i have a lot of data, my code (it can be
i have a lot of data, my code (it can be better) uses a lot of indexes to do its job which is fine, but i was shocked to see that my dashboard usage was this high
5 Replies
Heya, we don't expect the dashboard to consume this much bandwidth -- we use some idle heuristics to determine if the live loading of data on the data page should be paused, but if you're browsing data actively on the data page on a high-throughput table, you may consume a decent amount of bandwidth.
Can you share your deployment name so we can dig into this more? (feel free to DM it to me)
Hey @ari , I tried DM'ing it to you, but was not able to
The dashboard used 80GB since May 20, but I have not used the dashboard for looking at data for the past 3-4 days? Something is going quite sideways, I obviously don't know what I am doing, and I want to figure out what is wrong else its going to cost me too much, thanks for your help.
Sent you a friend request! You should be able to accept it and message me then.
Yes that’s certainly odd. Do the charts of usage by day show the day when you used the most?
That's the weird thing, unless I am going clinically insane, I am 99% certain I have not viewed data inside the dashboard at all. I do spend a lot of time inside convex, but I avoid going to
data. 🤣
The usage chart shows the opposite, which is super confusing.
Convo taken to DMs, thank you!
We figured out a few issues:
- The usage actually came from cloud backups and the dashbaord was marking that as "Dashboard" usage. We'll fix that so it says "Cloud backup" usage
- The customer uses the action-retrier component which only cleans up runs older than 7 days. This can lead to a very large table that is included in backups. We're going to release the ability to customize the max job age before cleanup
- We'll make the cloud backup dashboard page clearer that backups incur database bandwidth usage equal to the size of your backup when they run
Thanks @dejeszio for the debugging help 🙂