There has been a huge spike in usage today. The app is a website but has only 4 users accessing it. As you can see from the chart low usage every day and suddenly a massive spike, This is 95% fetching data. There are no backend flows running
Uploading: Screenshot 2024-07-18 at 11.34.39 PM.png…
Hi Ian,
The chart is not visible. But besides that, are you able to drill down by clicking in the chart on the segment for ‘fetchin data’? That way you should be able to see what is causing the spike. Additionally you can go to Logs to see what is happening, although that is on a very detailed level.
When the above options are not possible or do not provide more information, it might be good to contact Bubble Support.
In order to get back up and running fast you can go to Settings → App plan and for instance allow overages or add more workload (see image below). But it might be best to find the reason for the spike first.
Thank you. The workload usage went from a daily < 5,000 for the last 2 years to 187,000 yesterday. So I don’t want to pay for that extra workload. Clearly something weird is happening. 95% of it was accessing data. I’ve not got any backend workflows. I have so little data… Biggest database has 4,000 entries.
Looking at the metrics, the app workload is 35,000 per hour. How do I work out what that is?
As as we’ve been locked out of the app, none of the 4 users are in!!!
If you can’t get in at all then this probably isn’t helpful, but on the little pie chart underneath the bar chart in your screenshot above, you can click each segment to get increasingly granular data about what is causing the wf unit consumption.
Thanks you. I didn’t know you could just keep clicking on the pie chart. Super helpful. Now I know what the search is.
So it gets weirder. The huge usage is showing search for "FanURLLinkPlays. There are only 163 records in that database.
Do you have any recursive workflows that aren’t terminating properly? That would likely be the cause of a sudden spike like this. Happy to have a look if you’re happy to give me access.
No recursive flows. Clicking into the metrics it is pointing at a search inside the app on a page that no-one has access to (becasue app is locked). The page is doing 600 calls a minute, to a database of 163 records.
@ian18 Not great! 
In the last few days Bubble has quietly and vastly improved the documentation around WU. How to masure, track, optimise … set up alerts when things go haywire … put in recursion loop protection…
https://manual.bubble.io/help-guides/workload
And to immediately stop the bleeding (and your app) till you are back in control - Pause the scheduler!!!
Thank you. But I have no scheduled tasks in the app. I’ve gone in and cancelled all tasks
Look at “tracking workload” - https://manual.bubble.io/help-guides/workload/tracking-workload
(if you haven’t already) - It is entirely possible to drill down to exactly what is causing this usage.
Clicking into the metrics it takes me to a page it says the search that is running is in a popup that displays a RG searching a database that currently has 163 records. No-one can access that page, as the app has been locked for 2 days.
Go nuclear. Delete that RG. Stop the bleeding. If after deleting it is still going bonkers … that tells you something else. :-). Maybe export your App for posterity so in a calm moment you can inspect at your leisure
I’ve done that. Thanks.
Put a conditional onto the RG to set the datasource based on when the popup is visible. Make the default datasource of the RG empty. Set the conditional datasource to be static.
