NETWORK FREEZE bitcoin difficulty stuck high with a large hashrate drop

[stslimited]

If a large portion of the network went offline right after a difficulty increase, this scenario could freeze the bitcoin network for thousands of blocks. Basically blocks would be found MUCH more slowly to the point of it being nearly unusable.

This exact scenario happened with terracoins a few month ago. Basically what happened was that someone pointed an ASIC at terracoins and this push the difficulty EXTREMELY high, and then they stopped mining, leaving the difficulty extremely high for 8 days and no terracoin transactions were able to be processed because it took an extremely long time for the remaining miners to process any blocks.


If two of the largest miners of bitcoin went offline, the entire bitcoin economy would be at a standstill for at least several days. Although with today's makeup, just a pool could go offline and all the pool members would continue mining at a different pool. But in a different scenario, ASICMiner could have a larger portion of the network and lose connectivity right after a difficulty increase, or 6 months in the future, a different company would have 100s of terahashes more and the same thing could happen.

This is less likely with bitcoin today, but it is a real vulnerability.

In terracoin's situation the remaining hashrate was only 1/10th of what it was after the ASIC bumped the difficulty up. This isn't currently a possible scenario with bitcoin. But one could imagine a party with some future 14nm sha-256 chips developed solely to prop up the difficulty and then leave the network, making it take months until the next difficulty adjustment, making blocks take hours or days to solve, rendering transactions useless.

What can we do about that? Terracoin has fast changing difficulty

[1538089392]

1538089392

[1538089392]

1538089392

[techwtf]

As long as mining generate profit, that thing wont happen at all, miners have the intensive to configure backup pools & do other preparations in order to continue mining if the pool is down / other bad things happen.

[mustyoshi]

Why would somebody spend the money to develop, then deploy the miners just to pull them offline? They'd make more money by continuing to mine...

Even if they were to just do this out of hatred, the network would adapt rather quickly.

[gmaxwell]

If two of the largest miners of bitcoin went offline, the entire bitcoin economy would be at a standstill for at least several days.

Blocks taking 2 - 3x longer than nominal is hardly "at a standstill". If there is so much hashpower consolidation that this would be a severe concern then our security model would have already failed— it's not like you need a majority hashpower to make moderate reorganizations and transaction reversals with some success.

This is less likely with bitcoin today, but it is a real vulnerability.

I don't think it's clear that it is. Being just a small multiple slower would be annoying, but if left us short capacity competition for space would attract larger transaction fees, which would make mining more profitable, which would draw in more hashrate.

What can we do about that? Terracoin has fast changing difficulty

And ended up with exploitable vulnerabilities multiple times as a result of their tinkering there.  Continuous difficulty adjustment halves the cost for an attacker to mine down a fork for use in isolation attacks.

Presumably if the network were ever stranded— and people somehow still cared about Bitcoin at all (it not clear to me how those two things could ever be true)— then it wouldn't be too difficult to do a single point hardfork to step the difficulty back down. Considering that, I think this is not worth worrying about.

[TradeFortress]

A block taking 30 minutes is still acceptable. We get a hour or more without blocks frequently too. A block taking 6 hours would be a bigger problem, however it still would not represent the death of Bitcoin and it's implausible that we'd get 90% hashpower loss.

[BombaUcigasa]

This exact scenario happened with terracoins a few month ago.

You wanted to say "this exact scenario happened to bitcoins a few years ago"...



And yes, you are correct, it will be bad for a while... longer than needed, bitcoin does not have good recovery mechanisms like other altcoins where retargeting can occur through other means than simple block count. An increase of difficulty can happen in less than 50% the regular block time while a decrease might need 200% of the regular block time.

[Evolyn]

bitcoin does not have good recovery mechanisms

If transactions need longer to confirm due to lack of hashpower, people spend more fee to get their transactions confirmed while there's a blocksize limit or there are just more transactions (more fees) to grab up per block. Especially when basic block reward gets lower and fees become more important. Mining becomes more profitable and more miners appear. I think we have the best recovery mechanism could exist - greed. Thats my humble opinion.

[Carlton Banks]

This exact scenario happened with terracoins a few month ago.

You wanted to say "this exact scenario happened to bitcoins a few years ago"...

Sorry to be a smartass, but the February '09 - January '10 period had difficulty of "1". I don't think the difficulty uses decimal fractions, so this was actually just CPU miners struggling to get the hashrate to a point where the difficulty increased to >= 2.

[solex]

Yes. That example with difficulty <1 is a effectively a strawman example which undermines, rather than supports, the OP's concerns.

Considering the OP, surely, anyone with 50% of the hashing-power and prepared to switch it off suddenly would only create a period of 20 minute blocks (up to 2 weeks worth). Hardly a disaster. However, it begs the question of why someone with 50% could not stretch a bit further and perform all the disruptive 51% actions which have been discussed ad nauseum in this forum. Fortunately the hashing power is well distributed. If a major pool stopped, then individuals would quickly choose another pool and the block solving would continue close to the required rate.

[kjj]

This exact scenario happened with terracoins a few month ago. Basically what happened was that someone pointed an ASIC at terracoins and this push the difficulty EXTREMELY high, and then they stopped mining, leaving the difficulty extremely high for 8 days and no terracoin transactions were able to be processed because it took an extremely long time for the remaining miners to process any blocks.

This has been happening to scamcoins (or "alt"coins, if you prefer) for years.  There are plenty of threads from 2011 and 2012 discussing this.  I'll summarize them for you, since you don't seem to know how to search for them.

1.  Most of us think that a sudden major drop in hash rate would be a symptom of the effective end of bitcoin anyway, so there isn't much point worrying about it.

2.  Mechanisms to quickly reduce difficulty require an asymmetric difficulty adjustment system.

3.  Asymmetry in that system is a huge security problem.  It lets a small miner create fake chains to fool nodes that are booting for the first time, and it lets a (relatively) large miner dominate the entire network and manipulate it to his liking.

4.  #3 is not theoretical.  Artforz pwn3d one of the original scamcoins right after they patched to "solve" the problem described in the part of your post that I quoted.

[TierNolan]

If transactions need longer to confirm due to lack of hashpower, people spend more fee to get their transactions confirmed while there's a blocksize limit or there are just more transactions (more fees) to grab up per block.

This could work if minting was a small part of the reward.

Inherently, as time passes, the reward would increase.  However, atm, transaction fees are not significant.

[razorfishsl]

As long as mining generate profit, that thing wont happen at all, miners have the intensive to configure backup pools & do other preparations in order to continue mining if the pool is down / other bad things happen.

Fail.......

You are not taking into account the possibility of natural disasters, Or deliberate sabotage.

1. Several times over the last decade, Major communication cables have been cut, between Asia and the Backbone.
2. A bunch of clowns in a row boat attempted to cut a major communication cable in the Gulf recently.
3. A 'certain' big country decides it is going to block all mining traffic...

There is way more to maintaining a distributed network, than a few 'configurations' to roll a pool over.....

[gmaxwell]

In the case of a network partition we actually _want_ blocks to stop or go very slowly. Otherwise a large amount of partitioned hashpower we're not currently aware of could show back up and wipe out the recent history, reversing recent transactions on the losing side and leaving everyone who transacted in the span vulnerable to reversals.

This is basically the same risk as an attacker who mines a chain down to trick isolated nodes, but happening naturally without an attacker.