
From the documentation, it seems like requests to static pages (i.e. "dashboard" should not be hitting a worker but tails and baselime show that it does). In addition i'm suspecting that for whatever reason that this sometimes causes several hundred ms of CPU usage
Cloudflare Docs
How Workers assets' routing and its configuration works.
MThis is a screenshot from the documentation, and from my exerience, visiting "blog" would actually hit a worker. Since this example is seemingly identical to me visiting "dashboard" that is shown to be static on deployment (I'm using nextjs on workers)
JThose are the Workers docs not the Pages docs - next-on-pages is for Pages. The Cloudflare Pages docs has a section for deploying Next.js which touches on this a bit.
https://developers.cloudflare.com/pages/framework-guides/nextjs/ssr/
https://developers.cloudflare.com/pages/framework-guides/nextjs/ssr/
MOkay, so i should ask the question inside of the workers discussion instead. Thanks
JThere's a lot of work going on to re-use parts of the AWS adapter and share the same architecture between the two. This is mostly happening in the experimental branch but also requires a lot of changes in the AWS adaptor to make re-use easier. So there's still things happening, and you can try them out with
https://pkg.pr.new/@opennextjs/cloudflare@experimental. I would recommend using the OpenNext discord for discussions about that adaptor.