Backend Workflow bug

georgecollier · 2024-10-03T22:41:17.664Z

Click ‘resume tasks’

webiztlv · 2024-10-03T22:41:43.376Z

image1870×846 114 KB

done

what now?

Jici · 2024-10-03T22:42:17.668Z

What is the workflow that you tried to schedule? Does it appear there?

georgecollier · 2024-10-03T22:42:31.052Z

Okay, well your next workflow is scheduled for 2:00am your time, so it’ll run in 20 minutes.

If you’re not seeing workflows, you need to schedule them.

webiztlv · 2024-10-03T22:44:16.352Z

I didn’t quite understand what you were saying

But let’s say you said that a workflow should be scheduled in 20 minutes, right? So the problem is that it just won’t run because of the bug…

georgecollier · 2024-10-03T22:44:57.246Z

webiztlv:

So the problem is that it just won’t run because of the bug…

Cool, in 20 minutes, show us the log entries for 1:59am-2:01am.

webiztlv · 2024-10-03T22:46:13.008Z

Now I checked and this specific order has been deleted

But I understood what you want to see - I will now do something that will show you before and after

Jici · 2024-10-03T22:46:53.858Z

Also, share your backend WF settings, step, conditionnal.
Do you have privacy rules on your DB that could affect the WF?..

webiztlv · 2024-10-03T22:47:42.934Z

theres no privacy rules
theres conditionnal - but again it worked before the loop on saturday

Jici · 2024-10-03T22:48:38.332Z

When you said you “halted” the loop. What did you do?

webiztlv · 2024-10-03T22:50:43.654Z

image1877×471 56.7 KB

Before 01:50

After 01:50

image1888×951 168 KB

(i pressed “show” after 01:50)

webiztlv · 2024-10-03T22:53:03.122Z

image953×584 25.6 KB

I just made an api call to the mysql server that transfers my data

image674×766 26 KB

webiztlv · 2024-10-03T22:54:27.638Z

image1876×913 165 KB

updated - look at time

georgecollier · 2024-10-03T22:59:30.700Z

What does your app metrics tab show (both hourly and daily)?

Jici · 2024-10-03T22:59:55.481Z

You didn’t answer about what you did to stop the loop.

Can you show step 2 details of your WF?

In any of your screenshot I see your “update_every_day_metabase” scheduled… At this moment, it doesn’t seem to be scheduled…

webiztlv · 2024-10-03T23:03:08.091Z

Because I don’t use it anymore

For my part, if it should even be deleted

I just want to go back to how the app worked a week or two ago

Jici · 2024-10-03T23:08:46.804Z

Ok so don’t show them! We will not focus of what you don’t need! Actually, your backend WF are scheduled. The next one was planned for oct 4 1h50 min. Check after in logs if your workflow run (and disappear from your scheduler) Actually, there’s absolutely no sign that your backend WF doesnt run. They are all scheduled in the future. Wait for them to run and check logs.

You said your app is not working as before. Explain what exactly is not working? You said almost all backend workflows ceased to function properly. According to the scheduler, there’s no issue for scheduling them.
So maybe you can show a specific workflow that is scheduled and is not working anymore. But the best is to wait at oct 1h50 and check logs to validate your WF active wf maker is completed

webiztlv · 2024-10-03T23:15:17.153Z

I will describe the problem again

The application is intended for an office complex - it is connected to electrical components

As soon as someone makes an order and its due date arrives, the order changes to “active” status and the components in the office turn on, and as soon as the order ends, the components “turn off” - and the status changes to “finished”

The problem is that the backend workflows of turning on and off, etc. don’t work, that is, they don’t run - they don’t change status, they don’t make any api calls

I’m talking to you about any backend workflow, including those of sending a simple email

webiztlv · 2024-10-03T23:15:40.493Z

I can give you access to test or even do a zoom call with you if you can help that would be great…

What I did contact support and it seems they are investigating and also don’t understand why things aren’t working

JohnMark · 2024-10-04T00:00:13.070Z

I’ve just spent a few days dealing with the same experience using Bubble… and it’s not over yet! The first visible symptom was a hidden button that somehow created over 20,000 WUs in just a few minutes. Alarms went off, and I immediately stopped all processes. After that, other symptoms appeared—SMS sending stopped, verification routines failed, and so on. It feels like the Bubble code got, let’s say, ‘scrambled’… which makes everything very difficult to fix and even harder to explain to Bubble. So, I’m not wasting time asking them for help. The only solution is to rebuild in order to fix this kind of situation, which is a waste of dozens of hours, I must admit. Normally, I’d have made a backup by creating a copy with the database to experiment and try to identify the problem. But even this isn’t easy, as you literally need to strip out everything unnecessary… to isolate the issue and then fix it in the correct version.