web services - Basic webserver state - daemon or new process per web-hit?

Question

I am trying to write an online system, where info is distributed via websockets to interested clients in realtime from one or many datasources (also connected via websocket). (Something like SignalR)

This seems to be a three part process:

1. I can write all the client code to interface with a websocket process on the server
2. I can write the code for the datasource to send data to my web app via websocket
3. Some system needs to accept data from step 2, and optionally distribute to clients from step 1

I fundamentally don't know how step three operates. I understand serving a static or dynamic page to a browser, but that state isn't changing in a material way, and each server request will result in the same output. Easy.

Questions:

• Do I need a daemon running 24/7 on the server keeping realtime info from step 2 in memory, and distributing to step 1 clients?

• If the server-side websocket from step 2 gets info, how does it communicate with the websocket handling step 1 clients?
  It can't be via database, because then the step 1 websocket would have to keep polling the database, which seems extremely silly.

There is something fundamental that I don't understand, and haven't found a primer for on the web.
I am on a shared linux host.

question from:https://stackoverflow.com/questions/66066268/basic-webserver-state-daemon-or-new-process-per-web-hit

Answer 1

Waitting for answers