Pools down & total computation speed up - what does this mean?

[applecart]

Today, the total computation speed of the network moved from a low of 3000 GH/s to a high of 5000 GH/s, while the percentage of non-pool mining increased dramatically, involving deepbit and slush both going down.

Clearly something doesn't add up. Why does the computation speed go up when large pools are out of action?

There are also claims on the forum that two large FPGA clusters were brought online today. And other claims that the pools going down were due to DDOS attacks.

Draw your own conclusions.

http://bitcoin.sipa.be/speed-small-lin-2k.png
http://chart.googleapis.com/chart?chs=350x200&chd=t:16.88,8.06,61.23,2.63,10.33,0.88&cht=p&chf=bg,s,00000000&chl=slush|btcguild|other|Eligius|btcmine|swepool

[1715015141]

1715015141

[1715015141]

1715015141

[1715015141]

1715015141

[1715015141]

1715015141

[Anonymous]

Maybe the bitcoin network became self aware and decided the pools were a threat so it ddosed them ?

Seriously there is a theoretical attack scenario where the large pools who form a majority of the network can be ddosed and then the attacker has less surviving mining power to overcome. Perhaps someone is trying that. 

[applecart]

If not an outright attack, then just getting the competition out of the way to make mining more profitable?

[V2-V3]

sounds like a conspiracy.....

[martok]

Seriously there is a theoretical attack scenario where the large pools who form a majority of the network can be ddosed and then the attacker has less surviving mining power to overcome. Perhaps someone is trying that. 

How so? If you and I are mining and you stop, that doesn't improve my chances of finding a block. Not until your stopping lowers the difficulty. Until then, I am still attempting the same difficulty hashing and have the same probability of success per attempt.

[lizthegrey]

Seriously there is a theoretical attack scenario where the large pools who form a majority of the network can be ddosed and then the attacker has less surviving mining power to overcome. Perhaps someone is trying that. 

How so? If you and I are mining and you stop, that doesn't improve my chances of finding a block. Not until your stopping lowers the difficulty. Until then, I am still attempting the same difficulty hashing and have the same probability of success per attempt.

Because the amount of effort required to fork the block chain depends upon the relative amount of honest and corrupted hashing power, so less honest hashing power and more corrupted hashing power, even at high difficulty, results in the possibility of a forking attack.

[Horkabork]

If someone has enough power to DDOS a pool, then they're not in it for any plausible reasons other that pure mischief or to move the market and profit from that.

If your goal is to make your own mining more profitable, then it would be much more efficient to DDOS the exchanges. Mining would tank if people couldn't make trades from and to other currencies.

[gerex]

If your goal is to make your own mining more profitable, then it would be much more efficient to DDOS the exchanges. Mining would tank if people couldn't make trades from and to other currencies.

No, the attacker needs the exchanges to convert his own bitcoin into other currencies. But following the law of supply and demand, if there are less producers (mining pools), the value of all existing bitcoins will rise. So take out a pool and your bitcoins automatically increases in value.

[goatpig]

Seriously there is a theoretical attack scenario where the large pools who form a majority of the network can be ddosed and then the attacker has less surviving mining power to overcome. Perhaps someone is trying that. 

How so? If you and I are mining and you stop, that doesn't improve my chances of finding a block. Not until your stopping lowers the difficulty. Until then, I am still attempting the same difficulty hashing and have the same probability of success per attempt.

If the network is producing a block every 10 minutes, I need to be able to inject my 2-3 corrupt blocks within that time frame, which means I need astronomical hashing power, like 90% of the network. If the pools are taken out and we are in the situation were one block is solved every 30 minutes for the next hour or so, like we just experienced, during that window, I need much much less hashing power to fork the chain.

[goatpig]

If your goal is to make your own mining more profitable, then it would be much more efficient to DDOS the exchanges. Mining would tank if people couldn't make trades from and to other currencies.

This has the opposite effect. The miners won't stop mining, and when the echange gets back online, they'll flood the market with their days of unsold BTC, crashing the price. That's what happened when BTC hit over $4 and MtGox got DDoS'd. Trading was down for a week, and when Gox was finally stable, BTC was at $2.5

[kjj]

There is no way to measure hashing power globally.  The numbers you see on websites and in those charts are just estimates.

[Vladimir]

kjj, Network hashing power can be estimated very accurately using statistical methods by measuring how many blocks are solved during a period of time. It is just "websites" you've mentioned are doing rather poor job.

applecart, Anonymous must not be very fond of anyone approaching 50% of bitcoin hashing threshold. I would say they are too relaxed and 25-30% is not that good either.

[yunk3r]

if some one had a large botnet couldn't they also while ddos a pool also use there botnet to mine. even if all of the computers part of the botnet were cpu mining it could still be very powerful. 

[kjj]

kjj, Network hashing power can be estimated very accurately using statistical methods by measuring how many blocks are solved during a period of time. It is just "websites" you've mentioned are doing rather poor job.

applecart, Anonymous must not be very fond of anyone approaching 50% of bitcoin hashing threshold. I would say they are too relaxed and 25-30% is not that good either.

Over a long period of time, the estimates can be pretty accurate, yes.  Over short times, not even close.

[Vladimir]

kjj, fair enough.

sipa's charts do give you/us longer period and much more accurate estimation. Why would anyone concentrate so much on that green, short term line which jumps all around the place and why people do not treat it as noise is beyond me.

[Maxxx]

if some one had a large botnet couldn't they also while ddos a pool also use there botnet to mine. even if all of the computers part of the botnet were cpu mining it could still be very powerful. 

Yes, most definitely.

[kjj]

kjj, fair enough.

sipa's charts do give you/us longer period and much more accurate estimation. Why would anyone concentrate so much on that green, short term line which jumps all around the place and why people do not treat it as noise is beyond me.

Because they are under the mistaken belief that the green line represents some actual measured thing.