This is speedbrain (new blob post about it yesterday I think). It's turned on by default, but you ca

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...

texanOP•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...

texanOP•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

scheduled

scheduled

and

fetch

fetch

run the same function,

getIncidents

getIncidents

, 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.dev
```
```
workers.dev
```
domain only, not custom domain
Smart placement on
```
Content-Length: 0
```
```
Content-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:

Brett Willis•9/26/24, 8:50 PM

https://www.cloudflarestatus.com/incidents/1lrs7s9vc8lv

Workers Apps is Reportedly Having Trouble Hitting the Observability...

Mozzy•9/26/24, 8:50 PM

That makes a lot of sense :phew:

Mozzy•9/26/24, 8:51 PM

Good time to use the custom domain I guess

texanOP•9/26/24, 8:58 PM

Thanks for the reports all-- if you redeploy your workers, you should be back in action!

mallets•9/26/24, 9:01 PM

Please update the workers Static Assets docs with the size and file limits. Pages doc mentions 25 MB size and 20K files per site.

Moe•9/26/24, 9:11 PM

Whats going on with workers all POST requests giving 500 errors again

Advany•9/26/24, 9:12 PM

pfff... I just attached customdomains to all workers except one... quickly redeployed it and its working again!

MMoe Whats going on with workers all POST requests giving 500 errors again

Mozzy•9/26/24, 9:12 PM

Possibly: workers-and-pages-discussions

Two•9/26/24, 9:15 PM

i love wondering why my thing suddenly broke (on prod) and spending hours trying to resolve an issue only to realize it was an issue with cloudflare workers

TTwo i love wondering why my thing suddenly broke (on prod) and spending hours trying...

Moe•9/26/24, 9:19 PM

ikr this exact issue happened to me about 3 weeks ago i kept debugging for 2 hours why post requests are broken for this guy in iceland and only in iceland just to know its cloudflare trolling as usual

Moe•9/26/24, 9:20 PM

small apps can barely handle their own bugs and now big corps also having issues is making it even more doomed

MMoe ikr this exact issue happened to me about 3 weeks ago i kept debugging for 2 hou...

Two•9/26/24, 9:20 PM

cloudflare discriminating against people in iceland

(for legal reasons, this is a joke)

Two•9/26/24, 9:21 PM

cloudflare trolling is real

Moe•9/26/24, 9:23 PM

that guy was also an og enterprise user and after hours i was like "sorry it was cloudflare stopping thier service randomly in iceland" i was so confused i bought a premium vpn to access iceland srvers and it was actually broken for like 3/4 hours or even more

Moe•9/26/24, 9:23 PM

i think if i didn't mention it on discord prob would be still broken till now xD

Jakob Norlin•9/26/24, 9:31 PM

Trying out the new workers with assets. Just initiated a solid-js project as described in the tutorial and deployed. Looks like static assets aren't cached by default Cache-Control: public, max-age=0, must-revalidateCache-Control: public, max-age=0, must-revalidate.

How do I configure the worker with a proper cache-control header for hashed static assets built by the framework?

This is speedbrain (new blob post about it yesterday I think). It's turned on by default, but you ca

Similar Threads