Tor Shared Random Subsystem Specification This document specifies how the commit-and-reveal shared random subsystem of Tor works. This text used to be proposal 250-commit-reveal-consensus.txt. Table Of Contents: 1. Introduction 1.1. Motivation 1.2. Previous work 2. Overview 2.1. Introduction to our commit-and-reveal protocol 2.2. Ten thousand feet view of the protocol 2.3. How we use the consensus [CONS] 2.3.1. Inserting Shared Random Values in the consensus 2.4. Persistent State of the Protocol [STATE] 2.5. Protocol Illustration 3. Protocol 3.1 Commitment Phase [COMMITMENTPHASE] 3.1.1. Voting During Commitment Phase 3.1.2. Persistent State During Commitment Phase [STATECOMMIT] 3.2 Reveal Phase 3.2.1. Voting During Reveal Phase 3.2.2. Persistent State During Reveal Phase [STATEREVEAL] 3.3. Shared Random Value Calculation At 00:00UTC 3.3.1. Shared Randomness Calculation [SRCALC] 3.4. Bootstrapping Procedure 3.5. Rebooting Directory Authorities [REBOOT] 4. Specification [SPEC] 4.1. Voting 4.1.1. Computing commitments and reveals [COMMITREVEAL] 4.1.2. Validating commitments and reveals [VALIDATEVALUES] 4.1.4. Encoding commit/reveal values in votes [COMMITVOTE] 4.1.5. Shared Random Value [SRVOTE] 4.2. Encoding Shared Random Values in the consensus [SRCONSENSUS] 4.3. Persistent state format [STATEFORMAT] 5. Security Analysis 5.1. Security of commit-and-reveal and future directions 5.2. Predicting the shared random value during reveal phase 5.3. Partition attacks 5.3.1. Partition attacks during commit phase 5.3.2. Partition attacks during reveal phase 6. Discussion 6.1. Why the added complexity from proposal 225? 6.2. Why do you do a commit-and-reveal protocol in 24 rounds? 6.3. Why can't we recover if the 00:00UTC consensus fails? 7. Acknowledgements 1. Introduction 1.1. Motivation For the next generation hidden services project, we need the Tor network to produce a fresh random value every day in such a way that it cannot be predicted in advance or influenced by an attacker. Currently we need this random value to make the HSDir hash ring unpredictable (#8244), which should resolve a wide class of hidden service DoS attacks and should make it harder for people to gauge the popularity and activity of target hidden services. Furthermore this random value can be used by other systems in need of fresh global randomness like Tor-related protocols (e.g. OnioNS) or even non-Tor-related (e.g. warrant canaries). 1.2. Previous work Proposal 225 specifies a commit-and-reveal protocol that can be run as an external script and have the results be fed to the directory authorities. However, directory authority operators feel unsafe running a third-party script that opens TCP ports and accepts connections from the Internet. Hence, this proposal aims to embed the commit-and-reveal idea in the Tor voting process which should make it smoother to deploy and maintain. 2. Overview This proposal alters the Tor consensus protocol such that a random number is generated every midnight by the directory authorities during the regular voting process. The distributed random generator scheme is based on the commit-and-reveal technique. The proposal also specifies how the final shared random value is embedded in consensus documents so that clients who need it can get it. 2.1. Introduction to our commit-and-reveal protocol Every day, before voting for the consensus at 00:00UTC each authority generates a new random value and keeps it for the whole day. The authority cryptographically hashes the random value and calls the output its "commitment" value. The original random value is called the "reveal" value. The idea is that given a reveal value you can cryptographically confirm that it corresponds to a given commitment value (by hashing it). However given a commitment value you should not be able to derive the underlying reveal value. The construction of these values is specified in section [COMMITREVEAL]. 2.1. Ten thousand feet view of the protocol Our commit-and-reveal protocol aims to produce a fresh shared random value (denoted shared_random_value here and elsewhere) every day at 00:00UTC. The final fresh random value is embedded in the consensus document at that time. Our protocol has two phases and uses the hourly voting procedure of Tor. Each phase lasts 12 hours, which means that 12 voting rounds happen in between. In short, the protocol works as follows: Commit phase: Starting at 00:00UTC and for a period of 12 hours, authorities every hour include their commitment in their votes. They also include any received commitments from other authorities, if available. Reveal phase: At 12:00UTC, the reveal phase starts and lasts till the end of the protocol at 00:00UTC. In this stage, authorities must reveal the value they committed to in the previous phase. The commitment and revealed values from other authorities, when available, are also added to the vote. Shared Randomness Calculation: At 00:00UTC, the shared random value is computed from the agreed revealed values and added to the consensus. This concludes the commit-and-reveal protocol every day at 00:00UTC. 2.3. How we use the consensus [CONS] The produced shared random values need to be readily available to clients. For this reason we include them in the consensus documents. Every hour the consensus documents need to include the shared random value of the day, as well as the shared random value of the previous day. That's because either of these values might be needed at a given time for a Tor client to access a hidden service according to section [TIME-OVERLAP] of proposal 224. This means that both of these two values need to be included in votes as well. Hence, consensuses need to include: (a) The shared random value of the current time period. (b) The shared random value of the previous time period. For this, a new SR consensus method will be needed to indicate which authorities support this new protocol. 2.3.1. Inserting Shared Random Values in the consensus After voting happens, we need to be careful on how we pick which shared random values (SRV) to put in the consensus, to avoid breaking the consensus because of authorities having different views of the commit-and-reveal protocol (because maybe they missed some rounds of the protocol). For this reason, authorities look at the received votes before creating a consensus and employ the following logic: - First of all, they make sure that the agreed upon consensus method is above the SR consensus method. - Authorities include an SRV in the consensus if and only if the SRV has been voted by at least the majority of authorities. - For the consensus at 00:00UTC, authorities include an SRV in the consensus if and only if the SRV has been voted by at least AuthDirNumAgreements authorities (where AuthDirNumAgreements is a newly introduced consensus parameter). Authorities include in the consensus the most popular SRV that also satisfies the above constraints. Otherwise, no SRV should be included. The above logic is used to make it harder to break the consensus by natural partioning causes. We use the AuthDirNumAgreements consensus parameter to enforce that a _supermajority_ of dirauths supports the SR protocol during SRV creation, so that even if a few of those dirauths drop offline in the middle of the run the SR protocol does not get disturbed. We go to extra lengths to ensure this because changing SRVs in the middle of the day has terrible reachability consequences for hidden service clients. 2.4. Persistent State of the Protocol [STATE] A directory authority needs to keep a persistent state on disk of the on going protocol run. This allows an authority to join the protocol seamlessly in the case of a reboot. During the commitment phase, it is populated with the commitments of all authorities. Then during the reveal phase, the reveal values are also stored in the state. As discussed previously, the shared random values from the current and previous time period must also be present in the state at all times if they are available. 2.5. Protocol Illustration An illustration for better understanding the protocol can be found here: https://people.torproject.org/~asn/hs_notes/shared_rand.jpg It reads left-to-right. The illustration displays what the authorities (A_1, A_2, A_3) put in their votes. A chain 'A_1 -> c_1 -> r_1' denotes that authority A_1 committed to the value c_1 which corresponds to the reveal value r_1. The illustration depicts only a few rounds of the whole protocol. It starts with the first three rounds of the commit phase, then it jumps to the last round of the commit phase. It continues with the first two rounds of the reveal phase and then it jumps to the final round of the protocol run. It finally shows the first round of the commit phase of the next protocol run (00:00UTC) where the final Shared Random Value is computed. In our fictional example, the SRV was computed with 3 authority contributions and its value is "a56fg39h". We advice you to revisit this after you have read the whole document. 3. Protocol In this section we give a detailed specification of the protocol. We describe the protocol participants' logic and the messages they send. The encoding of the messages is specified in the next section ([SPEC]). Now we go through the phases of the protocol: 3.1. Commitment Phase [COMMITMENTPHASE] The commit phase lasts from 00:00UTC to 12:00UTC. During this phase, an authority commits a value in its vote and saves it to the permanent state as well. Authorities also save any received authoritative commits by other authorities in their permanent state. We call a commit by Alice "authoritative" if it was included in Alice's vote. 3.1.1. Voting During Commitment Phase During the commit phase, each authority includes in its votes: - The commitment value for this protocol run. - Any authoritative commitments received from other authorities. - The two previous shared random values produced by the protocol (if any). The commit phase lasts for 12 hours, so authorities have multiple chances to commit their values. An authority MUST NOT commit a second value during a subsequent round of the commit phase. If an authority publishes a second commitment value in the same commit phase, only the first commitment should be taken in account by other authorities. Any subsequent commitments MUST be ignored. 3.1.2. Persistent State During Commitment Phase [STATECOMMIT] During the commitment phase, authorities save in their persistent state the authoritative commits they have received from each authority. Only one commit per authority must be considered trusted and active at a given time. 3.2. Reveal Phase The reveal phase lasts from 12:00UTC to 00:00UTC. Now that the commitments have been agreed on, it's time for authorities to reveal their random values. 3.2.1. Voting During Reveal Phase During the reveal phase, each authority includes in its votes: - Its reveal value that was previously committed in the commit phase. - All the commitments and reveals received from other authorities. - The two previous shared random values produced by the protocol (if any). The set of commitments have been decided during the commitment phase and must remain the same. If an authority tries to change its commitment during the reveal phase or introduce a new commitment, the new commitment MUST be ignored. 3.2.2. Persistent State During Reveal Phase [STATEREVEAL] During the reveal phase, authorities keep the authoritative commits from the commit phase in their persistent state. They also save any received reveals that correspond to authoritative commits and are valid (as specified in [VALIDATEVALUES]). An authority that just received a reveal value from another authority's vote, MUST wait till the next voting round before including that reveal value in its votes. 3.3. Shared Random Value Calculation At 00:00UTC Finally, at 00:00UTC every day, authorities compute a fresh shared random value and this value must be added to the consensus so clients can use it. Authorities calculate the shared random value using the reveal values in their state as specified in subsection [SRCALC]. Authorities at 00:00UTC start including this new shared random value in their votes, replacing the one from two protocol runs ago. Authorities also start including this new shared random value in the consensus as well. Apart from that, authorities at 00:00UTC proceed voting normally as they would in the first round of the commitment phase (section [COMMITMENTPHASE]). 3.3.1. Shared Randomness Calculation [SRCALC] An authority that wants to derive the shared random value SRV, should use the appropriate reveal values for that time period and calculate SRV as follows. HASHED_REVEALS = H(ID_a | R_a | ID_b | R_b | ..) SRV = SHA3-256("shared-random" | INT_8(REVEAL_NUM) | INT_4(VERSION) | HASHED_REVEALS | PREVIOUS_SRV) where the ID_a value is the identity key fingerprint of authority 'a' and R_a is the corresponding reveal value of that authority for the current period. Also, REVEAL_NUM is the number of revealed values in this construction, VERSION is the protocol version number and PREVIOUS_SRV is the previous shared random value. If no previous shared random value is known, then PREVIOUS_SRV is set to 32 NUL (\x00) bytes. To maintain consistent ordering in HASHED_REVEALS, all the ID_a | R_a pairs are ordered based on the R_a value in ascending order. 3.4. Bootstrapping Procedure As described in [CONS], two shared random values are required for the HSDir overlay periods to work properly as specified in proposal 224. Hence clients MUST NOT use the randomness of this system till it has bootstrapped completely; that is, until two shared random values are included in a consensus. This should happen after three 00:00UTC consensuses have been produced, which takes 48 hours. 3.5. Rebooting Directory Authorities [REBOOT] The shared randomness protocol must be able to support directory authorities who leave or join in the middle of the protocol execution. An authority that commits in the Commitment Phase and then leaves MUST have stored its reveal value on disk so that it continues participating in the protocol if it returns before or during the Reveal Phase. The reveal value MUST be stored timestamped to avoid sending it on wrong protocol runs. An authority that misses the Commitment Phase cannot commit anymore, so it's unable to participate in the protocol for that run. Same goes for an authority that misses the Reveal phase. Authorities who do not participate in the protocol SHOULD still carry commits and reveals of others in their vote. Finally, authorities MUST implement their persistent state in such a way that they will never commit two different values in the same protocol run, even if they have to reboot in the middle (assuming that their persistent state file is kept). A suggested way to structure the persistent state is found at [STATEFORMAT]. 4. Specification [SPEC] 4.1. Voting This section describes how commitments, reveals and SR values are encoded in votes. We describe how to encode both the authority's own commitments/reveals and also the commitments/reveals received from the other authorities. Commitments and reveals share the same line, but reveals are optional. Participating authorities need to include the line: "shared-rand-participate" in their votes to announce that they take part in the protocol. 4.1.1. Computing commitments and reveals [COMMITREVEAL] A directory authority that wants to participate in this protocol needs to create a new pair of commitment/reveal values for every protocol run. Authorities SHOULD generate a fresh pair of such values right before the first commitment phase of the day (at 00:00UTC). The value REVEAL is computed as follows: REVEAL = base64-encode( TIMESTAMP || H(RN) ) where RN is the SHA3 hashed value of a 256-bit random value. We hash the random value to avoid exposing raw bytes from our PRNG to the network (see [RANDOM-REFS]). TIMESTAMP is an 8-bytes network-endian time_t value. Authorities SHOULD set TIMESTAMP to the valid-after time of the vote document they first plan to publish their commit into (so usually at 00:00UTC, except if they start up in a later commit round). The value COMMIT is computed as follows: COMMIT = base64-encode( TIMESTAMP || H(REVEAL) ) 4.1.2. Validating commitments and reveals [VALIDATEVALUES] Given a COMMIT message and a REVEAL message it should be possible to verify that they indeed correspond. To do so, the client extracts the random value H(RN) from the REVEAL message, hashes it, and compares it with the H(H(RN)) from the COMMIT message. We say that the COMMIT and REVEAL messages correspond, if the comparison was successful. Participants MUST also check that corresponding COMMIT and REVEAL values have the same timestamp value. Authorities should ignore reveal values during the Reveal Phase that don't correspond to commit values published during the Commitment Phase. 4.1.4. Encoding commit/reveal values in votes [COMMITVOTE] An authority puts in its vote the commitments and reveals it has produced and seen from the other authorities. To do so, it includes the following in its votes: "shared-rand-commit" SP VERSION SP ALGNAME SP IDENTITY SP COMMIT [SP REVEAL] NL where VERSION is the version of the protocol the commit was created with. IDENTITY is the authority's SHA1 identity fingerprint and COMMIT is the encoded commit [COMMITREVEAL]. Authorities during the reveal phase can also optionally include an encoded reveal value REVEAL. There MUST be only one line per authority else the vote is considered invalid. Finally, the ALGNAME is the hash algorithm that should be used to compute COMMIT and REVEAL which is "sha3-256" for version 1. 4.1.5. Shared Random Value [SRVOTE] Authorities include a shared random value (SRV) in their votes using the following encoding for the previous and current value respectively: "shared-rand-previous-value" SP NUM_REVEALS SP VALUE NL "shared-rand-current-value" SP NUM_REVEALS SP VALUE NL where VALUE is the actual shared random value encoded in hex (computed as specified in section [SRCALC]. NUM_REVEALS is the number of reveal values used to generate this SRV. To maintain consistent ordering, the shared random values of the previous period should be listed before the values of the current period. 4.2. Encoding Shared Random Values in the consensus [SRCONSENSUS] Authorities insert the two active shared random values in the consensus following the same encoding format as in [SRVOTE]. 4.3. Persistent state format [STATEFORMAT] As a way to keep ground truth state in this protocol, an authority MUST keep a persistent state of the protocol. The next sub-section suggest a format for this state which is the same as the current state file format. It contains a preamble, a commitment and reveal section and a list of shared random values. The preamble (or header) contains the following items. They MUST occur in the order given here: "Version" SP version NL [At start, exactly once.] A document format version. For this specification, version is "1". "ValidUntil" SP YYYY-MM-DD SP HH:MM:SS NL [Exactly once] After this time, this state is expired and shouldn't be used nor trusted. The validity time period is till the end of the current protocol run (the upcoming noon). The following details the commitment and reveal section. They are encoded the same as in the vote. This makes it easier for implementation purposes. "Commit" SP version SP algname SP identity SP commit [SP reveal] NL [Exactly once per authority] The values are the same as detailed in section [COMMITVOTE]. This line is also used by an authority to store its own value. Finally is the shared random value section. "SharedRandPreviousValue" SP num_reveals SP value NL [At most once] This is the previous shared random value agreed on at the previous period. The fields are the same as in section [SRVOTE]. "SharedRandCurrentValue" SP num_reveals SP value NL [At most once] This is the latest shared random value. The fields are the same as in section [SRVOTE]. 5. Security Analysis 5.1. Security of commit-and-reveal and future directions The security of commit-and-reveal protocols is well understood, and has certain flaws. Basically, the protocol is insecure to the extent that an adversary who controls b of the authorities gets to choose among 2^b outcomes for the result of the protocol. However, an attacker who is not a dirauth should not be able to influence the outcome at all. We believe that this system offers sufficient security especially compared to the current situation. More secure solutions require much more advanced crypto and more complex protocols so this seems like an acceptable solution for now. Here are some examples of possible future directions: - Schemes based on threshold signatures (e.g. see [HOPPER]) - Unicorn scheme by Lenstra et al. [UNICORN] - Schemes based on Verifiable Delay Functions [VDFS] For more alternative approaches on collaborative random number generation also see the discussion at [RNGMESSAGING]. 5.2. Predicting the shared random value during reveal phase The reveal phase lasts 12 hours, and most authorities will send their reveal value on the first round of the reveal phase. This means that an attacker can predict the final shared random value about 12 hours before it's generated. This does not pose a problem for the HSDir hash ring, since we impose an higher uptime restriction on HSDir nodes, so 12 hours predictability is not an issue. Any other protocols using the shared random value from this system should be aware of this property. 5.3. Partition attacks This design is not immune to certain partition attacks. We believe they don't offer much gain to an attacker as they are very easy to detect and difficult to pull off since an attacker would need to compromise a directory authority at the very least. Also, because of the byzantine general problem, it's very hard (even impossible in some cases) to protect against all such attacks. Nevertheless, this section describes all possible partition attack and how to detect them. 5.3.1. Partition attacks during commit phase A malicious directory authority could send only its commit to one single authority which results in that authority having an extra commit value for the shared random calculation that the others don't have. Since the consensus needs majority, this won't affect the final SRV value. However, the attacker, using this attack, could remove a single directory authority from the consensus decision at 24:00 when the SRV is computed. An attacker could also partition the authorities by sending two different commitment values to different authorities during the commit phase. All of the above is fairly easy to detect. Commitment values in the vote coming from an authority should NEVER be different between authorities. If so, this means an attack is ongoing or very bad bug (highly unlikely). 5.3.2. Partition attacks during reveal phase Let's consider Alice, a malicious directory authority. Alice could wait until the last reveal round, and reveal its value to half of the authorities. That would partition the authorities into two sets: the ones who think that the shared random value should contain this new reveal, and the rest who don't know about it. This would result in a tie and two different shared random value. A similar attack is possible. For example, two rounds before the end of the reveal phase, Alice could advertise her reveal value to only half of the dirauths. This way, in the last reveal phase round, half of the dirauths will include that reveal value in their votes and the others will not. In the end of the reveal phase, half of the dirauths will calculate a different shared randomness value than the others. We claim that this attack is not particularly fruitful: Alice ends up having two shared random values to choose from which is a fundamental problem of commit-and-reveal protocols as well (since the last person can always abort or reveal). The attacker can also sabotage the consensus, but there are other ways this can be done with the current voting system. Furthermore, we claim that such an attack is very noisy and detectable. First of all, it requires the authority to sabotage two consensuses which will cause quite some noise. Furthermore, the authority needs to send different votes to different auths which is detectable. Like the commit phase attack, the detection here is to make sure that the commitment values in a vote coming from an authority are always the same for each authority. 6. Discussion 6.1. Why the added complexity from proposal 225? The complexity difference between this proposal and prop225 is in part because prop225 doesn't specify how the shared random value gets to the clients. This proposal spends lots of effort specifying how the two shared random values can always be readily accessible to clients. 6.2. Why do you do a commit-and-reveal protocol in 24 rounds? The reader might be wondering why we span the protocol over the course of a whole day (24 hours), when only 3 rounds would be sufficient to generate a shared random value. We decided to do it this way, because we piggyback on the Tor voting protocol which also happens every hour. We could instead only do the shared randomness protocol from 21:00 to 00:00 every day. Or to do it multiple times a day. However, we decided that since the shared random value needs to be in every consensus anyway, carrying the commitments/reveals as well will not be a big problem. Also, this way we give more chances for a failing dirauth to recover and rejoin the protocol. 6.3. Why can't we recover if the 00:00UTC consensus fails? If the 00:00UTC consensus fails, there will be no shared random value for the whole day. In theory, we could recover by calculating the shared randomness of the day at 01:00UTC instead. However, the engineering issues with adding such recovery logic are too great. For example, it's not easy for an authority who just booted to learn whether a specific consensus failed to be created. 7. Acknowledgements Thanks to everyone who has contributed to this design with feedback and discussion. Thanks go to arma, ioerror, kernelcorn, nickm, s7r, Sebastian, teor, weasel and everyone else! References: [RANDOM-REFS]: http://projectbullrun.org/dual-ec/ext-rand.html https://lists.torproject.org/pipermail/tor-dev/2015-November/009954.html [RNGMESSAGING]: https://moderncrypto.org/mail-archive/messaging/2015/002032.html [HOPPER]: https://lists.torproject.org/pipermail/tor-dev/2014-January/006053.html [UNICORN]: https://eprint.iacr.org/2015/366.pdf [VDFS]: https://eprint.iacr.org/2018/601.pdf