Hi,
Does anybody know why the Bubble SQL connector is so slow? I’m loading 200 rows of data through my Bubble app and it takes 7-8 seconds. The same query takes 0.5 seconds when run through a website widget connected to the same MySQL database. The current performance makes the app feel very sluggish.
A less complex query takes less time, seemingly by a similar factor. So the query through the app might take 3 seconds and 0.2 seconds through the widget.
The servers hosting the MySQL database and widget are in a data centre in Germany. Could there be a distance factor from the Bubble servers? Previously the MySQL database server was in the US and the time lag seemed the same so it doesn’t seem so.
I have a very large MySQL db that I access using the db connector with some very complex SQL select statements (nested selects, joins, groups, sorts, etc). While I experience some small degredation in performance, it’s nowhere near a 15x like you are experiencing. Queries that are sub-second responses in MySQL workbench take maybe 3 seconds with Bubble. Hope that’s useful information.
Also, I’ve found setting up a test page to be very helpful in these situations. Simply put a button and a repeating group with no data source on the same page. Set it up to just display a couple fields in the repeating group. On button press, trigger a workflow that executes the SQL query and displays the repeating group using the results. Don’t do any sorting or filtering of the data source. If it takes less than 7 seconds, then you can tune from there.
FWIW, I’ve found if you do all the filtering and sorting in the SQL statement it’s way faster than bringing down all the results and filtering in Bubble.
I’m working through supabase right now on a very small dataset, but I can say that on my test page, the first call takes about 1.7sec. any calls while on that page take about 0.6 seconds, till I refresh the page.
