Traccar CPU performance
What i meant is that is does not seem to be logic to develop horizontal scaling while in certain circumstances a single server cannot use more the one core.
You're right, but why did you mention horizontal scaling? I don't see any discussion about scaling in this topic beside ours right here.
Because horizontal scaling is (as i understand it) used to take the load of 1 server and distribute the load over several servers (so that you can scale up better with more devices). So my point is, if traccar server is (not yet) optimized to use all cores from one server cpu (for some processes) then it would seem more logic to first optimize traccar server to use all cores on one server. Before they developed the horizontal scaling idear (just my opinion :)
That makes sense. But something tells me this will not be done (for free, at least) unless it becomes a critical problem worth fixing. Or until someone else fixes it and makes a pull request at GitHub. So far it seems I'm the only one who managed to hang and crash the server using Computed Attributes (and probably can still do it!).
You're right, this is not related to horizontal scaling. The whole question was about Traccar using only a single core to process computed attributes.