How does client_push exactly change the communication?

How does client_push changes the communication between the client and the server? According to the readme, without client_push, the client “polls” for changes on the server with some sort of scheduler. After installing client_push, this “polls” are reduced. But i found nowehere a explanation.

  • How often does a desktop client retrieve information from the server without client_push
  • How often does it retrieve information from the server when client_push is installed?
  • Or does the client only do this when starting or shutting down and otherwise wait for pushes?

The readme also mentions that not all changes on the server are pushed to the clients.

  • Under what circumstances does this not happen?
  • Does it even make sense to set up client_push for an instance with 4 users and 8 clients?

With 8 clients, I assume that each user uses one desktop and one mobile client. The mobile clients will occasionally be “online” at the same time, while the desktop clients will rarely or never be online at the same time.

You can configure this with the remotePollInterval and by default it is 30s. And you have a few more intervals, e.g. notificationRefreshInterval (1min).

github readme:

With many clients all checking for updates a large portion of the server load can consist of just these update checks.

You can always check the server load for your specific use case. In case your system is a bit slow, there might be other points to optimize first (different caches, database, file locking, …)

Many thx. I could swear that I read this document, but I can’t remember that section. Thank you very much.

This topic was automatically closed 8 days after the last reply. New replies are no longer allowed.