[tor-bugs] #23170 [Core Tor/Tor]: Include ed25519 relay id keys in the consensus

Tor Bug Tracker & Wiki blackhole at torproject.org
Tue Sep 26 01:05:38 UTC 2017


#23170: Include ed25519 relay id keys in the consensus
-------------------------------------------------+-------------------------
 Reporter:  asn                                  |          Owner:  nickm
     Type:  task                                 |         Status:
                                                 |  accepted
 Priority:  Very High                            |      Milestone:  Tor:
                                                 |  0.3.3.x-final
Component:  Core Tor/Tor                         |        Version:
 Severity:  Normal                               |     Resolution:
 Keywords:  prop224 tor-dirauth tor-hs ed25519   |  Actual Points:
  needs-proposal                                 |
Parent ID:                                       |         Points:  1
 Reviewer:                                       |        Sponsor:
                                                 |  SponsorR-can
-------------------------------------------------+-------------------------

Comment (by teor):

 Replying to [comment:6 arma]:
 > Hopefully useful thoughts:
 >
 > A) Is the only reason we need ed keys in the consensus for this hashring
 thing? That seems like a sad reason to put in all the overhead for every
 client for every hour forever.

 Consensus diffs will make this into a once-off addition for 0.3.1 and
 later clients.

 Also, if we're having trouble getting some descriptors for our primary
 guards, does this fix that issue as well?

 > A2) And if that is the only use, we could put like 4 characters of the
 key in, as just a hashring hint?

 No, because the key is hashed, so you need the entire key.

 > B) Is there a longer term plan for phasing out rsa keys entirely, and
 replacing them with ed keys?

 I hope so.

 > If not, maybe prop224 hashring should use rsa keys.

 No, it's too late to change this for 0.3.2.

 > C) Maybe the consensus overhead isn't so bad because of consensus diffs.
 Or maybe it is? Has anybody done any follow-up on the consensus diff
 deployment to see what effect it has had?

 Yes, ahf wrote an entire technical report on it. Search tor-project@ for
 Isabela's recent report.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/23170#comment:7>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online


More information about the tor-bugs mailing list