Re: [bitcoin-dev] Preventing/detecting pinning of jointly funded txs
On Sun, Nov 06, 2022 at 06:22:08PM -0500, Antoine Riard via bitcoin-dev wrote: > Adding a few more thoughts here on what coinjoins/splicing/dual-funded > folks can do to solve this DoS issue in an opt-in RBF world only. > > I'm converging that deploying a distributed monitoring of the network > mempools in the same fashion as zeroconf people is one solution, as you can > detect a conflicting spend of your multi-party transaction. > Let's say you > have a web of well-connected full-nodes, each reporting all their incoming > mempool transactions to some reconciliation layer. > > This "mempools watchdog" infrastructure isn't exempt from mempools > partitioning attacks by an adversary, A mempool partitioning attack requires the adversary to identify your nodes. If they're just monitoring and not being used as the initial broadcaster of your txs, that should be difficult. (And I think it would make sense to do things that make it more difficult to successfully partition a node for tx relay, even if you can identify it) > where the goal is to control your > local node mempool view. A partitioning trick is somehow as simple as > policy changes across versions (e.g allowing Taproot Segwit v0.1 spends) or > two same-feerate transactions. The partitioning attack can target at least > two meaningful subsets. An even easier way to partition the network is to create two conflicting txs at the same fee/fee rate and to announce them to many peers simultaneously. That way neither will replace the other, and you can build from there. In order to attack you, the partition would need to be broad enough to capture all your monitoring nodes on one side (to avoid detection), and all the mining nodes on the other side (to prevent your target tx from being confirmed). If that seems likely, maybe it indicates that it's too easy to identify nodes that feed txs to mining pools... > Either the miner mempools only, by conflicting all > the reachable nodes in as many subsets with a "tainted" transaction (e.g > set a special nSequence value for each), and looking on corresponding > issued block. Or targeting the "watchdog" mempools only, where the > adversary observation mechanism is the multi-party blame assignment round > itself. I think there's a few cases like that: you can find out what txs mining pools have seen by looking at their blocks, what explorers have seen by looking at their website... Being able to use that information to identify your node(s) -- rather than just your standardness policy, which you hopefully share with many other nodes -- seems like something we should be working to fix... > There is an open question on how many "divide-and-conquer" rounds > from an adversary viewpoint you need to efficiently identify all the > complete set of "mempools watchdog". If the transaction-relay topology is > highly dynamic thanks to outbound transaction-relay peers rotation, the > hardness bar is increased. I'm not sure outbound rotation is sufficient? In today's world, if you're a listening node, an attacker can just connect directly, announce the conflicting tx to you, and other txs to everyone else. For a non-listening node, outbound rotation might be more harmful than helpful, as it increases the chances a node will peer with a given attacker at some point. > Though ultimately, the rough mental model I'm thinking on, this is a > "cat-and-mouse" game between the victims and the attacker, where the latter > try to find the blind spots of the former. I would say there is a strong > advantage to the attacker, in mapping the mempools can be batched against > multiple sets of victims. While the victims have no entry barriers to > deploy "mempools watchdog" there is a scarce resource in contest, namely > the inbound connection slots (at least the miners ones). Anytime you deploy a new listening node, you use up 10 inbound connections, but provide capacity for 115 new ones. Worst case (if it's too hard to prevent identifying a listening node if you use it for monitoring), you setup all your monitoring nodes as non-listening nodes, and also set enough listening nodes in different IP ranges to compenasate for the number of outbound connections your monitoring nodes are making, and just ignore the mempools of those listening nodes. > Victims could batch their defense costs, in outsourcing the monitoring to > dedicated entities (akin to LN watchtower). However, there is a belief in > lack of a compensation mechanism, you will have only a low number of public > ones (see number of BIP157 signaling nodes, or even Electrum ones). Seems like a pretty simple service to pay for, if there's real demand: costs are pretty trivial, and if it's a LN service, you can pay for it over lightning... Fairly easy to test if you're getting what you're paying for too, by simultaneously announcing two conflicting txs paying yourself, and checking you get an appropriate alert. > Assuming we can solve them, there is still the issue of
Re: [bitcoin-dev] Preventing/detecting pinning of jointly funded txs
Hi AJ, Adding a few more thoughts here on what coinjoins/splicing/dual-funded folks can do to solve this DoS isse in an opt-in RBF world only. I'm converging that deploying a distributed monitoring of the network mempools in the same fashion as zeroconf people is one solution, as you can detect a conflicting spend of your multi-party transaction. Let's say you have a web of well-connected full-nodes, each reporting all their incoming mempool transactions to some reconciliation layer. This "mempools watchdog" infrastructure isn't exempt from mempools partitioning attacks by an adversary, where the goal is to control your local node mempool view. A partitioning trick is somehow as simple as policy changes across versions (e.g allowing Taproot Segwit v0.1 spends) or two same-feerate transactions. The partitioning attack can target at least two meaningful subsets. Either the miner mempools only, by conflicting all the reachable nodes in as many subsets with a "tainted" transaction (e.g set a special nSequence value for each), and looking on corresponding issued block. Or targeting the "watchdog" mempools only, where the adversary observation mechanism is the multi-party blame assignment round itself. There is an open question on how many "divide-and-conquer" rounds from an adversary viewpoint you need to efficiently identify all the complete set of "mempools watchdog". If the transaction-relay topology is highly dynamic thanks to outbound transaction-relay peers rotation, the hardness bar is increased. Though ultimately, the rough mental model I'm thinking on, this is a "cat-and-mouse" game between the victims and the attacker, where the latter try to find the blind spots of the former. I would say there is a strong advantage to the attacker, in mapping the mempools can be batched against multiple sets of victims. While the victims have no entry barriers to deploy "mempools watchdog" there is a scarce resource in contest, namely the inbound connection slots (at least the miners ones). Victims could batch their defense costs, in outsourcing the monitoring to dedicated entities (akin to LN watchtower). However, there is a belief in lack of a compensation mechanism, you will have only a low number of public ones (see number of BIP157 signaling nodes, or even Electrum ones). Outsource mempools monitoring will hit the same issue of bounded public resources, and as such be a "single-point-of-censorship" vector. Reminder, we would like LN mobile clients from low-budget users to access those fancy joint funding protocols (or at least I). So as a first partial conclusion, not only the security efficiency but also the economic scalability of such defensive "mempools watchdog" infrastructure remains an open question to me. Assuming we can solve them, there is still the issue of assigning blame reliably among a set of trust-minimized joint funding protocol participating UTXOs. Indeed, you're running quickly into issues like *two* double-spend from two sybilling participants, aiming to halt the assignment process. There is likely a need to introduce some "UTXO-satoshi-weight" vote to efficiently converge towards assignment. At the very least it would require the attacker to control more than 51% of the contributed UTXO to manipulate the outcome of the blame assignment process. Assuming an economically honest majority, you still have the timevalue cost inflicted for each round of blame assignment. Assuming 255 inputs (current LN's interactive construction protocol limit) and a transaction propagation delay of 2min (30s ?) on the p2p network, an attacker controlling all the inputs minus 1 might be able to DoS for ~50 blocks (do we have other factors to think of in the design of the blame assignment process ?). In a future where the timevalue of circulating coins is priced in (IMO when we have competitive LN routing markets), this is probably a significant damage. On the other hand, you have a full-rbf world, where instead to deploy or gain access to "mempools watchdog" and proceed to a timevalue-expensive blame assignment protocol, any participant should be able to fee-bump the joint transaction (assuming multiple pre-signed feerate version of the transactions, or ephemeral, nversion=3 and package-relay to do unilateral CPFP). Ideally, this would be a reduction to a "flood-and-loot" attack, i.e the attacker is constrained to buy the blockspace. A situation with a lot of visibility for the joint funding protocol victims, I think. Side-note: this alternative resolution process of relying on full-rbf, still assumes solving RBF pinning rule 3, I think a fact I underscored in my original full-rbf proposal of last year [0]. All that said, I think it's good to think more of the end-of-pipeline economic trade-offs of the two main directions to solve this DoS affecting joint funding protocol. Transaction signature withhold DoS should be defended on a different layer, and I think there are far more easy to deal with in a
Re: [bitcoin-dev] Preventing/detecting pinning of jointly funded txs
Assigning blame here seems to be the paramount concern here. If we can assign blame, most coinjoin-like protocols can terminate in bounded block time, assuming fees are properly set. It's also worth noting that in coinjoin cases, they're obviously coinjoins, so pinging explorers over Tor HS seems purely additive to me. The use-cases that can't use it are other privacy methods like coinswap and similar, where there's no blockchain indication anything different is happening. The larger the coinjoin, the more potential honest users, the more potential for a duplicitous double-spend to be gossiped among those peers. For dual funding LN channels, that number is pretty small(2), so I suspect the DoS concerns are fairly subtle. Might be worth talking to CLN/Eclair/Other LN teams that are working through those subtleties as we speak. Greg On Tue, Nov 1, 2022 at 11:52 PM Anthony Towns via bitcoin-dev < bitcoin-dev@lists.linuxfoundation.org> wrote: > On Fri, Oct 28, 2022 at 03:21:53AM +1000, Anthony Towns via bitcoin-dev > wrote: > > What should folks wanting to do coinjoins/dualfunding/dlcs/etc do to > > solve that problem if they have only opt-in RBF available? > > ref: > https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2022-October/021124.html > > So, having a go at answering my own question. > > I think ultimately the scenario here is: > > * you have a joint funding protocol, where everyone says "here's >an unspent utxo that will be my contribution", collaborates on signing >a transaction spending all those utxos, and then broadcasts it > > * everyone jointly agrees to pay some amount in fees for that >transaction, targeting confirmation within N blocks > > * the goal is to have the transaction confirm, obviously; but it's also >acceptable to discover a conflicting transaction, as that will >demonstrate that a particular participant has been dishonest (their >utxo was not "unspent"), allowing the overall protocol to make progress > > The question then is how much effort do you have to go to to make such a > protocol work? > > As an extreme example, you could always have each participant maintain > a dedicated amount of hashpower: eg, if each participant individually > controls 0.5% of hashpower, then having two honest participants would > give you a 75% chance of confirmation within 137 blocks (roughly a day), > even if your transaction failed to relay at all, and the only way to > prevent confirmation is for a conflicting transaction to be confirmed > earlier. Of course, needing to have 0.5% of hashpower would mean fewer > than 200 people globally could participate in such a protocol, and > requires something like $10M in capital investment just for ASICs in > order to participate. > > I think the next step from that pretty much requires introducing the > assumption that the vast majority of the bitcoin p2p network (both nodes > and hashrate) will accept your transaction, at least in a world where all > your collaborators are honest and don't create conflicting transactions. > You can limit that assumption a little bit, but without most p2p peers > being willing to relay your tx, you start having privacy issues; and > without most miners being willing to mine your tx, you start getting > problems with predicting fees. And in any event, I don't think anyone's > trying to make weird transactions here, just get their otherwise normal > transactions to actually confirm. > > I think the same approach used to detect double spend races by people > accepting zeroconf would work here too. That is setup a couple of > anonymous bitcoin nodes, let them sit for a couple of weeks so their > mempools are realistic, then when you broadcast a jointly funded > transaction, query their mempools: if your new tx made it there, it > likely made it to mining pools too, and you're fine; if it didn't, then > the transaction that's blocking it almost certainly did, so you can find > out what that is and can go from there. > > (If you don't see either your tx, or a conflicting one, then it likely > means the nodes that broadcasted your tx are being sybil attacked, either > because their peers are directly controlled by an attacker, or they've > been identified by an attacker and attacked in some other way; presumably > you could pick a couple of node that have been confirmed by both your > anonymous nodes' as valid and reachable, and connect to them to break > out of the sybil attack; if that doesn't work either, you probably need > to change ISPs or put your active node via a (different) VPN provider...) > > Your capital expenses are much lower that way: perhaps on the order of > $20/month to run a couple of nodes on AWS or linode or similar. > > But, you might say, what if I don't even want to run multiple bitcoin > nodes 24/7 indefinitely? Can we outsource that, like we outsource mining > by paying tx fees? > > That seems harder, particularly if you want to avoid whoever you're > outsourcing too
[bitcoin-dev] Preventing/detecting pinning of jointly funded txs
On Fri, Oct 28, 2022 at 03:21:53AM +1000, Anthony Towns via bitcoin-dev wrote: > What should folks wanting to do coinjoins/dualfunding/dlcs/etc do to > solve that problem if they have only opt-in RBF available? ref: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2022-October/021124.html So, having a go at answering my own question. I think ultimately the scenario here is: * you have a joint funding protocol, where everyone says "here's an unspent utxo that will be my contribution", collaborates on signing a transaction spending all those utxos, and then broadcasts it * everyone jointly agrees to pay some amount in fees for that transaction, targeting confirmation within N blocks * the goal is to have the transaction confirm, obviously; but it's also acceptable to discover a conflicting transaction, as that will demonstrate that a particular participant has been dishonest (their utxo was not "unspent"), allowing the overall protocol to make progress The question then is how much effort do you have to go to to make such a protocol work? As an extreme example, you could always have each participant maintain a dedicated amount of hashpower: eg, if each participant individually controls 0.5% of hashpower, then having two honest participants would give you a 75% chance of confirmation within 137 blocks (roughly a day), even if your transaction failed to relay at all, and the only way to prevent confirmation is for a conflicting transaction to be confirmed earlier. Of course, needing to have 0.5% of hashpower would mean fewer than 200 people globally could participate in such a protocol, and requires something like $10M in capital investment just for ASICs in order to participate. I think the next step from that pretty much requires introducing the assumption that the vast majority of the bitcoin p2p network (both nodes and hashrate) will accept your transaction, at least in a world where all your collaborators are honest and don't create conflicting transactions. You can limit that assumption a little bit, but without most p2p peers being willing to relay your tx, you start having privacy issues; and without most miners being willing to mine your tx, you start getting problems with predicting fees. And in any event, I don't think anyone's trying to make weird transactions here, just get their otherwise normal transactions to actually confirm. I think the same approach used to detect double spend races by people accepting zeroconf would work here too. That is setup a couple of anonymous bitcoin nodes, let them sit for a couple of weeks so their mempools are realistic, then when you broadcast a jointly funded transaction, query their mempools: if your new tx made it there, it likely made it to mining pools too, and you're fine; if it didn't, then the transaction that's blocking it almost certainly did, so you can find out what that is and can go from there. (If you don't see either your tx, or a conflicting one, then it likely means the nodes that broadcasted your tx are being sybil attacked, either because their peers are directly controlled by an attacker, or they've been identified by an attacker and attacked in some other way; presumably you could pick a couple of node that have been confirmed by both your anonymous nodes' as valid and reachable, and connect to them to break out of the sybil attack; if that doesn't work either, you probably need to change ISPs or put your active node via a (different) VPN provider...) Your capital expenses are much lower that way: perhaps on the order of $20/month to run a couple of nodes on AWS or linode or similar. But, you might say, what if I don't even want to run multiple bitcoin nodes 24/7 indefinitely? Can we outsource that, like we outsource mining by paying tx fees? That seems harder, particularly if you want to avoid whoever you're outsourcing too from associating you with the jointly funded transaction you're interested in. If you're *not* worried about that association, it's probably easy: just find some public explorers, and see if they list any conflicts in their mempool, or use the "broadcast tx" feature and see if it gives an error identifying the conflicting transaction. I think it's probably hard to make that behaviour a normal part of p2p tx relay though: if someone's trying to relay tx T but you reject it because of a conflicting tx C; then it's easy to tell the node that first relayed T to you about C -- but how does that information get back to the original broadcaster? One way would be to broadcast "C" back to whoever announced T to you, and let C propogate all the way back to whoever originally proposed T -- but that only works if everyone's running a mempool policy where there's a total ordering for tx replacement, ie for any conflicting txs, either T replaces C or C replaces T, and that's not something we have now or would have even with full RBF, and seems pretty hard to actually achieve. (And if it
