is this strictly a one dashboard for one tor daemon or will this also be usable to monitor multiple tor instances without running multiple dashboard instances in parallel?
It's being developed as one tor instance <-> one dashboard instance, although it'd be a nice feature to include what you mention into **future development**.
Hi nusenu. This is something I've long wanted to do with Nyx, and definitely something for long term consideration. It's a valid use case, especially for gigabit relay families that run multiple instances to saturate their connection. We did this, for instance, with Amunet. But that said, not a feature for the initial release.
Would that include data that is not available via the controlport, like exit/guard probability, (measured bw), cw and cw fraction (available via onionoo) or is that not something that you are planing to include?
Think I'm gonna need to say 'patches welcome' for these. First step would be to add them to Stem. Anything in Onionoo is easy to bring in (it's chiefly just distilled descriptor information), but I'm not positive about the exit/guard probabilities. Is it a simple ratio? Or do tor clients do something more sophisticated? If it's something simple I'd be happy to take a patch adding it to the Consensus class.
Cheers! -Damian