Author

Topic: Large pools - a threat to bitcoin network integrity? (Read 709 times)

legendary
Activity: 2506
Merit: 1010
would it not be possible to hijack the blockchain by compromising the top n pools?  If an attacker was able to gain 51+ % of the bitcoin network for a 10min period (the defined time for propagation of blocks through the network) they would be able to supress a legitimate transaction and have it invalidated/stomped out.

The attacker could supress a transaction for a block that the attacker has mined, but another miner might include the transaction in a later block.  If the attacker has a lot more than 50%, the blocks by other miners could be replaced.  Here's more on the subject:
 - http://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power
newbie
Activity: 30
Merit: 0
I took an interest in bitcoin last week and started reading as much as I could.  I love the idea from a philosophical and technology perspective and I am eager to see the community thrive (especially my small investment in a mining rig heh).

My question is meant to highlight a possible weak point in the integrity of the bitcoin network OR clear a misconception from my mind about the propagtion of trusted blocks and the trust modle of the P2P network.

QUESTION:
Since 'mining' pools such as BTCGuild and deebbit were not considered in the design of the trust model and distribution of work, would it not be possible to hijack the blockchain by compromising the top n pools?  If an attacker was able to gain 51+ % of the bitcoin network for a 10min period (the defined time for propagation of blocks through the network) they would be able to supress a legitimate transaction and have it invalidated/stomped out.

[1] According to this graph: http://bitcoin.sipa.be/speed-ever.png the total computational capacity of the bitcoin  network is just over 10G Hash. (*Note this number needs to be verified -- it seems low and the points on the graph grow dont grow linearly -- could be much higher than 10GHash but no higher than 50.  Clarification on this point would be nice is anyone has a better number)

[2]
Deepbit pool rate: 5320 Gh/s
BTC Guild pool rate: 2340.95 GH/s

If an attacker was able to compromise the TWO bitcoin work distribution servers at BTCGuild and then one at deepbit (I realize thee are probably load balanced but the compromise could be of DNS servers (social engineering?), or even a local attack at a data center (if you're a particular nation state that is against bitcoin for example).

The question summed up: Is 7600GH/s enough to disrupt the bitcoin network and if so is there a method of preventing this if pools continue to be such a significant portion of the vote?



Jump to: