I’d assume it uses both as in if local cache doesn’t have it will go to the upper tier (if it isn’t)
I’d assume it uses both as in if local cache doesn’t have it will go to the upper tier (if it isn’t)
UI’m pretty sure that’s how tiered cache works at least is that every datacenter can still cache the value and not just upper tiers
Dahhh, that makes sense! What happens if the local cache is the location of the tiered cache, and the cache API attempts to replace it?
UWould just update its own cache the other tiers would eventually pull from it if needed
Dbut then that sounds like the Cache API does work with tiered caching? 
UNot really. You would have to use fetch to pull from tiered cache, unless the datacenter you are in already has the cache. At least thats how I understand it 
CI assume it depends on the topology as well, if its a location other colos pull from you might update it for tiered cached to use, if its at the bottom/just a local colo it wouldn't ever be used, just a side effect of not storing things twice just for tiered caching/normal cache
NHey! Is there any chance for CloudFlare Workers to support compression algorithms other than just gzip?
We provide a library for developers which has a Rust core compiled to WASM. Currently, gzip is able to compress 3MB WASM module into roughly 1.2MB. We have exhausted all options of compressing the code size on our end (like dependency reduction, unused code elimination, etc). Considering other options, we noticed that LZMA and ZSTD algorithms provide us with 800KB and 770KB WASM modules respectively, which is very promising. Our target is roughly 800KB.
CloudFlare Workers do not allow to run WASM code generated in runtime, so we cannot just unpack the WASM module in runtime ourselves and run it. An ugly alternative to that is including a WASM runtime of our own (like wasm3) and running WASM inside WASM. We would like to avoid that, but this is the only option we have at the moment.
Thus, we are wondering if native support for compressed WASM modules can be added to the platform.
We provide a library for developers which has a Rust core compiled to WASM. Currently, gzip is able to compress 3MB WASM module into roughly 1.2MB. We have exhausted all options of compressing the code size on our end (like dependency reduction, unused code elimination, etc). Considering other options, we noticed that LZMA and ZSTD algorithms provide us with 800KB and 770KB WASM modules respectively, which is very promising. Our target is roughly 800KB.
CloudFlare Workers do not allow to run WASM code generated in runtime, so we cannot just unpack the WASM module in runtime ourselves and run it. An ugly alternative to that is including a WASM runtime of our own (like wasm3) and running WASM inside WASM. We would like to avoid that, but this is the only option we have at the moment.
Thus, we are wondering if native support for compressed WASM modules can be added to the platform.
ZWhat's preventing you from just using the 1.2MB WASM module?
NFree tier worker limit size is 1MB. Since we are a library we want (total library size + total user code size) to be less than 1MB
ZAh, I see. I think your best bet is to create a discussion post on the workerd (workers runtime) repo: https://github.com/cloudflare/workerd. Unless you are an enterprise customer, in which case reaching out to support might work better.
Dstupid question, does workerd allow you to load WASM from a URL?
Dor does it have to be bundled?
Dhave you tried compiling with
-Os?
NYes, but the size reduction is only a couple percent
NIt has to be bundled
NBefore I create a discussion at workerd repo, does anybody else have WASM code size issues? The more use cases we have, the better
Dis there no way to access
envis it a bad idea to just do
process.env.API_KEY = env.API_KEY in my fetch handlerDWhy does this tutorial use the fetch adapter? https://developers.cloudflare.com/workers/tutorials/openai-function-calls-workers/#2-make-a-request-to-openai
Doesn't look like it's needed. https://developers.cloudflare.com/ai-gateway/tutorials/deploy-aig-worker/
Doesn't look like it's needed. https://developers.cloudflare.com/ai-gateway/tutorials/deploy-aig-worker/
K1st one is OpenAI SDK v3, 2nd is v4
KGiven that the first tutorial doesn’t pin to the major version, I guess it’s broken since it’ll install v4. cc @rickyrobinett
RHmm... yes, good catch. Let me work on a PR to update / fix.
Dis it possible to use local node modules in a worker (with wrangler)? i'm setting up a monorepo and i need some types from a local package in my worker
DAhhh, thanks! Totally missed that.
HIs it to use cloudflare's workers for security research purposes such as brute forcing web hosts, fingerprinting, crawling, etc ?
Hall legally ofcourse (bug bounty etc)
UCloudflare
This page is for those who are interested in our Self-Serve Subscription Agreement.
Hthanks for this! i don't see anything in the Restrictions part that denies this, would love to get confirmation tho as i may have misread it
DI'm getting really weird console.debug results today.
DSo this is my code... and this is my log:
Dwhat on earth is going on?
CThere is a bug open related to log messages with
at in them. I know there is a fix PR up but don't remember the numberUGitHub
Fixes #4668.
What this PR solves / how to test:
Previously, logs containing at were always treated as stack trace call sites requiring source mapping. This change updates the call site detection t...
What this PR solves / how to test:
Previously, logs containing at were always treated as stack trace call sites requiring source mapping. This change updates the call site detection t...
D@Cyb3r-Jok3 thank you! Thought I was going crazy.
Is there a temp workaround?
Is there a temp workaround?
Udo a replaceAll on console.log for "at" and replace with "@" or something 
Dlol
Dhmm, downgrade to
3.19.0 is my only safe option right?DI had a panic moment thinking I had incorrectly stored JSON in >100,000 files.
NHeyo guys!
I'm developing/testing a worker project using the
Hope someone could help me figure it this out. Thanks in advance.
I'm developing/testing a worker project using the
wrangler dev --remoteHope someone could help me figure it this out. Thanks in advance.
RApi tokens are still not associated with specific pages or workers projects. This is a rather large security problem when using gha or other build systems, because it allows any developer to deploy on any wrangler project with or without a branch tag.
RI want to do some file operations using ffmpeg as soon as a file (it should only trigger when the uploaded file is mp4), is uploaded to my storage bucket, and then want to upload the resulting files to another R2 storage bucket using workers
I have gone through the doc, but can't figure out how to do it. can someody please help
I have gone through the doc, but can't figure out how to do it. can someody please help
RI am migrating a script from aws lambda to cloudflare, it was effortless in aws, but cannot figure it out here
KWorkers are just JS and can’t spawn child processes, ffmpeg isn’t going to work
Rcan't we use ffmpeg wasm node module?
KIt's designed for use in a browser, and I'm not aware of anyone who has got it working on Workers, but you might be able to.
LI'm trying to make a fetch request from my worker handler using fetch api to ebay, and the auth headers im passing are not being used?
KWhat's the code snippet and what makes you think they're not being used?
LI have the same setup in postman and it works, but with workers it doesnt
Land the issue is that the access token is missing, i.e authorization header
SMake sure this isn't applying to your situation: https://stackoverflow.com/questions/71441897/cloudflare-worker-redirect-stripping-auth-headers
Stack Overflow
I set up a Cloudflare worker to redirect to our API gateway since we don't have control of the DNS and can't just set up a CNAME. The redirect works and it passes along the body and all the headers
