Document factors that contribute to user server start time

[choldgraf]

Context

In the LEAP hub, there was some concern about the length of start time for some of the users. In some cases this can be improved with technology changes, but it will also help if we can help users understand the different factors to consider with start time.

@rabernat provided some good explanation about the things to keep in mind here:

leap-stc/leap-stc.github.io#8 (reply in thread)

this is an issue to keep track of that so that we can incorporate these ideas into our documentation somehow.

[GeorgianaElena]

I believe we should prioritize this issue, so I will add it to the sprint board and increase its priority.

Documenting what contributes to node startup times is especially important now that we are transitioning to have bigger nodes be shared by multiple users. It's important to document that each user that will generate a new node event, will see a longer wait time than the other users that will fit onto an existing node.

Also, in the context of an event, such a setup enables the instructor to "pre-warm" the hub by themselves, by just starting a server before the event. That server star event, will spin up a node and that node will be ready for the initial hub users to use.