Improving performance of dashboard with multiple parallel callbacks

Dash Python
1

Hi everyone!
I’m building a dashboard that aggregates metrics from different online sources (it makes API requests and parses html pages). The requests frequencies range from every 10 seconds to every minute. Here is a callback graph for them.

callback_graph
callback_graph1522×543 53.5 KB

Sometimes, when callback timers overlap, new items are not added to containers on the front-end, animations don’t play, etc. But they are added to storages on the back-end. Logs show, that timing of some callbacks is not right. The interval for this function is set to 10 seconds:

logs
logs984×657 73.9 KB

I use partial updates where i can (updating news-feed-div and mir-feed-div). I’ve tried making those long callbacks background, but their time only went up. I can’t use caching, because i need to parse data and check it for updates everytime.

I work on windows.

Are there any ways to improve performance and solve these issues?

2

Welcome to the community, @MrTwister.

Do you use client-side callbacks? These usually help speeding things up.

3

No, I don’t. But i’m going to try (never wrote in JS, so it’ll take some time)! Thank you!
Another question then. I use API Tokens for requests, but clientside callbacks can’t use back-end variables or files. What is a best-practice solution to storing/passing sensitive info to perform clientside callbacks? Or is it better to re-write only callbacks without such info?