Cloudflare Developers•2w ago

Feeling like alone on the problem here 🙂 At the end I finally found that I can reproduce my problem

Feeling like alone on the problem here 🙂 At the end I finally found that I can reproduce my problem locally with cloudflare only, so I guess I will switch to another queue system, except if someone can help me, I lost my message after a weird delay, I can't see why with current API or UI on cloudflare

9 Replies

Mitya•7d ago

Yeah it sucks when these CF channels mothball and you can't get help. A CF dev once told me this Discord is volunteer-led, which surprised me as I had assumed it was part of their job to offer support on here. I suppose if you need a guaranteed response other (non-Discord) means are preferable.

boop•7d ago

Cloudflare makes it pretty clear that you need to pay for support

Charles DOP•6d ago

I have an enterprise account in fact My co-founder told me it was super long (like 2 day) to have an answer with an issue open on the support, so I just tried here, but I was already having a non answer message 3 month ago on golang sdk -> I was thinking nobody was on golang sdk, but maybe somebody was here. I can open an issue on support, but at the end I will just loose time, so I will just move out asap. I see here is a global issue, not my use case right now, I've debug everything on the whole thing and loose too much time compare to another provider A path to reproduce, and maybe it's my account with a quota or anything else, but I don't care I have no warning, no information, no observability : - New queue from web UI - Put a message in it from web UI - Check the retention message (set to more than one day) - Go to the list message from UI after 1-2 minute -> Message disapear - And I have no consumer at all plug to this queue It's not an alpha product from what I see, so I don't understand at all what happen here, my only fix is to : - Acknowledge a message asap after receive it - Use 100 retry on CF queue, to handle CF failure, not mine -> But I don't want to code a retry system on my hand, that's something where I want to rely on the queue, so I won't search more on that and find another solution

boop•6d ago

If you're on enterprise, you should get in touch with your account rep to push this harder. If what you say is true and there is an issue with queues on your account, you're much more likely to get it resolved faster through your official channels

Isaac McFadyen•6d ago

That's correct, we're all volunteers on this DIscord. Even the employees are here in their free time (and aren't specifically assigned to the Discord or anything).

Unknown User•2d ago

Message Not Public

Mitya•2d ago

Ah well then that's fair enough, and I thank you for all the help. It's helped me many times. Could someone please clarify when it is and isn't necessary/a good idea to use waitUntil() inside a queue consumer? The docs suggest using it, however queue consumers already run asynchronously, in the background (not as the result of a HTTP trigger), so unless the consumer is doing a LOT of work, and is therefore at risk of exceeding Worker limits, what's the point? My consumer receives a bunch of messages and sends an email (via an API) for each one. Traffic to this queue is low, and the settings are quite conservative:

"max_batch_size": 10,
"max_batch_timeout": 5,
"max_concurrency": 3

"max_batch_size": 10,
"max_batch_timeout": 5,
"max_concurrency": 3

digitalpoint•2d ago

Have also been seeing elevated 10000: Authentication error messages from our users. Not for Queues, but other API endpoints. No change to the API tokens, and across multiple sites/accounts. Seems sporadic... sometimes it works, sometimes it doesn't, but seems to have started in the last 8 hours.

Unknown User•2d ago

Message Not Public

Gaming

Programming

Feeling like alone on the problem here 🙂 At the end I finally found that I can reproduce my problem

Did you find this page helpful?