Bitcoin Forum

Should we be concerned that over 50% of the mining power is controlled by two pools? https://blockchain.info/pools

This seems like a really bad concentration of power for a network that is supposed to be decentralized.  Sure, it's handy when you need to fix a bug in the system, but it's also a weak point that can be exploited.  Thoughts?

No, it is not a concern.  Why?  Are you concerned?  If so, can you explain exactly why you are?

Because it makes it easier to pull off a 51% attack?

Any of the major pools with cap themselves if they even got close. BTCGuild has had to do that a few times. Nothing to worry about unless they suddenly announce a merger.

what if the same guy already owns them both ?

It is not a good thing that is for sure.

Do you even know what a 51% attack is?  Do you know what can be done using it?  Do you know what would happen if a pool tried to do it?

Let's do a mind experiment.  Assume BTC guild, right now, has 60% of all the hashing power.  Now explain to me exactly what they would do in order to pull off a "51% attack", exactly what they would accomplish by doing it and most importantly exactly why they would do it.

Well if you are that concerned start your own pool that offers paying transaction fees, mines Altcoins and has low pool fees.

No one will mine in a pool that pays less....

This ^^^^

My biggest fear is that the Star Trek Enterprise will come back through time to our century and pull off a 51% attack.

I think they are only ones with their super advanced technology that could pull it off.

Hell, Commander Data could probably do it with his positronic brain.  ;)



 

the whole point of bitcoin is being a decentralized currency... yeah, I don't think any mining pool will want to have control over the blockchain, that would probably destroy bitcoin and obviously miners don't want that.. but I'm still concerned  :-\

what happen with p2pool?

Please don't talk down to people like that, it's not constructive, and given the nature of a lot of the conversations on this board, I don't think this was an invalid question.

Answers:

Yes. Yes.  : https://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power

No.  I'm not sure what would happen.  If you have a good idea, and think that the consequences make it not something to worry about, I think many people would like to know that information.  That's pretty much why I asked.  Should we be concerned?  It sounds like no, but you haven't really explained why.

Here are my conjectures:

It seems like the double spend is probably not practically useful, since it's less of a double spend and more of a "spend and then unspend", which only helps if you're getting something from someone else for spending the bitcoin.  And in that case, the other party would start squawking up a storm about it, so it  wouldn't go unnoticed.

Forcing transactions not to be confirmed seems like it might be something to be concerned about.  It could be hard to determine why its happening, and could be used as a weapon against people that piss off the pool.  Of course, that requires the pool to know which transactions belong to that person, which is sorta the whole point of anonymity in the blockchain.... but I'm not sure how hard it would be in practice.

Preventing other miners from getting blocks seems like the most likely attack vector, since it could be the hardest to recognize.  The top two pools could just give themselves 10% more block awards by rejecting 10% of blocks submitted from outside their pools, and waiting until one of them finds the block.  I'm not sure how obvious that would be to people watching transactions on the block chain.  The benefits are obvious - they get 10% more block awards than they're entitled to.

I'm relatively new to bitcoin, so some or all of this may be off, feel free to tell me where I've gone wrong, after all, I am trying to learn.

They would be able to do everything.

Maybe modify just some transactions... because someone pay them for that... you know, corruption... something wich will pass unnoticed but important...who know... no proof that "they" did it...

Support p2pool

A 51% Attack doesn't have to be intentional, does it? If any pool had 60% of the hashing power, an attack would be forthcoming. The pool doesn't have to try to do it, it will, which is why the pools cap themselves. Blocks other miners solve would be orphaned left and right... The longest chain always win the conflict, right? That isn't to say it was done intentionally or with malice. But whoever has the upper hand would benefit from the block rewards and transactions fees of those resolved 'conflicts.'  On the other hand, if a pool wanted to orchestrate an attack... well, I guess that's another topic.

You did good, hehe.

No, that's not how it works. There is no problem with a pool having more than 50% of the hash power as long as they don't abuse that power.

Can you explain how it works, or how the pool can prevent from abusing that power? Can that power be abused - oh, hell yes! I'm not talking about the many things that a pool powerhouse has the ability to do.... But how can said pool prevent from NOT taking other miners' block rewards and transaction fees? This isn't intentional - As the authority on the network (so to speak), clients will believe the pool's hashed transaction block over any other. That means the pool will win all transaction conflicts, and get the rewards.

I'll accept what you said that there's no problem with having more than 50% of the hashing power, but you've got to explain just a lil... after all, I'm still an infant in this world too.

Arguably it is still a marketplace problem in that there may be lack of confidence that the pool is honest, will stay honest, and won't be compromised by (internal or external) dishonest actors.

The pool model is highly problematic and isn't solved by starting a new low-fee pool as someone else suggested because a new (presumably smaller) pool will have payout variance that is too high. There are some very nice smaller pools right now that don't even reliably solve one block per week. The incentive for the individual miner is always to sign up with the bigger pools to get consistent payouts, and that's bad for the system as a whole.

The way out of this trap is with p2pool (or build a new system like it if you don't like p2pool).

Look at you OP... got me challenging veteran members and I don't know what the hell I'm talking about  :P

No.

Epic confusion.

A 51% attack is then the attacker blockchain voluntarily ignore every other blocks. An honest 51% holder would accept the blocks found by other people.

Here's a thought experiment...How about have just one big pool and every miner is automatically a member of that pool.  That pool would of course get every block award.  The pool operator sends out the appropriate fractions of block award based on the hashing contribution of each miner.  As long as the pool operator didn't abuse their power this would be the most fair system.  Of course that degree of centralization would be ripe for abuse and corruption, if history is any guide.  Even if the operator *wanted* to be fair others could find ways to coerce the operator to allocate shares unfairly and cook the books to hide it by manipulating the statistics showing how many miners there are and their relative hash power.  It seems clear to me that having 1 pool with %100 of the hash power is bad.  That raises the question: Is 2 pools split 80%/20% any better?  How about 3 pools 60/35/5%?  Actually, when bitcoin really hits the big time (e.g. $5000 per coin?) then I believe any sufficiently large pool operator becomes a target of the sorts of people who are good at forced manipulation (e.g. governments, criminal organizations).


I think as long as people are in charge of operating mining pools there are opportunities for extortion, abuse, and corruption.

P2pool is doing fine. It's on the chart linked above, but it's too small. The problem with p2pool is that the interface is not really user friendly and it can be finicky to get working especially with certain miners. If you really want to help the pool concentration problem, volunteer to help with p2pool. That doesn't have to be coding, it can be better guides and cheat sheets, pretty stats sites, etc.

Even if the 2 pools hypothetically pulled off a 51% attack, who gains? If they do it, the market will instantly crash, the value for their saved coins is nothing.

Cash out before the panic crash or invest in whichever altcoin is likely to rocket upon news of a 51% attack on bitcoin. Millions to be made.

So does anyone here are really know what 51% attack is? The worst thing it could do is to ignore some transactions by request even though it will be obvious to everyone and because of this it will immediately trigger the lowering price of bitcoin and the pool will collapse, all its participants will leave it as a free will. No one want to mine coins when its price lowering and the pools are almost identical.
51% attack do not allow you to steel money from someone else.
In case of any possible 51% attack market will not even feel this it will be like a sudden drop of price for a minute and it will raise back in hours.

Epic confusion.

A 51% attack is then the attacker blockchain voluntarily ignore every other blocks. An honest 51% holder would accept the blocks found by other people.
[/quote]

Well I'm sure as hell confused now if I weren't before. Are you talking about a blockchain fork, pseudo blockchains and such? No... wait, that can be done if the powerhouse pool wanted to do such a thing... but just to make sure I understand.... you're saying that if the pool with the majority of computing power is honest, it will accept other pools' blocks? Forgive my confusion... I just don't understand what this has to do with the price of rice in China.

Yes I always thought part of the allure of BTC was how many many people would mine and make sure there was never a 51% attack/collusion to maintain fairness.  I suppose pool members could switch to a different pool if the pool gets too big and the owner abuses their power, assuming those miners care about BTC as a whole.

Well I'm sure as hell confused now if I weren't before. Are you talking about a blockchain fork, pseudo blockchains and such? No... wait, that can be done if the powerhouse pool wanted to do such a thing... but just to make sure I understand.... you're saying that if the pool with the majority of computing power is honest, it will accept other pools' blocks? Forgive my confusion... I just don't understand what this has to do with the price of rice in China.


[/quote]

dont believe the hype
the possibility of a  51% attack has been analysed ad hominum already and it will not result in everyones previously saved coins
being suddenly worth "nothing" as someone just claimed :D
this has been talked about for years even when satoshi was still an active developer i think  and it hasnt changed the general opinoion much
if there was any chance  all the  previously mined btc would become worthless due to a 51% attack i think they would have factored it in already :)

its not like were talking about monkey nuts here ,people have millions or even possible billions at stake so i think theyve analysed the risk sufficiently for now anyway

Wrong.
It can. It can rewrite the whole blockchain.

rewrite the entire bloackchain from the genisis block until now  ? i dont think so

my undertanding was that such an attack would only work in the short term for a few recent blocks 
but if the network was this vulnerable it would have happened already

If the two pools collude for a 51% attack, what would they gain?  A double spend?  What would the double spend get them?  

Suppose it was for something like a Lamborghini which would be a good sized purchase.  Probably won't happen because the dealer is going to wait a reasonable number of confirmations - more than 6 no doubt - or use escrow.  So large purchases are pretty much out.  So it would be for something less.  What happens after a double spend for $10,000 then?  The double spend is seen (whomever the double-spend is against will point it out ASAP if it is of any size and/or it will be noticed by others), and the pools lose 95% of their members within hours or days.  The pool is done and the operator loses a lot of recurring income from fees.  So it seems unlikely.

Alternatively, what happens to the bitcoin price if the two pools collude and the miners stay with them to help keep the collusion going?  The price of bitcoin plummets and the pools and miners lose out again.  

No rational miners will stay with pools colluding so any attack will be short-lived, so it would only be attempted for a large reward, but no large transaction will be accepted without many confirmations or escrow making a double-spend unlikely to be profitable for the pools.

Perhaps a state-sponorsored pool might try it solely to destroy bitcoin, but every day that becomes less likely due to the increasing hash rate. Perhaps 3 years ago pre-GPUs it would have been easy.  With the GPUs, it was still possible, but 2013's ASIC explosion has made it much more difficult and much more expensive to even consider.  With the miners shipping now and next year it will become even more unlikely.

Obviously more decentralization is better as various people have said with regard to p2pool, but some ASICs can't mine with it reliably, and it is not nearly as easy to set up.  I believe this will improve with time.  I certainly hope it will because p2pool-like mining is an important part of the future of bitcoin.

Regarding the quote above, people get frustrated because the same question has been asked and discussed hundreds of times on here.  His questions are valid because if the person can't answer them, then they may not be clear on some of the concepts, and so the question comes from misconceptions.  

;-)

Can I do this one!!! Give me a shot! --

Epic Confusion.

No, IT (not really sure what you're referring to here) can't rewrite the whole blockchain, but someone can create a fake or alternate blockchain with all the previous transactions up to the point... wait, why am I even explaining this?

You're confused; and I thought it was me. Consult with Google on Blockchain forks.   

Concur.

There's more to gain than just double spending. Double spending is something the pool can do if it chooses. If they created a fork in the blockchain, they could double spend some coins and get away with it, which is what I think another post was referring to, coins getting lost forever. But say for example we're the powerhouse pool and we wanted to be bad... {Girl power!} Crap, am I by myself here? Damn it.

Anyhoo, we created our version of the blockchain and it has all real transactions up until yesterday's. Today's transactions are bogus btw... we sent some coins there, and over there. Since we're the authoritative bitch (majority of computing power) theoretically we could force other pools to point to our 'fake' blockchain and accept our made up transactions from today. But I suspect we'd eventually get caught doing it this way. We should probably just pluck random, semi-large transactions.

I think we're getting too caught up with the term attack. We want to think of a 51% attack as being intentionally malicious... Agreed, a pool out for gain intentionally exploiting their power/network can do some damage. But what about the unintentional consequence of having over 51% of the hashing power? Are we just going to dismiss this and focus on the bad stuff they can accomplish?



Anybody? Or am I over-analyzing. For the purpose of your response, let's say it's an "honest pool."

It is already described in the wiki what the 51% attacker can and can't do:

https://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power

The attacker can:
    Reverse transactions that he sends while he's in control. This has the potential to double-spend transactions that previously had already been seen in the block chain.
    Prevent some or all transactions from gaining any confirmations
    Prevent some or all other miners from mining any valid blocks

The attacker can't:
    Reverse other people's transactions
    Prevent transactions from being sent at all (they'll show as 0/unconfirmed)
    Change the number of coins generated per block
    Create coins out of thin air
    Send coins that never belonged to him

Note: So the attacker can potentially do double-spend, but, please correct me if I am wrong, this only applies if it is really one attacker with 51% hashpower, in other words, he can deploy his own patched "dishonest" client to all his nodes. But can this be applied to pools? How will the pool owner force the attacking behavior to his poll members?

ok
Someone tracked down the one guy who owns BTC guild and put a gun/warrant to his head.

I think a bigger question is the how.

What happens if a coalition of governments publishes a modified client that allows reversing transactions and freezing of funds by a central authority regulated by the WTO.
They mandate by law that all must use this client.
They raid the biggest guilds for IP address of their users to track down and arrest those who do not make the switch.
90% of users & 90% miners switch to this new version. What happens to the 10% that don't switch? does the currency split up into 2 seperate ones or are the 10% who don't switch suddenly unable to get any transactions to complete?

Also, the above assumes it is still fairly small, in a few years with major banks using it as well as many merchants accepting it they can merely release a law saying to use the government made versions, provided to you by your bank, and 99% of user switch over because they aren't libertarians. No need for raids and arrests

Hypothetical 2:
What happens if instead of the above, you have JP morgan, intel, samsung, etc build giant mining facilities that end up controling 90% of mining. Those companies are approached by governments who bribe/coerce/cajole/whatever them into using such a modified client. Such that 90% of mining is done by new client, but 99% of users are still using the old client.

PS. these aren't rhetorical questions meant to insinuate that bitcoin is doomed. I am actually curious as to what the answer is. If the protocol is designed in such a way that the above attempts will merely result in governments wasting a lot of its time and money and failing to do anything I would be very happy

Wouldn't surprise me. And I guess some of us would become mining criminals. But if there were a 'split' technically you'll have two coins, right? So they can keep client they made.

Don't think the government would have to do much bribing. But all they'd have to do is build the biggest, baddest miner to take over.

• Addresses would become the new social security number.
• We'd be taxed of course.
• All transactions would be monitored; anything suspect and it'll never get confirmed.
• Our balance would be allowed to go in the red.
• Bitcoin loan companies would sprout out like liquor stores.
• Uncle Sam will always get its cut first. Hope you don't owe taxes or gov't debt.
• They'd create a world wide lottery via the blockchain. No need to go out and buy tickets anymore.
• It's the gov't so expect slow confirmation times.
• Blockchain shutdown during holidays perhaps.
• It's the gov't so expect lots of errors.

Man, that list could go on and on!

if everyone created a back-up pool, then there shouldn't be a problem.

Temptation for skimming hash rates, now there is a potential problem - does anyone carry out an impartial audit of these pools?

No concern because im a blind bitcoin follower.

The whole blockchain is now very long... it will require little bit more computational resources than just 51%. 51% mining power allows you to just ignore some recent transactions. Also, in case of mining pools, if the pool administrator will ask nodes to do some work to prolong old blocks it will be transparent to nodes. Everyone knows what blocks pool is trying to prolong so the admin will be required to explain this behaviour. Rewriting the whole blockchain is possible only through the attack on SHA2 hash function. So it is better to worry about this.

Who knows... every now and then new supercomputers are being created with massive computational power. Will create interesting scenarios in the future.  ;D

but the current network is only being strengthed ,not weakened as btc  becomes more famous

its already bigger than anything anyone could attack it  with and its growing with every new bitcoiner who joins
and gpus and asics  are sold out everywhere 

nobody is going to invent a supercomputer anytime soon that can outperform the btc network

You would be breaking the rules of time traveling if you made such changes.
Note for everyone: You can recover from a 51%.

Haha, very interesting thinking :)

This has all been discussed ad nauseum http://en.wikipedia.org/wiki/Ad_nauseam (not ad hominem http://en.wikipedia.org/wiki/Ad_hominem, unless that was meant as a joke, then, good one!)

Here is a pretty good thread that I moderated a while back discussing some of the technical aspects of a fork of this nature:

https://bitcointalk.org/index.php?topic=352734.0;all

EDIT:  BTW the other thread mentioned in the above referenced moderated thread, the one that motivated me to open the moderated thread is:

https://bitcointalk.org/index.php?topic=349869.0;all

which is a pretty good thread on the whole "attack" idea being discussed (again) in this thread.