Bitcoin Forum

So seeing as there are only like 7771 nodes (https://getaddr.bitnodes.io/) running, couldn't a wealthy individual or company convince a large majority of people running nodes (not necessarily miners) to, say for instance, switch their node software to the one this individual sends them?

For example, let's say someone offers full node operators, $200 a month to switch over to their 'slightly modified' node software. Maybe they bill it as just a simple data aggregation thing, or something that seems more innocuous that will benefit bitcoin, maybe it's even better software.

Anyway, so once they have this control over a large number of nodes, then they can implement software changes and such that might benefit their plans or they could shut it down suddenly or do other nefarious things.

Would this have an impact on Bitcoin as a whole? How vulnerable are we to node control as opposed to hash control?

All that wealthy individual needs to do is open up Pool, set 0% fee, add something like merged mining, and wait for hash to come to him. No need to bribe anything or anyone. Sheeps will fill him up to 100% if needed. Buying nodes or ASIC's to do harm to network is for newbie evil doers (Like Dr. Evil asking for 1 MILION DOLLARS! https://www.youtube.com/watch?v=cKKHSAE1gIs&feature=kp for those less fortunate that have no idea what i'm talking about)

Isn't that effectively the GHash situation? Small number of nodes feeding large number of miners. Controls hash but not network. Or were you thinking a p2pool setup? Anyway I wanted to investigate owning network and what effect that could have.

Yes, it is. Ghas situation. Do you job, do it great, and you get ability to attack network basicly for free, or even get payed for it (GHash is making crazy amount of money). When i see how people are dreaming of other kind of attacks, it's just funny. Like anyone needs to invest into hash power....how silly is that  :D

Well, that's comforting.  Thanks Ghash.

-B-

Sell the node. Add two more real nodes. Sell them too. Then add four new nodes...

As Bruce Schneier once said, security and convenience are almost always a tradeoff. SPV clients are convenient so everybody uses them. As a result, only 7771 nodes. I think it would be possible to control the majority of full nodes, but instead of paying people off, buy a bunch of cheap VPSs.

He was referring to ghash

How would you find out who the nodes belong to?

I run a full node, but I'm taking it offline.  It has been a pain in the ass.  If I had the capacity to keep running it I would, but having to re download the freaking chain every time MS throws out a patch Tuesday fix my system reboots.  And then it starts all over again.  And again.  And again.  So, I'll use online storage services distributed throughout to deter a sweeping loss of everything. 

I might consider keeping it if there was an incentive.  Doesn't have to be dollars.  Shoot, give discounts on cloud hashing, exchange purchases, etc.  That's one way to incentivize.  Okay, so I'd do it for a teeshirt that has the BTC emblem with a caption saying "In your face, Federal Reserve!".  Lol

Bitcoin is based on trustless verification.

To the greatest extent possible we do not believe our peers at all.  They make claims— and we verify them.  Because we check for ourselves, we cannot be deceived by peers lying to us about most of the properties of the system.  Most of the time the most such an attack could hope to do is isolate us from the true state of the network— a denial of service attack... but as soon as we find a single peer that tells us the honest state we'll recognize it and accept it.

The only element of Bitcoin which cannot be trustlessly verified is the ordering of transactions— one ordering of one set of valid transactions is just as good as any other ordering of any other set of valid transactions, so we can't distinguish the real one without the help of a consensus algorithm. So we use mining to produce the ordering, and here again your node modification attack doesn't help because to substantially change the mining ordering you need to apply more computing power than the rest of the network.

Really? MS patches require re-downloading the entire blockchain? I find this rather astonishing.

I run a full node - but only for ~ten minutes - several hours every other day or so. Only takes a couple minutes for QT to sync to the last few days of blockchain activity. It is in no way an onerous task.

If I leave QT running on my computer 24/7 I am running a full node correct? Or is there some kind of specialized equipment or software I would need to do it?

yes, enable the "node " function ... to avoid problem.
(add -disablewallet in the shortcut or ... in bitcoin.conf, add disablewallet=1)

each node could have a BTC address associated with it

I don't understand why MS patches should make a difference (it deletes you blockchain!?). I run a node with -disablewallet on my linux box and I never hear a peep out of it. The network usage doesn't seem much of an issue at all. It does take up 650MB of RAM, which might be an issue on a lower spec machine.

Good luck paying off all/majority of them. Likely will never happen.

It's funny...I actually used to hate seeing posts like this but once my friend explained to me how to trying to take advantage of the system with various methods of attack ultimately makes the overall system that much more resilient (provided it survives).  I think there is a lot to be said for that argument.

Even if the majority were paid off, it should be obvious to other nodes that the paid off nodes are reporting incorrect information and would not keep these nodes in the network (not accept TXs from them, not relay found blocks and TXs to them). The result would be the paid off nodes would only be connected to each other with no miners processing any blocks to them and the non-paid off nodes running business as usual (there would be less nodes)

I'm not sure it would be that obvious or that all vectors of attack like this have been considered. The current bitcoin core code does have some detection mechanisms for really obvious stuff (void Misbehaving(NodeId pnode, int howmuch)), but not sure it captures everything. I don't know if a someone controlling 70% of the non-mining bitcoin nodes has been fully considered. Can you point to any discussions on this?

What would stop me from taking your money then switching my mine back on? Now "all ur bitcoin R belonging to me!" Oh and some of your cash.

What did you get? A huge bill.  ;)

Say what?? How would you do that?

If I understand the idea correctly. Miners would be paid to run alternate software. Assuming a majority of the worlds miners participate, this mystery code is really a way to fork the chain. At that point the mystery software takes over the network.
There are a million problems with this idea. Here are a few.

1. I would NEVER run a closed source program involving bitcoin. That would be insane.
2. The attacker has absolutely zero ability to keep miners from taking his money then not running his software. If they did then difficulty would be lower and the attacker is now back to competing for the same block rewards. Except he has paid money to do so.
3. If all went perfectly somehow and the attacker took control of a majority of hashing power then forked the chain... He looses. Why. All he owns is another stupid alt coin that he paid millions of dollars to get. I assume I am like most here and would not even consider using this new fork. Therefore his "fork coins" are basically valueless. The rest of us would still be using the original bitcoin.

Actually the idea was to own the majority of 'non-mining' bitcoin nodes. What damage can be done without controlling huge amounts of hash power?

Maybe something more subtle than forking the chain.