[tor-dev] Scalability or Onionbalance for v3 ephemeral/ADD_ONION services

[Holmes Wilson]

Would this return a list of currently-online onion addresses in possession
of the frontend address key?

Or would it just route traffic to one of those addresses invisibly?

For our application (a messaging app) it would be super useful to get the
full list of known online (or recently seen online) onion addresses in
possession of some frontend key. This would let us use onionbalance for
peer discovery instead of blindly trying the set of all known peers, which
won't work well for large groups / large numbers of peers.

I'd be interested in working with others on a spec for this!

On Mon, Jun 14, 2021 at 6:25 AM George Kadianakis <desnacked at riseup.net>
wrote:

> Chad Retz <chad.retz at gmail.com> writes:
>
> > A quick glance at the code shows that ADD_ONION (i.e. "ephemeral"
> > onion services) doesn't support setting an Onionbalance
> > frontend/master onion address (specifically
> > https://gitlab.torproject.org/tpo/core/tor/-/issues/32709 doesn't seem
> > to have a control-side analogue). Would a feature request for adding a
> > `*(SP "OnionbalanceMasterKey=" OBKey)` (or "OBMasterKey" or whatever)
> > to ADD_ONION be reasonable? If so, just add in Gitlab?
> >
>
> Hell Ched,
>
> that's indeed something that is missing and a reasonable feature
> request. A spec/code patch would be particularly welcome ;)
>
> > Also curious alternative scalability and load balancing options for
> > ephemeral v3 onion services. I have read
> >
> https://www.benthamsgaze.org/wp-content/uploads/2015/11/sucu-torscaling.pdf
> > but unsure if anything more recent has been written. Beyond that and
> > Onionbalance, any other interesting approaches I could employ
> > (assuming I can dev anything from a control port pov, but am wanting
> > to work w/ an unmodified Tor binary)?
> >
>
> Another complementary approach is to split the 'introduction' and
> 'rendezvous' functionalities to different hosts:
>
> https://gitlab.torproject.org/tpo/core/torspec/-/blob/main/proposals/255-hs-load-balancing.txt
> However it hasn't been implemented yet...
>
> Cheers!
> _______________________________________________
> tor-dev mailing list
> tor-dev at lists.torproject.org
> https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.torproject.org/pipermail/tor-dev/attachments/20210616/96048471/attachment.htm>