Does your bubble live app works fine? Suddenly, my live users can’t see or interact with any data: [image] I can’t see anything as well, I have a dropdown with a list of agents and it shows empty: [image] Is someone experiencing similar issues? It’s like privacy rules are not being applied? +20 minutes live apps down with no data, and who knows what else is not working on the backend but everything green, such a trusty status page [image]

Thanks for fixing quickly but come on… something like this must surely be picked up during testing (automated or otherwise). It feels like Bubble continuously uses the shared cluster as a test environment at the expense of our live apps.

Hey @josh, I hope you’re doing well. I’m getting straight to the point here: the recent Bubble updates have been a pain. Our app, along with many others, has broken a few times after these core updates. Most times, nobody realizes there’s an issue until someone reports it on the forum. This means our live apps are down for an hour in the best scenario, or more, which is super frustrating for both us and our users. I think we already suggested this, but once again, here’s what I suggest: Let us control updates: Can we get an option to opt-out of automatic core updates on the shared clusters? It’d be great if we could choose when to upgrade. This way, we can test things out and make sure nothing breaks before making the move. Exactly the same thing we have with version control when a breaking-change is introduced. Keep an eye on the forum post-update: After you guys roll out a new update, it’d be cool if someone from the team could actively monitor the forum for at least the next hour. Issues usually pop up quick, and this would help in catching them early. AI-forum watcher: I’m sure you guys can use an AI such as OpenAI to detect possible forum threads that are reporting systems down or a massive failure like this thread and internally fire the alarms. I know you guys are working hard, and it’s a headache to detect this, but this can’t happen every few weeks. Just these couple of tweaks would make things smoother for all of us. Redu.

it appears to be working normally

There is a way to opt out of updates but it’s only available on higher tiers, but agree with your points about watching the forum post-code push. Also — is there any code reviews before the code is pushed to production? I think really the only way to get through to flag this as a real big issue to Bubble right now is by submitting a big volume of the same type of bug (which is why these forum posts are made, I think? To see if others are having issues, if so, post a bug report) then their success team escalates. Also — does doing @ Bubble still ping the internal team? My question to that is how long does it actually take to escalate. Although I wasn’t directly affected, would love to see how this is going to be prevented in the future @josh! It’d also be great to get a Postmortem

I mean to be fair, since users have full control (just about) over their app, it’d be almost impossible to find all the edge cases. We’re just a little prone to more errors because, at least in the last year, there’s been so many updates. The standard and best practices have changed versus someone who has joined bubble recently. That’s my opinion at least. Could there be improvements, yes. I’d say if you notice issues, check to see the release page from bubble as well. You can tie recent releases to bugs if you notice them. Releases | Bubble

Yeah, I know there’s an option for higher tiers, that’s why I remarked for shared clusters. [image] yusaney1: Can we get an option to opt-out of automatic core updates on the shared clusters? They can’t use us as beta testers.

Agreed! Looking forward to the postmortem

I also do admit that the unstableness is sorta getting old I do get that updates are being pushed at a rate that maybe internal processes aren’t being updated at the frequency they should be but maybe there should be a pause for at least 1 week to figure out the internal processes before continuing to ship potentially broken, buggy code

Can we please have an explanation as to why the Status page is all green today even though the main cluster had issues for more than 1h? [image] How can we explain to users of our apps why their sessions weren’t working correctly when no issue was reported? What’s the point of subscribing to the Status page if metrics do not represent the true states? Many thanks.

@josh Has there been any postmortem on this issue? What is Bubble doing to prevent this from happening again? Also, my biggest frustration was that after checking out the status page which was all green, I thought the issue was on my end. So I started to dig everywhere until I (fortunately) came across this post on the forum

Something is down?

Questions

georgecollier October 18, 2023, 3:21pm 55

I still feel the exact same way as after the previous editor issues a month ago:

Echoing this.

Topic		Replies	Views
Anyone's live data glitching? Questions	83	2912	August 3, 2024
Page data not updating Database	77	11007	September 1, 2023
Incident postmortem from 11/29 Announcements	45	5096	December 21, 2022
Bubble Down May 6 Bugs	103	3580	May 8, 2024
Performance issues / test live issues [Sept 19] Bugs	47	1050	September 24, 2025

Something is down?

Related topics