doing a service binding to worker B with function textLableMatch, and getting the error when doing a

doing a service binding to worker B with function textLableMatch, and getting the error when doing a fetch to worker A that is calling that function from worker B

Corsix👾OP•9/26/24, 4:40 PM

is what im doing

Corsix👾OP•9/26/24, 4:41 PM

doing a ai worker on worker B to match text to lables as micro servie to worker A

CCorsix👾doing a service binding to worker B with function textLableMatch, and getting th...

ac•9/26/24, 4:55 PM

they were pointing out that you misspelled the word "label" and wondering if that is the issue.

ac•9/26/24, 4:56 PM

otherwise, I would make sure that you have a WorkerEntrypoint class that has that method on it.

dexter•9/26/24, 4:57 PM

does anyone know how to fix the error:

✘ [ERROR] Wrangler cannot validate that your Worker name matches what is expected by the build system. Please retry the build.

✘ [ERROR] Wrangler cannot validate that your Worker name matches what is expected by the build system. Please retry the build.

when deploying a worker via git? (Worker Builds)
the name is obviously the same in the dashboard and wrangler.toml

Keebs•9/26/24, 5:05 PM

any reason the AI get started demo

      const response = await env.AI.run("@cf/meta/llama-3.1-8b-instruct", {
        prompt: "What is the origin of the phrase Hello, World",
      });

      const response = await env.AI.run("@cf/meta/llama-3.1-8b-instruct", {
        prompt: "What is the origin of the phrase Hello, World",
      });

takes 15 seconds to respond from local dev?

Keebs•9/26/24, 5:06 PM

Response time: 25032 milliseconds

Keebs•9/26/24, 5:06 PM

25 seconds

adeg•9/26/24, 5:07 PM

has anyone got the source map to work in the cf worker? I couldn’t get it to work

Leon338•9/26/24, 5:37 PM

What's that speculation request I'm seeing in a simple Worker returning some html from kv? I feel like this is new and it adds one more Round Trip

Edit: And how do i disable it?

LLeon338 What's that speculation request I'm seeing in a simple Worker returning some htm...

texan•9/26/24, 5:46 PM

This is speedbrain (new blob post about it yesterday I think). It's turned on by default, but you can disable in your account by selecting a domain, then Speed > Optimization > Content Optimization > Disable Speed Brain

Ttexan This is speedbrain (new blob post about it yesterday I think). It's turned on by...

Leon338•9/26/24, 5:48 PM

Ah, okay, thank you! I have one more question. It seems like the initial load time of the Worker has slowed down as well. I don't remember it being around 120ms six months ago. I know it's not a huge difference, but it's just something I've noticed. It's an Worker to only test speeds and I tried it multiple times ofc. Btw the KV Part takes 5ms its also messured

LLeon338 Ah, okay, thank you! I have one more question. It seems like the initial load ti...

texan•9/26/24, 5:50 PM

I might be the wrong guy to ask about that

though you might benefit from smart placement if you're not already using that:
https://developers.cloudflare.com/workers/configuration/smart-placement/

Ttexan I might be the wrong guy to ask about that 😅 though you might benefit from sma...

Leon338•9/26/24, 5:52 PM

Ah, okay. No, that wouldn't help since I'm not calling any backend or external services outside of Cloudflare.

taczzy•9/26/24, 5:52 PM

I've been following along with the commits in the workers-sdk repo for the new experimental-assets. Played around with it a bit in a small repo. It is so close to providing the perfect developer experience for building sites on workers. The main feature I would love to see is support for running wrangler dev and having vite live reload the browser whenever I make changes to my HTML, CSS, JS that runs through the worker. You would get the best of both worlds, (quick update in the browser when you modify your code/you are actually running your app via the worker so its more true to how it will work in production).

You can get live-reload partially working with workers + assets, but really what I want, and I'm sure many others want is full integration with wrangler dev and vite to do real hot module reloading. It feels like this is possible with the new Environments API in the latest Vite beta but I havent dug in.

Is the team that is working on the new experimental assets thinking about any vite integration that would support HMR/live reload with workers? If not, what's the best way to pass along that use-case, open an issue on the workers-sdk repo?

taczzy•9/26/24, 5:53 PM

Also was sad to see the --remote flag will not work with the new expiremental assets feature. When I'm running wrangler dev, I like using the --remote flag so I can get access and interactive with my preview KV instance, was hoping to have the same ability when using experimental assets.

Mr. Technician•9/26/24, 6:00 PM

do I need to do anything special to enable compression on my websockets connection?

Mr. Technician•9/26/24, 6:00 PM

I'm sending some very large messages 500Kb+

Ohad•9/26/24, 6:24 PM

congrats on the new UI and all, but where did my cron triggers logs disappeared to?

Walshy•9/26/24, 6:25 PM

Settings -> Trigger Events -> View events

WWalshy Settings -> Trigger Events -> View events

Ohad•9/26/24, 6:25 PM

not trigger, view events

WWalshy Settings -> Trigger Events -> View events

Ohad•9/26/24, 6:26 PM

used to have logs of previous events

Walshy•9/26/24, 6:26 PM

the cron events that used to be there was moved to what i mentioned - there was no logs in those though - for real time logsclick "Live" in that UI

WWalshy the cron events that used to be there was moved to what i mentioned - there was ...

Ohad•9/26/24, 6:26 PM

yes there were logs of them, you could see each event status

WWalshy Settings -> Trigger Events -> View events

Walshy•9/26/24, 6:27 PM

that's the events I mentioned here

Ttaczzy Also was sad to see the --remote flag will not work with the new expiremental as...

texan•9/26/24, 6:27 PM

There are plans to support this, but we can't share a timeline yet

WWalshy that's the events I mentioned here

Ohad•9/26/24, 6:27 PM

ok need a magnifying glass to find it, please tell UI team to bring back to normal place like before

Ohad•9/26/24, 6:28 PM

settings to view logs != intuative

Raymond Camden•9/26/24, 6:37 PM

Hi - sorry to repeat

would anyone have a clue as to why a worker with a scheduled event that uses fetch() gets an error in the network code. but if i hit my worker via http, i can then make that remote API call just fine. both scheduledscheduled and

fetch

fetch

run the same function, getIncidentsgetIncidents, which makes a fetch call to the API. exact same code works in one way, not the other

adeg•9/26/24, 7:52 PM

Sorry to repeat. has anyone got the source map to work in the cf worker? I couldn’t get it to work

DanTheGoodman•9/26/24, 7:52 PM

are workers having downtime right now? 2 major services of ours have just gone mostly offline

DanTheGoodman•9/26/24, 7:53 PM

getting all 500s back from them, no changes to code

Walshy•9/26/24, 7:53 PM

Workers should be fine - got anymore info?

DanTheGoodman•9/26/24, 7:53 PM

yeah sec

Walshy•9/26/24, 8:01 PM

I'm looking into it

WWalshy Such a weird situation haha, for now you can disable smart placement @Brett Will...

Brett Willis•9/26/24, 8:01 PM

@Walshy | Deploying sorry for the ping. Just in the last half hour or so it seems like this issue has reawakened.

curl -i https://api-dev.hhc.workers.dev/ --header 'Content-Length: 0'curl -i https://api-dev.hhc.workers.dev/ --header 'Content-Length: 0'

This time it's returning the "Worker threw exception" HTML page instead of 520, but in the same way there are no entries (exceptions or otherwise) in the realtime logs for the request. Refresher:

Occurs on workers.devworkers.dev domain only, not custom domain
Smart placement on
Content-Length: 0Content-Length: 0 header set
May or may not be related to AKL colo (it was last time) but it just happens I'm only getting paired with AKL so I can't verify otherise

Brett Willis•9/26/24, 8:04 PM

Actually this time seems unrelated to smart placement. Happens whether on or off.

ac•9/26/24, 8:06 PM

Slight bug with new worker CI/CD: adding a root directory for the worker in the inital drawer page doesn't work, it goes back to

after you create the integration. Editing it afterwards works fine.

ac•9/26/24, 8:07 PM

Small plug: Adding build watch paths for Workers CI/CD (like we have for Pages) would be fantastic. We have a monorepo so I'd love to specify which directories matter for actually re-deploying

Mozzy•9/26/24, 8:45 PM

I think something broke about an hour ago with Workers

Mozzy•9/26/24, 8:47 PM

My Discord bot that was making POST requests to my workers.dev domain gets the <title>Worker threw exception | bfstats.ragnarok.workers.dev | Cloudflare </title><title>Worker threw exception | bfstats.ragnarok.workers.dev | Cloudflare </title> HTML page back. Hitting my custom domain works as usual though

I don't even see a log entry for the exception on the Logs page

Mozzy•9/26/24, 8:49 PM

workers.dev vs custom domain: