51% Attack And The Effects It Poses

[Asiska02]

Sequel to my previous post on bitcoin 51% attack and its effect, this is an updated (new) post of the previous one after taking correction and done more research about it.


Blockchain transactions that have already been mined are modified by the 51% attack.
The only way this is feasible is if one's mining power is greater than other miners'. The ability to mine the next block is determined by one's mining power in comparison to other miners. Due to the significant costs associated with controlling more than 50% of the network's total computing capacity, a 51% attack has never been attempted in the history of bitcoin and is unlikely to ever happen.

Here's an illustration of how it functions:



The above image shows (5 miners) whose total computational power is 500,000 hashes per second (h/s). Different hashes per second were generated by each of the 5 miners as seen in the image above. The first miner was able to produce 175,000 h/s, followed by the second miner that generated 100,000 h/s, the third generated 75,000 h/s, the fourth generated 125,000 h/s, and the fifth generated 25,000 h/s. They’re all struggling to find a hidden number that is unknown to them all and the person that gets the correct number will be able to unravel the puzzle (mine the next block). The likelihood of mining the next block is 35% for the first miner, 20% for the second, 15% for the third, 25% for the fourth, and 5% for the fifth.
The miner with the highest percentage of the total hashing power has the likelihood of finding the number first. Meanwhile, this is just a probability as the person with the lowest hashing power can also be lucky to unravel the puzzle first before others.

The bitcoin network's nodes have produced 190 million tera hashes per second as of February 2022. A tera hash is equal to 1,000,000,000,000. With 51% computational power, a miner or node (a group) may produce 96.9 million tera hashes per second, increasing his chances of creating the longest chain.
Chain reorganization occurs when nodes receive details of a valid chain with more proof of work than the chain they are currently following. A miner who has more than 50% of the hashrate will not have the ability to alter the upcoming block because it is not known yet, but can modify previous block in the blockchain by replacing them in favor of malicious chain with more work.

The bad person (who owns 51% hashrate power) has more control over the block reward and the ability to add the block to the blockchain with more than 51% hashrate power. The miners following the correct rule-see this and decline to accept the block. And as a result, the bad guy's work is pointless. You must manage several nodes and operate them simultaneously if you want to swiftly confirm your block and upload it to the blockchain.
Currently, a miner owning 51% hashrate and using it for an attack swiftly in seconds is impossible because at least 20% of the hashrate is owned by public companies. Attempting this will only cause the attacker problems and will face many lawsuits for such action.



Source: https://learnmeabitcoin.com/technical/51-attack

The graphic above shows how much mining capacity every of the miner has. Only one miner (GHash.io), with a total mining power of 38.24% in June 2014, was close to crossing 50% threshold. This poses a serious risk of security for people who are learning about bitcoin interestingly for the first time. The distribution of power has improved, and that company is no longer running.
You can check how hash rate distribution has been shared over the years in this clip:
https://twitter.com/in3rsha/status/1159468260661895169

If a 51% attack takes place, what will the consequences be for the bitcoin network?

Effects and consequences
In the circumstances mentioned above, this person might take advantage of that to carry out the following transactions while in control:

* Use your bitcoins for reverse transactions and double spending.
* Changing confirmations for any transaction in the block chain that has already been confirmed before by other miners.
* Not accepting transactions that should be authenticated normally.
* Preventing fellow miners not to mine valid blocks.

The great inventor of bitcoin once said:

“As long as a majority of CPU power is controlled by nodes that are not cooperating to attack the network, they’ll generate the longest chain and outpace attackers”. – Satoshi Nakamoto

51% might be possible theoretically but as far as the bitcoin network is concerned, 51% is not practically feasible.


Source: https://bitcoin.org/bitcoin.pdf
Source: https://medium.facilelogin.com/51-attack-a72d26786b23
Source: https://academy.bit2me.com/en/51-bitcoin-attack/
Source: https://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power
Source: https://learnmeabitcoin.com/technical/51-attack

[hatshepsut93]

When people talk about 51% hashpower, they often only think about the hardware, but power is hardware + energy. Energy that has to be supplied to continue the attack. If the attacker stops, the honest miners could continue to mine on the original chain and eventually overtake attacker's chain. This can undo the damage from the double spending. And as denial of service the 51% is so expensive that no one can afford it for long.

[KingsDen]

When people talk about 51% hashpower, they often only think about the hardware, but power is hardware + energy. Energy that has to be supplied to continue the attack. If the attacker stops, the honest miners could continue to mine on the original chain and eventually overtake attacker's chain. This can undo the damage from the double spending. And as denial of service the 51% is so expensive that no one can afford it for long.

Most times I let things be without further disturbing myself about them. When I was new to this forum, I read the 51% attack theory and was afraid. I had to personally research more to under how possible it could be. But after disturbing myself for much, I understood that if this attack was possible, it would have happened since before I knew bitcoin or at least during the last bear that so bitcoin as high as $69k. That attack is not possible and I have stopped to disturb myself about it. I have also stopped to disturb myself on what happens if the last btc is mined etc. Those that will get to the bridge will surely cross it.

[NotATether]

And as denial of service the 51% is so expensive that no one can afford it for long.

Well, the only people who can do something like that are those with massive energy reserves and who don't have to pay for electricity. Because a 51% attack is an extremely high-risk venture for attackers, because the electricity costs will be millions of times higher than for the average mining farm.

So only governments with impunity on how their budget is spent would attempt something like this. Not even the NSA can try it, because the energy consumption would immediately become obvious (and then they'd come under attack themselves by green energy activists, and in particular those in Congress). So, PoW is protected by the status quo.

Whereas PoS which is loudly touted as "environmental-friendly" and "green" by a lot of people is more susceptible to subverion: All the agencies have to do, is seize the bitcoin funds of some drug dealer or money launderer and lock them in mining.

[Charles-Tim]

I always remember this Andreas Antonopoulos YouTube video: https://youtu.be/ncPyMUfNyVM

Even if 51% attack occur on bitcoin blockchain, the attackers will at last be a loser because the kind of mining hashrate that will be 50% more than the present total bitcoin mining hashrate would become known and would be taken down, yet bitcoin mining hashrate continue to increase. No attacker would want to do what he is not capable of doing, the attackers would also know it won't be profitable.

[OcTradism]

To have enough at least 51% hashrate on the network, it is not a cheap cost. The cost would be bigger because if you are attacker, you will not bet that you will have a fluctuation between 51% and 49%, the percent of hashrate you control should be bigger than 51%.

More important, there are many full nodes on the network will detect your attack very fast and they will react accordingly.

Exchanges will react fastly too. They even react more fastly than with altcoins. Coinbase detected Deep Chain Reorganization Detected on Ethereum Classic (ETC)

They should have stronger detective mechanisms for Bitcoin because if they allow attacks to happen, they will bear serious loss.

[o_e_l_e_o]

Chain reorganization occurs when nodes receives a new block in the longest chain, they deactivate the previous longest chain and update it with the new block provided. It is not certain that a miner who has more than 50% of the hashrate will alter the upcoming block.

This statement is very muddled.

Chain reorganization occurs when nodes receive details of a valid chain with more proof of work than the chain they are currently following. This would usually, but not necessarily, be a longer chain. They don't "deactivate" the chain they are leaving - they simply leave it. The chain still exists and is entirely active for anyone else who chooses to follow it. Further, "altering the upcoming block" doesn't really make sense. Nobody knows what the next block will be until it is mined, so it is impossible to "alter it". A 51% attacker could instead alter already existing blocks by replacing them in favor of a malicious chain with more work.

The bad person (who owns 51% hashrate power) has more control over the block reward and the ability to add the block to the blockchain with more than 51% hashrate power. The miners following the correct rule-see this and decline to accept the block. And as a result, the bad guy's work is pointless. You must manage several nodes and operate them simultaneously if you want to swiftly confirm your block and upload it to the blockchain.

This isn't right either. A 51% attacker's blocks will still be entirely consensus valid (unless they do something really dumb) and therefore accepted by all other nodes who are following the consensus rules. And operating a single well connected is more than enough to have your winning block propagated and confirmed.

[stompix]

If the attacker stops, the honest miners could continue to mine on the original chain and eventually overtake attacker's chain. This can undo the damage from the double spending.

And what will happen to all the transactions that were present on chain a and not on chain b and the other way around?
But never mind that you're assuming that the attackers will just mine their chain and that loyal miners will not abandon their older smaller chain and only continue to build on it even when both are public, this won't happen since most nodes will reject the smaller chain.
So no, if this shit really happens it will be a devastating blow, there is little recovery from this to be done, not when billions are flowing left and right each week.

And as denial of service the 51% is so expensive that no one can afford it for long.

The daily reward for mining is currently 18 million dollars, that is what miners get to run their mining gear and offset the equipment purchase, so the actual energy costs should be way lower than that, let's do with 15 million, half a billion a month, or a Nimitz class each year.
Compared with 400 billion in market cap, it's just a tiny fraction, not really an Achilles heel, but still somewhat of a weakness during these times.

So only governments with impunity on how their budget is spent would attempt something like this. Not even the NSA can try it, because the energy consumption would immediately become obvious (and then they'd come under attack themselves by green energy activists, and in particular those in Congress). So, PoW is protected by the status quo.

You have to admit is quite funny:
- we talk about how much energy Bitcoin mining burns, it's way less than Christmas lights!!!!!!
- we talk about somebody executing an attack, an attack would require so much energy, not even governments would be able to pull it off!

If we go full tinfoil hat in order for the NSA to pull this off, assuming they would have the 1.5 million s19XP already in place all they need is to "detect" something being wrong at Grand Coulee, "shut the power" for two hours and that damn alone would be able to run all those miners.
But as always, the main question is, why would they do it?

Anyhow, I don't see any attacker trying to do this, why spend resources on trying to destroy it when you could spend those on actually making a ton of money, and even if you somewhat manage to "destroy" bitcoin with it, it will just leave a hundred thousands alternatives out there.

[Asiska02]

This statement is very muddled.

Chain reorganization occurs when nodes receive details of a valid chain with more proof of work than the chain they are currently following. This would usually, but not necessarily, be a longer chain. They don't "deactivate" the chain they are leaving - they simply leave it. The chain still exists and is entirely active for anyone else who chooses to follow it. Further, "altering the upcoming block" doesn't really make sense. Nobody knows what the next block will be until it is mined, so it is impossible to "alter it". A 51% attacker could instead alter already existing blocks by replacing them in favor of a malicious chain with more work.

Getting to understand the operation more on what happens during chain reorganization and 51% attacker’s capability of altering blocks in the blockchain.
OP updated!

This isn't right either. A 51% attacker's blocks will still be entirely consensus valid (unless they do something really dumb) and therefore accepted by all other nodes who are following the consensus rules.

A 51% attacker doesn’t follow the consensus rule, won’t the operation be known to other miners immediately assuming he doesn’t do anything dumb to be noticed?

And operating a single well connected is more than enough to have your winning block propagated and confirmed.

Operating a single well propagated nodes will cost a lot. Is it something that can ever be feasible at all? The other miners won’t dictate any fowl connection with the node possessing such power? and then mining a block swiftly in the chain?

[o_e_l_e_o]

A 51% attacker doesn’t follow the consensus rule, won’t the operation be known to other miners immediately assuming he doesn’t do anything dumb to be noticed?

My point is that no 51% attacker would do this.

You can churn out as many invalid blocks as you want, and the rest of the network will ignore them. There's no point going to the huge cost and effort of accumulating 51% of the hashrate to then mine a chain with more proof of work but which contains invalid blocks which will be rejected. All the blocks a 51% attacker generates are still perfectly valid and will be accepted by the rest of the network. There is no process by which nodes can reject blocks based on who found them, or indeed, even say for sure who found them, if the miner doesn't publicize that.

Operating a single well propagated nodes will cost a lot. Is it something that can ever be feasible at all? The other miners won’t dictate any fowl connection with the node possessing such power? and then mining a block swiftly in the chain?

You are confusing miners and nodes. It is trivial to run a single well connected node. You can do it on almost any home computer.

[Asiska02]

A 51% attacker doesn’t follow the consensus rule, won’t the operation be known to other miners immediately assuming he doesn’t do anything dumb to be noticed?

My point is that no 51% attacker would do this.

You can churn out as many invalid blocks as you want, and the rest of the network will ignore them. There's no point going to the huge cost and effort of accumulating 51% of the hashrate to then mine a chain with more proof of work but which contains invalid blocks which will be rejected. All the blocks a 51% attacker generates are still perfectly valid and will be accepted by the rest of the network. There is no process by which nodes can reject blocks based on who found them, or indeed, even say for sure who found them, if the miner doesn't publicize that.

Now I get what you mean. It is extremely futile to continue mining a chain that has greater proof of work but contains invalid blocks that will still be rejected; the miner will be wasting their time and effort.

[hatshepsut93]

If the attacker stops, the honest miners could continue to mine on the original chain and eventually overtake attacker's chain. This can undo the damage from the double spending.

And what will happen to all the transactions that were present on chain a and not on chain b and the other way around?
But never mind that you're assuming that the attackers will just mine their chain and that loyal miners will not abandon their older smaller chain and only continue to build on it even when both are public, this won't happen since most nodes will reject the smaller chain.
So no, if this shit really happens it will be a devastating blow, there is little recovery from this to be done, not when billions are flowing left and right each week.

If someone mined a new chain that goes 10 blocks back, it means they didn't participate in mining those 10 blocks. Wouldn't honest miners want to protect the rewards that they have mined and try to outrace the attacker?

[o_e_l_e_o]

Now I get what you mean. It is extremely futile to continue mining a chain that has greater proof of work but contains invalid blocks that will still be rejected; the miner will be wasting their time and effort.

Correct. A 51% attacker will mine a completely valid blocks which will be accepted by the rest of the network. Otherwise there is no point to the attack at all.

Wouldn't honest miners want to protect the rewards that they have mined and try to outrace the attacker?

Which raises an interesting question - are these honest miners still honest? Surely honest miners should just follow the chain with the most work, regardless of who generated that work? Once miners start choosing to follow a minority chain to protect their own rewards, are they still honest?