Something in your Cloudflare Worker is throwing an error.
Something in your Cloudflare Worker is throwing an error.
IYou'd need to look at your code and see
BTo service bindings is free [1], to Durable Objects is 1 request per session/method call [2] (e.g. when you use await)
Ref:
[1] https://twitter.com/irvinebroque/status/1776331933129253290
[2] https://developers.cloudflare.com/durable-objects/platform/pricing/#billing-metrics
Ref:
[1] https://twitter.com/irvinebroque/status/1776331933129253290
[2] https://developers.cloudflare.com/durable-objects/platform/pricing/#billing-metrics
DHi 
I'm looking for a way to trigger workers in response to Messages from an external Kafka topic or to pipe messages to a Cloudflare Queue. Is there any way to accomplish it without deploying an additional service outside of CF to call an REST endpoint in response to messages?
I'm looking for a way to trigger workers in response to Messages from an external Kafka topic or to pipe messages to a Cloudflare Queue. Is there any way to accomplish it without deploying an additional service outside of CF to call an REST endpoint in response to messages?
MNo, dear, there is no error. The old worker I created three days ago with the same code, I want to copy the new worker again in another account, but it does not delete and gives an error, even though there is no error in the code
because in mode overview worker It outputs correctly
because in mode overview worker It outputs correctly
MWhat do you think is the other problem?
IThat page is only ever shown if there is an error in your Worker. You can try
wrangler tail to view the logs of your Worker, or you can view the logs in the Cloudflare dashboard. I'm guessing when you moved your Worker to the new account you forget to re-add environment variables or bindings from the other account.
FIs it somehow possible to calculate a checksum of a stream? I also upload this stream to R2 but the docs there say checksum will only be returned if it's already provided on upload. So in my eyes it's currently not possible to calculate a checksum of a stream without loading the whole thing into memory?
1Anyone get handshake_failures for some clients when using workers? My users with really old computers can't connect to the worker
1I suspect it's because they don't have TLS 1.2. Probably stuck with 1.1 or 1.0 even
1I guess my question is more, is it possible to allow older SSL/TLS versions? I know you can do this for web pages, but I see no option for workers.
CUsing a custom domain, you should be able to lower the TLS setting but not for the
workers.dev domain1:NotLikeThis:
1Ok I'll have to just issue an update and do that. Thanks
Vanyone?
HCan you try wrapping each I/O call in a
Date.now() log like so? HWon't work for any heavy computation you are doing, but if nothing else you can see if KV/the connection to your DB is the issue
HHello is anyone else getting 202 status code from the mailchannels API, with the email never actually being sent ?
KHTMLRewriter doesn’t actually run until you return the body, so I’m guessing it’s that
Ktransform(…) doesn’t actually call your element handler yet
Ktransform(...) returns a Response, you can await the text() method on that to make HTMLRewriter runKThat consumes the body, so you would want to save that if you want to return the body to the user still
KKTry that
KI ran into the same issue myself when trying to use it to extract css/js/imgs in order to add Link headers onto the response awhile back
KIt wasn't a fun debugging session
Cis transform even awaitable?
Cyou can pr anything to the docs if you think it'd be helpful
Cto me it looks like transform returns a Response, not a promise. so you could even just do
await new HTMLRewriter().on('head meta[property^="og:"]', metadata).transform(source).text();
await new HTMLRewriter().on('head meta[property^="og:"]', metadata).transform(source).text();
YHi, I would like to know if workers could be used to host a complete API (medium size) in several regions of the world? Or are workers intended for simple functions? I'd like to abandon my idea of hosting some of my microservices on k8s clusters because of the price so I've started thinking about the full serverless alternative.
DYou can definitely host big projects on Workers! With the 10MB size limit of paid plans, you can go quite far (libraries and wasm can contribute a lot to your bundle) before needing to shard, but with recent new additions such as the JSRPC api, you can also deploy components of your workers separately and have them interact nicely :meowparty:
You just have to make sure that your code does not rely on node.js apis like fs or similar - though even some of those are available, and many libraries now have an alternative that works, or a way to make them work
You just have to make sure that your code does not rely on node.js apis like fs or similar - though even some of those are available, and many libraries now have an alternative that works, or a way to make them work
Yokay, that sounds really interesting in this case. I started looking into WebSocket with workers and wondered if it wasn't possible to put almost my entire architecture in serverless. I'm going to look into JS RPC, it looks interesting. Thank you very much 
Dyup, websockets also definitely work - I recommend taking a look at hibernated Durable Objects, they have a special API for websockets that makes them super efficient (an instance can do up to 32k connections in theory) and cost-effective!
YThat's exactly what I was looking at. While I'm at it. In my use case, I'm planning to host pre-game lounges with chat, configurable options etc... so several different lounges. Managing the rooms won't be a problem, do we agree? I'd rather ask before jumping in
Ddefinitely possible, simplest is probably to track them in a D1 database
