How are you referencing the secret in your app?
How are you referencing the secret in your app?
GI've just started paying for the pro workers functions and am using the unbound model, though I'm still getting
Exceeded CPU Limits errors (requests range from 50ms to 2000ms), I thought I would just get billed extra for the additional time, but it's breaking my app
HDid you redeploy after changing models? IIRC updates to a project are only applied on new deploys
GRedeployed and cleared cache, still seem to be getting the error
GEven on the deploy specific pages.dev domain, client gets a 503 (Service Unavailable) error and functions logs show another cpu time error
GI still see 0 unbound requests though... all of them are going to 'bundled'
GNevermine I think i was being stupid! I was setting the usage model for pages but needed to set it in the project too!
GSeeing unbound requests going through, but half the time the request just hangs and never resolves 
Ri set it under Github settings > Secrets and variables > Actions
afaik the environment variables are fine, someone told me the problem might have to do with me not having set permissions on 'integration as an app' (not sure what this is but i'll have a look)
as a temporary workaround tho, i'm using CF pages deployment webhooks
afaik the environment variables are fine, someone told me the problem might have to do with me not having set permissions on 'integration as an app' (not sure what this is but i'll have a look)
as a temporary workaround tho, i'm using CF pages deployment webhooks
GQuite disappointed with performance here, requests consistently take multiple seconds and sometimes don't resolve at all (at least after 20+ seconds)
KWhat are you doing in the Worker?
GIt's a very basic DB lookup from post request, have tested database speeds and all sub 10ms
GThat was for the login page, where I'm guessing the hashing stuff took a while... I've tested extensively locally, with wrangler, and on vercel and none of them take over a second ever - I can see in the network tab that the requests are not resolving with cloudflare
GThough, this is a seperate issue, unbounded seems to be working now, the requests are just never hitting the CPU i guess? Not a caching thing as I've been trying all day from the same location
Gbro? I'm expecting my basic requests to resolve in <10 seconds!
GGyup, never even comes close.
GI guess so, I still don't feel like any of the operations I'm doing are compute intensive at all though... Doesn't explain the huge variance in response time though
GAnd to be honest, I'd expect to have decent barebones login support on a platform like this
SEverything we’ve ever done with workers/functions has been incredible fast unless there was an issue in our code, especially when comparing to something like lambda
GI've never had issues with normal workers, I'm unsure if this is a pages function related issue, though I don't think basic password hashing is a very big ask from a full stack platform
SWe’ve been moving some of our platforms to next-on-pages and the response times have far surpassed our expectations, even some of more complicated fetch and store actions are efficient
Ghmm intersting! I've seem some speed issues using next surfacing and seem to be ongoing in the channel... I'm trying out sveltekit here, it's a toy project so nothing critical, though even loading my site on mobile with 4 bars of 4G hung!
SI wanna learn sveltekit eventually and do some testing too
GThough to be fair, it's a fresh deploy and the pages.dev domain seems to be a lot more consistent, I'm hoping it's just some caching stuff
GIt's phenomenal! Fraction of the next bundle size, I'm refactoring our app at work onto it, 80% savings coming from next+vercel!
SDo you have your zone hosted at Cloudflare? I hope they eventually automate best practices for how all their services interact, for example zone caching messing up pages apps
GHmm my nameservers are on CF and all my DNS stuff is managed over there, I thought that took care of all the zone setup?
GSWe’ve been seeing weird caching issues and I’ve been told on here you need to bypass zone caching on a Pages app
SIt isn’t documented yet, I need to put in a feedback request on it. I’m sure behind the scenes the pages team is doing all sorts of magic with other Cloudflare products, we just need to figure out what the best practices are
CThe recommended way is to have your framework minify the output but otherwise you can use your zone level minify settings to do it
DI have a (next on) pages site that uses Firebase + the firebase client sdk for some file serving, and we pay a not pretty amount of money to Google for the bandwidth. Google has special pricing for being accessed by CDN networks but I don't know how to... do that... with my setup.
DI've had a few ideas but they've all hit random dead ends, most notably my idea to just move to R2 lol
DWould love some help brainstorming 
DIs there a way to see the traffic information (unique visitors, total requests, etc...) for the pages.dev site directly for static webpages? Just like we can see in the "Analytics" tab for a standard website handled by Cloudflare. I know it can be done by using custom domains or that there's more detailed information by using the Web Analytics and setting up a beacon, but I'm just looking for a short, no setup needed overview of the traffic.
DSearched around a bit, but I haven't seen that question asked anywhere yet.
HIf you go to Project Settings, hit
Enable Web Analytics, then redeploy your project, it should inject the Web Analytics Beacon for you automagically.
MIs anyone having issues with automatic deployment from GitHub?
MI sent changes in two different repositories and the deploy didn't happen...
BLooks like Github actions / api is down/having trouble right now, are cloudflare pages builds also failing for others for the same reason
Uas long as webhooks and git ops are still working which the status page seems to indicate they are then pages should still be building just fine assuming you have a project that uses the git integration
