Next one does all nodes connected to the modified node
Last one does only the node(s) changed.
What really happens when you deploy is the node is completely reinitialised/recreated. I.e connections are closed & reopened, internal state is lost, etc.
I stick with the last option, modified nodes. It's quicker and less disruptive.
What im looking for is to have let's say 500 tabs or so and only deploy via API ONLY the changes that i've made inside the tabs. For example, i have 500 tabs and i only make changes on tabs 1, 300 and 456.
I know there is an API funcionality but i don't know how to use it.
Ok, i got some Documentation ready for the reading...so i'll continue...
The thing is that my project will require to have several clients and i want to separate them by that. I mean, every tab is a client. So if i modify something from one client i don't want to "broke" any other client. But....it will be a good thing to use the API because of versioning.