[tor-dev] Draft of proposal "Stop HS address enumeration by HSDirs"

[Nick Mathewson]

On Fri, Aug 16, 2013 at 10:31 AM, George Kadianakis
<desnacked at riseup.net> wrote:

As before: let me know when there's something you should I should give a number.


> Filename: xxx-onion-antienumeration-hsdirs.txt
> Title: Stop HS address enumeration by HSDirs
> Author: George Kadianakis
> Created: 16 August 2013
> Target: 0.2.5.x
> Status: Draft
>
>                                          [More draft than Guiness.]
>
> 0. Proposal overview and motivation:
>
>   Currently, it is the case that, HSDirs can read the descriptors of the
>   Hidden Services they host. Even if the descriptor was encrypted,
>   HSDirs could still learn the address of the HSes by logging the client
>   directory requests (which contain the hidden service address).
>
>   This proposal:
>
>   a) Defines cryptographic procedures to be followed by Tor clients
>      and hidden services to allow HS address antienumeration.
>
>   b) Defines a new descriptor format which hides the descriptor
>      content from entities who don't know the public key of the HS.
>
>   c) Changes the size (and semantics) of Hidden Service addresses.

To be fair, the other proposal would need to do that too.

> 1. Acknowledgments
>
>   The cryptography behind this proposal was originally proposed by
>   Robert Ransom in a private email threadi and subsequently posted in
>   tor-dev [0]. Discussion was continued in trac ticket #8106 [1].
>
>   During the past 6 months many bright people have looked at the
>   cryptography behind this scheme. The list of people includes Nadia
>   Heninger, Leonid Reyzin, Nick Hopper, Aggelos Kiayias, Tanja Lange,
>   Dan J. Bernstein and probably others that I can't recall at this
>   point. Thanks!

One more:  Christian Grothoff told me in Garching that GNUnet does
something quite similar for its keys. So we should probably check out
their approach too, and include them in the "related work" section of
any hypothetical publication here. :)


> 2. Parameters
>
>   TIME_PERIOD is XXX
>
> 3. Scheme overview
>
>   Currently, Hidden Services upload their unencrypted descriptor to
>   hidden service directories (HSDirs). HSDirs store the unencrypted
>   descriptor in an internal map of: <hs address> -> <hs descriptor>
>   When a client wants the descriptor of an HS, it asks an HSDir for
>   the descriptor that corresponds to <hs address>. If the HSDir has
>   such an index in its map, it returns the <hs descriptor> to the
>   client.

See also the not-entirely-baked "stealth authentication" approach in
the current hidden service design.

>   This proposal asks Hidden Services, to periodically derive a new
>   ephemeral keypair from its long-term identity key; the new keypair
>   is a function of the identity key and a time-dependent nonce. The
>   derivation should be one-way; if you know the identity key you
>   should be able to derive the ephemeral key, but not the other way
>   around. Finally, a client should be able to derive the ephemeral HS
>   public key from the long-term HS public key without knowing the
>   long-term HS private key (#KEYPAIRDERIVATION)
>
>   Hidden Services then encrypt their descriptor with a symmetric key
>   (derived from the ephemeral public key)

Derived from the _permanent_ public key, surely?  The ephemeral one is
the one attackers know.

>      and sign the ciphertext and
>   the ephemeral public key with their ephemeral private key. Then they
>   place the ephemeral public key, the encrypted descriptor and the
>   signature in a "metadescriptor" document (#METADESC) and send it to
>   the HSDir (#DESCPUBLISH).
>
>   The HSDir validates the signature of the "metadescriptor", and if
>   it's legit it stores the metadescriptor in an internal map of:
>   <ephemeral public key> -> <metadescriptor> .
>   (#DESCPUBLISH)
>
>   Now, out of band, the HS gives to its clients a <z>.onion
>   address. <z> in this case is the long-term public key of the HS
>   (this is different from the current situation where <z> is the hash
>   of the long-term public key).

We should discuss the referer problem here, and make sure there's a
good solution.

The design should allow an additional out-of-band key that doesn't
appear in the URL.

>
>   A client that knows the <z>.onion address and wants to acquire the
>   HS descriptor, derives the <ephemeral public key> of the HS by using
>   <z> and the same key derivation procedure that the HS uses. She also
>   derives the symmetric key that decrypts the encrypted HS descriptor
>   (#DESCFETCH).
>
>   The client then contacts the appropriate HSDir and inquires for the
>   descriptor with index <ephemeral public key>. If the HSDir has such
>   an index in its internal map, it passes the <metadescriptor> to the
>   client (#DESCFETCH).
>
>   The client then verifies the signature of the metadescriptor, and if
>   it's legit she decrypts the encrypted descriptor with the symmetric
>   key. The client now has the Hidden Servide descriptor she was so
>   looking for (#DESCDECRYPTION).
>
> 4. Security proof
>
>   XXX A security proof of the above scheme is under development. We
>   are not going to implement or deploy anything before we have a solid
>   proof of this.
>
> 5. Changes to the current HS protocol
>
> 5.0. Related proposals
>
>     This proposal is supposed to be applied on top of the "Migrate HS
>     identity keys to Ed25519" proposal.
>
> 5.1. Ephemeral keypair derivation (#KEYPAIRDERIVATION)
>
>     XXX Leaving this unpsecified for now till the security proof comes
>     along.
>
>     For now, let's assume that after this paragraph each HS has a
>     per-TIME_PERIOD ephemeral keypair. It also has a symmetric key
>     derived from the ephemeral public key (maybe even the hash of the
>     ephemeral public key) to encrypt its descriptor.
>
>     Let's also assume that each client can generate the ephemeral public
>     key of a HS given only its long-term public key (and knowledge of
>     the current time-dependent nonce).
>
> 5.2. HS descriptor publishing
>
> 5.2.1. Metadescriptor format (#METADESC)
>
>     The format of the metadescriptor that is uploaded to the HSDir is:
>
>       "ephemeral-public-key" NL public-key
>
>         [At start, exactly once]
>
>         The ephemeral public key for this time period in base64 encoding.

This also needs a publication date and/or expiry indicator, so that
HSDirs know how long to hold them.

>       "encrypted-descriptor" NL encrypted-descriptor
>
>         [Exactly once]
>
>         An encrypted v3 hidden service descriptor (as specified in
>         xxx-hs-ecc-id-keys.txt). It's encrypted using the ephemeral
>         symmetric key of the HS, encoded in base64 and surrounded with
>         "-----BEGIN MESSAGE-----" and "-----END MESSAGE-----".

How is this symmetric key derived, exactly?  What format does the
encryption take?  (Also, the encryption should probably have a MAC on
it.)

>       "signature" NL signature
>
>        [At end, exactly once]
>
>        A signature of all fields above with the ephemeral private key
>        of the hidden service.

As before, let me suggest that all of this take a format analogous to
that used in proposal 220.


> 5.2.2. Metadescriptor publishing (#DESCPUBLISH)
>
>     The metadescriptor specified above is published to the appropriate
>     HSDir by sending an HTTP 'POST' request to
>     "/tor/rendezvous3/publish" as specified in the "Migrate HS
>     identity keys to Ed25519" proposal.
>
>     # XXX should this get its own URI, even though we assume that these
>     two proposals will be implemented and deployed simultaneously?
>
>     Upon receiving a descriptor, the directory server checks the
>     signature, and discards the descriptor if the signature does not
>     match the enclosed public key. If the signature matches, the
>     directory server saves the descriptor in a map of:
>     <ephemeral-public-key> -> <metadescriptor>.
>
> 5.3. Metadescriptor fetching (#DESCFETCH)
>
>     A client that knows the long-term public key (onion address) of a
>     hidden service can derive the ephemeral public key and the
>     ephemeral symmetric key as specified in section 5.1.
>
>     A client that wants to fetch the metadescriptor of an HS, does an
>     HTTP 'GET' request to the appropriate directory server asking for
>     "/tor/rendezvous3/<z>" where <z> is the ephemeral public key of the
>     HS.
>
>     # XXX should this get its own URI, even though we assume that these
>     two proposals will be implemented and deployed simultaneously?

I suggest that we deploy _only_ the format in this proposal.  Is that
what you meant?


> 5.4. Metadescriptor decryption (#DESCDECRYPTION)
>
>     The client then verifies the signature of the metadescriptor, and
>     discards it if it doesn't match the ephemeral public key that was
>     previously derived.
>
>     If the signature matches, the client uses the derived ephemeral
>     symmetric key to decrypt the 'encrypted-descriptor' part of the
>     metadescriptor.
>
> 6. Discussion
>
>   [Points here might deserve their own sections]
>
>   Do we need the HSDir hash ring, even though the HS address and the
>   descriptor are now hidden from HSDirs?

Probably!

>   An ed25519 public key is 32 bytes. 32 bytes in base32 encoding is 56
>   characters (or 52 with the '=' padding removed). Do we want a
>   different URL encoding or are we happy with addresses like:
>   mfrggzdfmztwq2lknnwg23tpobyxe43uov3ho6dzpjaueq2eivda.onion ?

That's okay with me.


(The migration proposal is sure going to be interesting, isn't it?)


One more thing to think about here: any design here should think about
HSDir prediction resistance, as in
https://trac.torproject.org/projects/tor/ticket/8244 )

yrs,
-- 
Nick