And if you're calling out to stuff in C++ land like hashing I expect it to use multiple cores?
And if you're calling out to stuff in C++ land like hashing I expect it to use multiple cores?
II might be wrong though 
HOh, you meant CPU Time in general. I was referring to specifically on CF, where you only get single thread(unless you call out to a DO or smthn)
IOn multi-processor machines, a computer program can use two or more CPUs for processing using parallel processing scheduling. In such situations, the notion of total CPU time is used, which is the sum of CPU time consumed by all of the CPUs utilized by the computer program.
IYeah your isolate gets single-thread
IBut I expect things in
workerdI don't know though.
HI would hope that you aren't billed for multicore ops because they are an implementation detail?
IWell you're billed for the amount of CPU time you use.
HIt would make it pretty difficult to figure out what your pricing might be
IIt uses the same CPU time on a single-core vs multi-core.
IJust multi-core uses less wall time because it's usually faster.
IHOh, I thought you said it multiplies it by number of cores?
INo, so if you cap 4 cores at 100%, then in 1 wall-clock second you'll consume 4 cpu-time seconds.
IBut if you wanted to do the same task on a single core, then you'd cap that single core at 100% for 4 wall-clock seconds and consume the same 4 cpu-time seconds (assuming the scaling was linear)
HI was also just referring to while you can pretty easily figure out single core pricing, by timing total duration length, then subracting I/O, it would be a bit more complex for multicore
IYeah... I'm not sure.
HSince while a request may take 5 cores 10 seconds to process, you don't know if it is utilizing all 5 cores the entire time
ISorry, I kind of went on a tangent - this isn't related to your use-case at all.
IYou just need to look at the CPU Time measurements on the Cloudflare dashboard. On the free plan you get 10ms of CPU time, on the paid plan you get up to 30s.
JFun fact: you could create next-on-pages projects without needing to do additional setup for a long time. The pages project create command accepts a compatibility flags argument which automatically sets the flag in your new project
Lhuh? that doesn’t answer the question at all.
Jkind of. you said you can't push a next-on-pages project without having to set it up in the dashboard, but it's always been possible to do that have the compat flag added when creating the project. sorry, not a helpful response to you i know.
Janyway, looking at the changelog for wrangler there's been nothing to change the new behaviour
Sis there a entry point for installing an application-level instance of asynclocalstorage?
Sfor next-on-pages
Si could wrap all my /api/ handlers (and server components if that works)
Sbut if there's one place, i'd rather do that
JThere is not. Are you able to share an example use case you're thinking of for being able to do so?
DHi all, if I want to use turso db with a full stack nextjs app deployed on pages, do I need to initialize the db on every next API endpoints? Or can I just create a file and export the db and import to endpoints could it work?
BHi all, Cloudflare page not working NextJS14 Node v18.17.1
HWhat is your deploy command?
HShould be something like
wrangler pages deploy .vercel/output/staticJyeah that looks like using the wrong output directory
BI changed the build configurations
DHi, there can I ask how did managed using turso with nextjs?
Mok so got allot of stuff working but i am trying to debug my pages but the errros keep showing up like this
is there any way for it to show the actual data it was passed instead of object ?
is there any way for it to show the actual data it was passed instead of object ?
Mmigh hvae figured it out
DAnyone used turso with nextjs?
DIn the web UI there only D1 biding integration
NHey, I just found out that nextjs default image resizing doesn't work with next on pages, is this intended?
NThe returned image is always at its original size
Jyes, it's not possible for us to do image resizing on the server in pages until the images team adds support to do that
Bsrc/server/utils.ts
is this fine?
is this fine?
getRequestContext() and process.env is still very confusing. sometimes they work in local, sometimes they dont work in prod, so i made this utility functionBright now process.env works on local, but getRequestContext doesn't, when it always did in the past, very weird
Dboth
(and you should ideally always use
If either doesn't work in some scenario, could you please open an issue (ideally with a minimal reproduction): https://github.com/cloudflare/next-on-pages/issues/new?assignees=&labels=bug&projects=&template=bug-report.yml&title=%5B%F0%9F%90%9B+Bug%5D%3A+
so that I can take a look and address the bug?
(most definitely you should not have to define utility functions like the one you shared!
)
getRequestContext() and process.env should with next dev, wrangler pages dev(and you should ideally always use
getRequestContext())If either doesn't work in some scenario, could you please open an issue (ideally with a minimal reproduction): https://github.com/cloudflare/next-on-pages/issues/new?assignees=&labels=bug&projects=&template=bug-report.yml&title=%5B%F0%9F%90%9B+Bug%5D%3A+
so that I can take a look and address the bug?
(most definitely you should not have to define utility functions like the one you shared!
)Bill make a repro once i can
Bbut the idea is
server component -> server action -> server utility - (getRequestContext() doesn't work here)
server component -> server action -> server utility - (getRequestContext() doesn't work here)