for my part, I’m not on the latest version and I’m also experiencing erratic backend workflow behavior.
Sometimes SAWOL (list of 20 or 30 items) doesn’t execute completely, but the next time it works without changing anything.
For some time now, everything to do with lists has been relatively unstable. I’ve come to the point where I don’t use list of items as a data field at all.
Another example: I have a use case linked to appointment management. There’s a creation phase for various items which is triggered the day before the appointment. the BE id is recorded (data field for that), but sometimes the associated process disappears (from the scheduler)! and so the creation is incomplete. For no reason at all, if I force the workflow manually on the associated data, it passes.
So I had to add a verification step when opening the appointment to check.
Using the “:make static” expression when scheduling the affected back end workflow (thanks to the advice of @Roody_BubbleSupport ) seems to have fixed my issue for now - would recommend giving this a try in the meantime!
The issue with the API workflow seems to have been silently fixed. However, I still need to go through numerous discrepancies caused by this malfunction. Additionally, I’m noticing inconsistencies in the database where deleted items still appear in the Modify dialog, which is quite strange and makes it difficult to run workflows for corrections.
Here is an example: this payroll batch is deleted but Bubble still reads it:
it doesn’t show up on the table and shows up as deleted_thing when I export to CSV:
I am curious why there’s no official post on the resolution from Bubble other than @fede.bubble saying the team is pushing a fix they “hope” will resolve the issue. I also haven’t had any reply to my bug report saying that it was resolved (or any response at all).
Is that the typical process when a bug is found and reported? At minimum, it would be nice to receive an acknowledgment that the issue existed and that it has been definitely resolved. I spent a fair bit of time troubleshooting internally, and then having to go back and forth to provide documentation of my issue to Bubble. It all just feels a little lackluster of a response given the severity of the issue.
I used the word “hope” because I knew the issue was affecting a lot of people and I didn’t want to sound definitive and then have to come back 5 mins later explaining that the fix had uncovered another issue that now had to be worked on. So that’s my bad, sorry.
The engineering team is working on a post-mortem and once that’s ready we’ll share the info with everyone
