just service binding, you can also invoke it the HTTP way (and not RPC)
just service binding, you can also invoke it the HTTP way (and not RPC)
森the rest i guess is implementation details, i don't think there's a guarantee that the two services would be in the same workspace or even server at all; and in reality, most cross application boundary communications uses some form of RPC anyways, where they are on the same machine or not
Dmy worker sends requests to a specific server on the cloudflare network. it's been working for months and it was working this morning but i just started receiving this error out of nowhere (no deployed code changes in the last few weeks on my worker).
can an employee confirm whether a change was made recently to block access to certain cloudflare ips from workers?
can an employee confirm whether a change was made recently to block access to certain cloudflare ips from workers?
Lbut that's just a fake HTTP, it's still using same magic under the hood
I^ RPC is invariant of transport
IYou can RPC over a unix socket, over IPC boundaries, or over a network connection
IOr even RPC in shared memory as is the case with some apps (see Tauri for example)
森... sorry can i vent for a bit?
森or if someone could let me know is there a better solution
森how exactly do you detect if the worker is in dev mode or in production mode?
森it looks like the official solution is to add [env.dev] and add something like ENVIRONMENT = "dev"
森if that's the case....
森then why is the majority of keys "NOT INHERITABLE"
森why?
if it works both in staging / prod and dev, why should i have to copy every single vars under env.dev ?
if it works both in staging / prod and dev, why should i have to copy every single vars under env.dev ?
森i have like an array of vars that define a certain default output
森does that mean i have to copy the whole thing under [env.dev]?
森ok, if that's the case, and I decided that I would change the non-dev vars just because i want a different set of output, does that mean i need to remember to also change env.dev too? I already can see myself debugging only to realize i forgot to change the [env.dev] too because the non dev vars were changed...
森....omg
森now running
wrangler dev --env dev causes service binding to freak out and screams Cannot access '{functionName}' as we couldn't find a 'wrangler dev' session for service "{service}" to proxy to.森I found the solution
森kind of hacky, not intended; not sure if the wrangler dev session thing is a bug or not, don't care to reproduce it on a more basic setup.
but if brave traveler if you stumble upon this message; instead of using [env], use
(note: this would also cause 'but this works on my machine problem', best to document this and note that include
but if brave traveler if you stumble upon this message; instead of using [env], use
.dev.vars.dev.vars(note: this would also cause 'but this works on my machine problem', best to document this and note that include
.dev.varsnot sure why that cf-workers has "node:process" included, but never
process.env.NODE_ENV === 'development' which should be a simple implementationIBecause only recently 
node:processNODE_ENV to determine whether it's running in Workers or Node would break 
森oh yeh, great argument!
Well, I still think people would benefit a lot if it's simple to detect what env they are in via something similar, perhaps
Since a lot of frameworks does things similar to this (for example
Well, I still think people would benefit a lot if it's simple to detect what env they are in via something similar, perhaps
process.env.WORKER_ENV or a standalone global variable since env is per fetch and not global, so it be also nice to check this can be checked globally sometimes, I guess?Since a lot of frameworks does things similar to this (for example
app.Environment.IsDevelopment() in ASP like webservers), I was sort of bummed out that the official way to detect development/staging/prod is via [env], which comes with a whole baggage of concerns like non inherit keys; and also aformentioned bug about not find the session (I'll write up a bug report for it :/)
NGot the same issue
AI am, its working pretty good. The durable object isn't that big of a cost factor if you use it at scale. Maximum like $10 dollars a month orso? That you use it like 24/7. As currently there are not session costs?
T
Is there known ongoing issues with Workers Builds? Two of my Workers have been timing out at cloing repo step for the past >3 daysI was in the Builds cache beta and I'm using watch paths so not sure if that's related
TI've tried reconnecting GitHub integration and still same issue
WCan you send me a build ID?
POut of curiosity, is your ISP Deutsche Telekom by any chance? Or for the users who report having slowness with your worker endpoints
Pthose high connect times make me think it's some weird routing issue and not necessarily Workers
PAnyone knows why i have logpush = true and observability on my wranglet toml but still when i deploy i disables it?
PAnd I have to enable it via the dashboard
PIm using env and i want to enable it just for mu env.prod environment
PHow should i define it in my wrangler?
Wwhat's your wrangler.toml look like?
PIt starts with zone_id, account id, compatibility date and node_compat =true, then i have my envs
PBut i dont know where to add the observability part and the logpush
Chey there any ideea about a wasm pdf generator lib (best from html)?
KCloudflare远程隧道管理是不是不支持18.4.6版本的Ubuntu？
KDoes Cloudflare remote tunnel management not support Ubuntu version 18.4.6?
BI use Browser Rendering to generate PDFs from HTML. Don't use it at a large scale yet but works fine for me so far.
Cloudflare Docs
As seen in the Getting Started guide, Browser Rendering can be used to generate screenshots for any given URL. Alongside screenshots, you can also generate full PDF documents for a given webpage, and can also provide the webpage markup and style ourselves.
Awhats the best way to figure out where my cpu usage is? I'm suddenly running into cpu limits for queues (even with a batch size of 1) and ive made all the code changes i can think of to reduce it and it has made no difference
TDid you get a chance to take a look? It seems I managed to get one Worker to pass but the other one still times out cloning the repo
IHey guys! Any chance I can make an authenticated request fetching an image (from google drive directly through api with bearer access token in the auth header) (not google cloud!) that I can run through cloudflare images to resize it on the fly? saw in the docs (https://developers.cloudflare.com/images/transform-images/control-origin-access/#authenticated-origin) that it is possible but cant get it working
Ehi. I know static assets is in beta
Ebut im not sure how the technicality of it works
Edoes it live with the worker globally? or it works like KV where its just stored i n1 location and cached
