Anyone Using WebSocket API for Custom Dashboards?

EODHD API
1

Hey folks,
I’ve recently started playing around with the WebSocket API to stream real-time data into a simple dashboard I’m building for personal use (mainly tracking a few US tickers).

Just curious — is anyone else here using WebSocket for something similar? Would love to hear how you’re handling reconnections, formatting the data for charts, or any cool tricks you’ve figured out.

Also open to any libraries or tools you’d recommend to make the whole setup smoother. Thanks!

2

Hi.
I’ve been using the EODHD Python library for this — it wraps the WebSocket stuff nicely and lets me subscribe to live prices with minimal setup. I feed the real-time data into a pandas DataFrame, apply a few moving average filters, and chart the output using Plotly Dash. Works great for monitoring short-term price action across selected assets.

3

I reconnect every time the socket drops. Simple loop but works fine for now.

4

I built a Node.js app that connects to the WebSocket and listens for updates across a filtered list of tickers. I batch those updates every 10 seconds and dump them into a local SQLite DB. From there, I do some light processing before visualizing it in a small web UI I built with Chart.js. The biggest issue I ran into was handling dropped connections — I ended up using a heartbeat + reconnect strategy. Took a while to tweak it, but it’s been stable for weeks now.

5

Just a tip: avoid overloading your WebSocket connection with too many tickers. I split them into chunks by exchange and manage separate streams. It helps reduce lag and makes debugging easier when something breaks.

6

Hey everyone!

I’ve also been using the WebSocket API to feed data into a custom dashboard — I’m running Node.js + Socket.io on the backend, and React with D3.js on the frontend.

A few practical takeaways:

  1. Auto-reconnect logic: I send heartbeat pings every 30 seconds, and if there’s no pong, I trigger a reconnect immediately.
  2. Batching updates: I collect incoming quotes into 1-second slots and push them as small batches to the UI — this keeps rendering smooth.
  3. Selective subscriptions: I subscribe only to the symbols I actually need — it makes a big difference in performance when tracking a large list.

Tools/libs I use: reconnecting-websocket — super helpful for stability, and lodash.debounce to reduce re-renders caused by noisy tick updates.

Curious how others handle high-traffic periods — do you adjust your update frequency dynamically? Like throttling during earnings releases or FOMC announcements? Would love to hear strategies.

Thanks for the insights!

7

Yeah, same here — batching + lodash debounce helps a lot.

For high-traffic, I just throttle harder (up to ~1s). Haven’t had issues unless tracking 50+ tickers.

Reconnect logic looks solid. Curious if you buffer missed data after reconnect or just pick up from live?

8

I just pick up from the live stream after reconnecting. I don’t buffer missed data — for my use case, short gaps aren’t critical. If needed, I handle historical recovery separately with a different process.