Add details about CI architecture, graphs and troubleshooting

This is an introduction for CI architecture with the description of data that we have for monitoring.

I would like it to be a base for CI shared runners training that I would like to organize in order to pass as much of my knowledge.

We could then collaboratively try to work on more accurate run books for most common problems that we faced recently.

The current state is that the resolutions are part of the docs, and this is also what I did in a couple of recent days trying to figure out and solve it.

cc @stanhu @pcarranza @ernstvn