Miner dilution attack on Bitcoin - is that something plausible?
7 messages, 18 to 19 Jun 2018. Started by Артём Литвинович.
Артём Литвинович, Alexander Leishman, Laszlo Hanyecz, Bram Cohen, Bryan Bishop, Eric Voskuil, Richard Hein
Dilution is a potential attack i randomly came up with in a Twitter arguement and couldn't find any references to or convincing arguments of it being implausible.
Suppose a malicious actor were to acquire a majority of hash power, and proceed to use that hash power to produce valid, but empty blocks.
As far as i understand it, this would effectively reduce the block rate by half or more and since nodes can't differentiate block relay and block production there would be nothing they can do to adjust difficulty or black list the attacker.
At a rough estimate of $52 per TH equipment cost (Antminer pricing) and 12.5 BTC per 10 minutes power cost we are looking at an order of $2 billion of equipment and $0.4 billion a month of power costs (ignoring block reward) to maintain an attack - easily within means of even a minor government-scale actor.
Is that a plausible scenario, or am i chasing a mirage? If it is plausible, what could be done to mitigate it?
-Artem
Well miners already regularly mine empty blocks. However, it is usually in the economic interest of the miners to collect transaction fees. This incentive should hopefully be enough to prevent miners from choosing to produce many empty blocks.
If a nation state attacker decides to allocate billions in resources to attack Bitcoin, then that is a bigger discussion. The risk there is double-spends, not empty blocks.
-Alex
On Mon, Jun 18, 2018 at 11:39 AM Артём Литвинович via bitcoin-dev < bitcoin-dev@lists•linuxfoundation•org> wrote:
Quoted text (28 lines)
> Dilution is a potential attack i randomly came up with in a Twitter > arguement and couldn't find any references to or convincing arguments of it > being implausible. > > Suppose a malicious actor were to acquire a majority of hash power, and > proceed to use that hash power to produce valid, but empty blocks. > > As far as i understand it, this would effectively reduce the block rate by > half or more and since nodes can't differentiate block relay and block > production there would be nothing they can do to adjust difficulty or black > list the attacker. > > At a rough estimate of $52 per TH equipment cost (Antminer pricing) and > 12.5 BTC per 10 minutes power cost we are looking at an order of $2 billion > of equipment and $0.4 billion a month of power costs (ignoring block > reward) to maintain an attack - easily within means of even a minor > government-scale actor. > > Is that a plausible scenario, or am i chasing a mirage? If it is > plausible, what could be done to mitigate it? > > > -Artem > _______________________________________________ > bitcoin-dev mailing list > bitcoin-dev@lists•linuxfoundation•org > https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev >
On 2018-06-18 18:34, Артём Литвинович via bitcoin-dev wrote:
Quoted text (2 lines)
> Suppose a malicious actor were to acquire a majority of hash power, and > proceed to use that hash power to produce valid, but empty blocks.
https://github.com/libbitcoin/libbitcoin/wiki/Empty-Block-Fallacy
Not sure what you're saying here. The block rate can't be particularly increased or decreased in the long run due to the work difficulty adjustment getting you roughly back where you started no matter what. Someone could DOS the system by producing empty blocks, sure, that's a central attack of what can happen when someone does a 51% attack with no special countermeasures other than everything that Bitcoin does at its core. An attacker or group of attackers could conspire to reduce block sizes in order to increase transaction fees, in fact they could do that with a miner activated soft fork. That appears both doable and given past things which have happened with transaction fees in the past potentially lucrative, particularly as block rewards fall in the future. Please don't tell the big mining pools about it.
On Mon, Jun 18, 2018 at 11:39 AM Артём Литвинович via bitcoin-dev < bitcoin-dev@lists•linuxfoundation•org> wrote:
Quoted text (28 lines)
> Dilution is a potential attack i randomly came up with in a Twitter > arguement and couldn't find any references to or convincing arguments of it > being implausible. > > Suppose a malicious actor were to acquire a majority of hash power, and > proceed to use that hash power to produce valid, but empty blocks. > > As far as i understand it, this would effectively reduce the block rate by > half or more and since nodes can't differentiate block relay and block > production there would be nothing they can do to adjust difficulty or black > list the attacker. > > At a rough estimate of $52 per TH equipment cost (Antminer pricing) and > 12.5 BTC per 10 minutes power cost we are looking at an order of $2 billion > of equipment and $0.4 billion a month of power costs (ignoring block > reward) to maintain an attack - easily within means of even a minor > government-scale actor. > > Is that a plausible scenario, or am i chasing a mirage? If it is > plausible, what could be done to mitigate it? > > > -Artem > _______________________________________________ > bitcoin-dev mailing list > bitcoin-dev@lists•linuxfoundation•org > https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev >
On Mon, Jun 18, 2018 at 3:40 PM, Bram Cohen via bitcoin-dev < bitcoin-dev@lists•linuxfoundation•org> wrote:
Quoted text (13 lines)
> Not sure what you're saying here. The block rate can't be particularly > increased or decreased in the long run due to the work difficulty > adjustment getting you roughly back where you started no matter what. > Someone could DOS the system by producing empty blocks, sure, that's a > central attack of what can happen when someone does a 51% attack with no > special countermeasures other than everything that Bitcoin does at its > core. An attacker or group of attackers could conspire to reduce block > sizes in order to increase transaction fees, in fact they could do that > with a miner activated soft fork. That appears both doable and given past > things which have happened with transaction fees in the past potentially > lucrative, particularly as block rewards fall in the future. Please don't > tell the big mining pools about it. >
Bram, actually I thought the previous discussions determined that less than 51% hashrate would be required for certain soft-hard-forks employing empty blocks?
I don't have a specific reference: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2016-February/012377.html https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2016-February/012457.html https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2016-December/013332.html
- Bryan http://heybryan.org/ 1 512 203 0507
https://github.com/libbitcoin/libbitcoin/wiki/Other-Means-Principle
Quoted text (17 lines)
>> On Mon, Jun 18, 2018 at 11:39 AM Артём Литвинович via bitcoin-dev <bitcoin-dev@lists•linuxfoundation•org> wrote: >> Dilution is a potential attack i randomly came up with in a Twitter arguement and couldn't find any references to or convincing arguments of it being implausible. >> >> Suppose a malicious actor were to acquire a majority of hash power, and proceed to use that hash power to produce valid, but empty blocks. >> >> As far as i understand it, this would effectively reduce the block rate by half or more and since nodes can't differentiate block relay and block production there would be nothing they can do to adjust difficulty or black list the attacker. >> >> At a rough estimate of $52 per TH equipment cost (Antminer pricing) and 12.5 BTC per 10 minutes power cost we are looking at an order of $2 billion of equipment and $0.4 billion a month of power costs (ignoring block reward) to maintain an attack - easily within means of even a minor government-scale actor. >> >> Is that a plausible scenario, or am i chasing a mirage? If it is plausible, what could be done to mitigate it? >> >> >> -Artem >> _______________________________________________ >> bitcoin-dev mailing list >> bitcoin-dev@lists•linuxfoundation•org >> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
It's important therefore to ensure that everyone can make ASICs, IMHO.
On Mon, Jun 18, 2018 at 2:34 PM, Артём Литвинович via bitcoin-dev < bitcoin-dev@lists•linuxfoundation•org> wrote:
Quoted text (30 lines)
> Dilution is a potential attack i randomly came up with in a Twitter > arguement and couldn't find any references to or convincing arguments of it > being implausible. > > Suppose a malicious actor were to acquire a majority of hash power, and > proceed to use that hash power to produce valid, but empty blocks. > > As far as i understand it, this would effectively reduce the block rate by > half or more and since nodes can't differentiate block relay and block > production there would be nothing they can do to adjust difficulty or black > list the attacker. > > At a rough estimate of $52 per TH equipment cost (Antminer pricing) and > 12.5 BTC per 10 minutes power cost we are looking at an order of $2 billion > of equipment and $0.4 billion a month of power costs (ignoring block > reward) to maintain an attack - easily within means of even a minor > government-scale actor. > > Is that a plausible scenario, or am i chasing a mirage? If it is > plausible, what could be done to mitigate it? > > > -Artem > > _______________________________________________ > bitcoin-dev mailing list > bitcoin-dev@lists•linuxfoundation•org > https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev > >