added some asynclocalstorage stuff to my nextjs app, and it gets bundled into the frontend code???
added some asynclocalstorage stuff to my nextjs app, and it gets bundled into the frontend code???
LHey folks, I'm having issues caching fetch requests within Server Components when we deploy to CF Pages. The Next docs mention time-based revalidation, e. g.
This works as expected in dev - the server will only re-fetch 60 seconds after the last fetch occurred. Once we deploy though, each request to the page makes a fetch every time.
Is this where the KV bindings come into play? I don't believe my team set that up.
const users = await fetch(USERS_URL, { next: { revalidate: 60 } }).then(…)This works as expected in dev - the server will only re-fetch 60 seconds after the last fetch occurred. Once we deploy though, each request to the page makes a fetch every time.
Is this where the KV bindings come into play? I don't believe my team set that up.
Rhi team, for better reach reposting an open question here Worker using multiple modules in a single bundle
Basically seeing warnings as ✘ [ERROR] workerd/jsg/_virtual_includes/jsg/workerd/jsg/modules.h:443: warning: NOSENTRY Worker using multiple modules in a single bundle.
Basically seeing warnings as ✘ [ERROR] workerd/jsg/_virtual_includes/jsg/workerd/jsg/modules.h:443: warning: NOSENTRY Worker using multiple modules in a single bundle.
Jcan you open an issue on the next-on-pages github repo please
Jare you using it only in server components? it'll be how nextjs builds your app so im not sure we can do anything about that
LHey @James thanks for getting back to me. I tested deploying a hello world next-on-pages project to see if the same issue occurred and it didn't! The caching works great. So it is definitely something wrong with my company's project setup. One thing I noticed is that we don't have a wrangler.toml -though it's unclear to me how we're deploying without it. Thanks again!
Ji see. that's interesting. you shouldnt need a wrangler.toml for fetch caching to work though...
LO rly. I'll have a look and report back once they grant me access to the account, probably tomorrow
Nthe issue was a bit different
Nwell
Nthe code for it was called
Nbecause
Ni was importing a package
Nthat was importing it somewhere deep
Ni just refactored the pkg
Nso i only import the files i actually need
Jah makes sense
Bhi, how did you solve calling D1 binding? I get an error.
Short explanation including minimum code to reproduce the error: How can I access `env.DB` in my Pages/Worker on production?
Short explanation including minimum code to reproduce the error: How can I access `env.DB` in my Pages/Worker on production?
AI am getting the error
I can reproduce this locally when running npx @Cloudflare/next-on-pages@1 and then using wrangler to do a preview, otherwise this works on local.
The error I was getting before reproducing on local (in my dev cloudflare url) was
The script will never generate a response when trying to use fetch() in my generateMetaData function in NextJS 14. I've seen a lot of similar bugs reported on this but haven't been able to find a solution. Is anyone familiar with this?I can reproduce this locally when running npx @Cloudflare/next-on-pages@1 and then using wrangler to do a preview, otherwise this works on local.
The error I was getting before reproducing on local (in my dev cloudflare url) was
A hanging Promise was canceled. This happens when the worker runtime is waiting for a Promise from JavaScript to resolve, but has detected that the Promise cannot possibly ever resolve because all code and events related to the Promise's I/O context have already finished.
NHey, it seem like I have this problem again, after deploying it seem that the website load infinitely
Nmaybe a cache problem because after some minutes after deployment the app works normally
Nor if I use a different browser it works
Hcan I somehow see the logs of a running deployment?
IYou can tail them with Wrangler, or tail them in the web UI (Deployments > View Details (on your production deployment) > Functions > Begin Log Stream.
INote that in both cases you won't see prior logs, only things that happen after you start tailing.
HThank you ill try that, since I have an issue that doesn't happen in the local preview and only on cloudflare.
TI'm facing this same issue: where should I put the
_headers file so I don't need to manually edit ./.vercel/output/static/_headers after every build?
Bi've been encountering some weird issue, has anyone else had anything similar?
whenever im refreshing like multiple times repeatedly, somewere between 5th or 10th refresh results in a 500 internal server error and the logs say:
whenever im refreshing like multiple times repeatedly, somewere between 5th or 10th refresh results in a 500 internal server error and the logs say:
BAlthough the main page that's displayed is a server component, some child compnoents might be client, but thats about it
Bwhy you use 14.1 ?
GWhen using zero-trust and next-on-pages I don't see the CF_Authorization cookie is this expected?
Gn/m it's in the header but it's not being parsed
Bjust asking no need to be confrontive
Bbut try 14.2.3, it seems to be working
IThat was not a rude question at all; they were legitimately asking why you needed a particular version. That stuff matters when troubleshooting. Please don't bring a hostile attitude like that in here and expect help.
IThis is often an issue with server components + suspense I think. Make sure you really follow the rules of how you can use server components within client components (can only be passed as children) and mark client components actually as client components, even though the parent is a client component.
Bi have that issue on a server component page that isnt even async, but has some client components beneath
IFor people having issues with "The script will never generate a response", I ran into this too with Next.js + Storyblok & Server Components. After following the rabbit hole of the source code of Storyblok I noticed that it stores the client in a globally scoped variable. I had a suspicion that that might have something to do with it, since it might re-use these in subsequent requests, which workers doesn't like? I forced the re-initalization of the Storyblok client on every request we do to it, which seems to have solved the problem for us.
HHello, I am still struggling to understand on how to run a next-on-pages dev environment.
How do I run it like
If I run
If I run
How do I run it like
npx next dev would (with instant updates as I change the code, instead of needing to compile everything all the time).If I run
wrangler pages devIf I run
wrangler dev
Mhas anyone had a chance to try next15? curious how it goes, going to give it a shot a lil later hopefully
Jthank you for this!
RI have a weird build error. I have the .node-version, .tool-versions and mise.yaml config files to tell the builder to build using 20.5.1. I see asdf installing 20.5.1 correctly, but then during the build it all of a sudden says it found 18.17 to build with. My site will not build with 18.x so I can no longer use the cloudflare builder. It works fine building it locally and uploading it using wrangler pages deploy and the function works correctly (ie. SSR using nexjs with a next-on-pages vercel build.)
RIt will also work using github actions, of course, but then it's the role problem I mentioned in the other channel making this really suboptimal for large enterprises.
Bdevelop with
By default, the entry point is the build directory, so add the following to your
Note:
next devwrangler pages devBy default, the entry point is the build directory, so add the following to your
wrangler.tomlNote:
wrangler devpages feature (which is a subset of workers). So, wrangler devwrangler pages dev
HThank you for clearing that up!
NHi, there is a problem with fetch request with "force-cache" that are loading forever randomly when its a file bigger than 1Mb in server components
