Surely this isn't a 51% attack on bitcoins, right?

[BurtW]

I am connected to deepbit and the miners are all still running.  However the web site does appear to be down at this time.  There does seem to be an intermittent communication problem between my miners and deepbit.  They may be under attack (again).

If the two largest pools were really down the overall hash rate would have dropped a lot.  So I expect the pools are running but they are unable to report back their results (back to Bitcoin Watch) due to some other web site issue/attack.  Since Bitcoin Watch cant associate the hashing with any pool it lumps the remainder of the current hash rate into "other".

So I belive the "other" just represents the normal hashing activity of the pools.

[1714961564]

1714961564

[1714961564]

1714961564

[1714961564]

1714961564

[1714961564]

1714961564

[grod]

This is very simple.  Bitcoinwatch averages blocks over the last 24 hours.   If you have everyone producing 0 blocks (this isn't the case) and reporting that the "Other" pie chart is 100% based on past block rates alone.

This is used to happen during difficulty jumps up.  For the 24 hours following the jump in difficulty that chart would compute how much hashing power would be required to get the average # of blocks over 24 hours at current difficulty, compare that to reporting hashing power of the pools and assign the rest to "other."  Since the average was hashed at a lower difficulty you'd get these giant "Other" slices showing up every retarget.

[jjiimm_64]

The overall security of the network is not really under threat, since 10 Terahashes per second isn't a mark reachable by any known network or supercomputer.  This does highlight the risks to the bitcoin network at large associated with pool mining.  Concentrating the mining to a single server for the purpose of income stability also creates central points of failure/attack.  I don't mine in pools for this exact reason, and don't think anyone with any substantial hashing ability should do so.  10K small lights are harder to shoot out than a dozen really big ones.

This brings me to another point..  in a conversation with a colleague he said " What if IBM or some other entity with tons of computing power bring it online all at once?"  He mentioned something on the order of "3 trillion teraflops/sec" or something that one company has.  

I did some rough calculations of the TFlops of of all the vid cards encompassing 11Thash.  I think I came to about 100,000 on our network.

I know that teraflops does not equate directly to hashing power, but I would like to know if anyone knows of the biggest competition to our network?

[digital]

I think the K Computer in Japan does 8 petaflops, and im pretty sure that is the current fastest supercomputer.

But not sure about any farms out there...

[BurtW]

Converting the number of hashes to petaflops is a controversial issue.  The cacluation done at bitcoin watch is considered "high" by most.  However using it for the sake of argument the Bitcoin network would be about 137 petaflops.  If we take even 1/4 of that as the real number the Bitcoin network is pretty darn safe from an outside attack.

Here are some numbers for other distributed networks (from wikipedia).  Since Bitcoin only does integer operations it is hard to compare apples to apples:

BOINC – 5.634 PFLOPS as of April 4, 2011.[4]
Folding@Home – 5 PFLOPS, as of March 17, 2009[5]
As of April 2010[update], MilkyWay@Home computes at over 1.6 PFLOPS, with a large amount of this work coming from GPUs.[6]
As of April 2010[update], SETI@Home computes data averages more than 730 TFLOPS.[7]
As of April 2010[update], Einstein@Home is crunching more than 210 TFLOPS.[8]
As of June 2011[update], GIMPS is sustaining 61 TFLOPS.[9]

Also there is a thread here somewhere discussing all this and attempting to claim Bitcoin is the largest for the world record book.

[DeathAndTaxes]

DOSing the major pools isn't what I think of as a "51% attack", but whatever.  Whether or not the pools are under attack or not is of little concern to myself.  The overall security of the network is not really under threat, since 10 Terahashes per second isn't a mark reachable by any known network or supercomputer.  This does highlight the risks to the bitcoin network at large associated with pool mining.  Concentrating the mining to a single server for the purpose of income stability also creates central points of failure/attack.  I don't mine in pools for this exact reason, and don't think anyone with any substantial hashing ability should do so.  10K small lights are harder to shoot out than a dozen really big ones.

No but if you had say 4TH of computing power knocking (or trying to) the largest pools off the network could be a precursor.  The 3 largest pools have ~6TH combined.  The 5 largest pools have ~7TH combined.  Disrupting the network softens it up for a 51% attack.

Still it looks like operator of BTC Guild thinks this happened because he pissed off a botnet operator (by blocking him from his pool)

I don't think it is something you did.  Slush indicates his pool is under attack and deep bit users are indcating connectivity issues.

I'm referring to the fact that this all happened a few minutes after I posted that I had submitted a giant list of IPs that I had been collecting from the botnet miners the last few days.  Could very easily be the childish mentality of these "hackers":  "Oh yeah, well I'm going to take my ball and go home, but first I'm going to pop your ball."

[BurtW]

Well now that is interesting.  I have connected my miners to Slush, BTC Guild and Deepbit and they are all three down as of now, so I moved on to a smaller pool.  I am monitoring the "big three" and I cannot connect to any of them at this time!

[DeathAndTaxes]

Well now that is interesting.  I have connected my miners to Slush, BTC Guild and Deepbit and they are all three down as of now, so I moved on to a smaller pool.  I am monitoring the "big three" and I cannot connect to any of them at this time!

Weird I was kicked off slush but about 30 minutes ago I was able to reconnect and have been able to stay connected.  Looks like denial of service is partial?

[DeathAndTaxes]

Well now that is interesting.  I have connected my miners to Slush, BTC Guild and Deepbit and they are all three down as of now, so I moved on to a smaller pool.  I am monitoring the "big three" and I cannot connect to any of them at this time!

Weird I was kicked off slush but about 30 minutes ago I was able to reconnect and have been able to stay connected.  Looks like denial of service is partial?

Spoke to soon.  Had to know I was going to jink myself. I can't connect anymore.  Blerg.  Well I will drop 6GH to solo mining at least for rest of the day.

[BurtW]

O well I have decided to spread my hashing power between the big three plus my favorite small pool (BTC Mine).  I have been monitoring them and they are up and down and up and ...  But on the average I am hashing at one or more at all times.  Good enough for today.  They will eventually get tired of wasting their power on the DDoS and go away.

[bulanula]

Clearly just a preview from BitcoinExpress of what he has in store for us if the price falls to low and miners drop out and the network is weakened.

BUY BUY BUY !!!

[BurtW]

BCE has stated that he has a lot of Bitcoins so he has no plans to mess with the Bitcoin network.  But I think it is a good idea to introduce at least one new alt coin chain per week in order to keep him busy killing them off - sort of like the whack-a-mole game

BTW I am "doing my part" and turning on all of my idle miners and running them "just in case".  Don't know if my piddly amount will help but there you go.

[deslok]

It's possible this is not a "51% attack" but something much simpler, bringing down a pool as large as deepbit, slush or btc guild this close to a difficulty change will 1, decrease the number of new blocks found slowing the flow of new blocks into exchanges theoretically temporarily stablizing the prices(this would happen at any point a major pool went offline however) and 2 cause the difficulty drop to be longer than normal by affecting the blocks/hour ratio even if done for a day this would cause a greater than predicted difficulty drop and allow said attacker to mine more coins than prior durng the next cycle

[BurtW]

That is what I was thinking.  I believe we are running at about 5 blocks per hour and the retarget is about 1.5m and only about 100 blocks away so if this keeps up long enough the difficulty may drop below 1.5m.

[abracadabra]

This is exactly why concentrating too much hash power in a few pools is not a good idea.
Better to mine at a smaller mid-sized pool

[slush]

This is exactly why concentrating too much hash power in a few pools is not a good idea.
Better to mine at a smaller mid-sized pool

No, it's basically the same. With such massive botnet power there's no problem with flooding tens of smaller pools. Now attackers don't do that because 70% hashrate is on three pools, but they'll change attack vector if hashrate distribution will change.

[c_k]

https://rfcpool.com/ still working just fine 

[acostanza]

It's possible this is not a "51% attack" but something much simpler, bringing down a pool as large as deepbit, slush or btc guild this close to a difficulty change will 1, decrease the number of new blocks found slowing the flow of new blocks into exchanges theoretically temporarily stabilizing the prices(this would happen at any point a major pool went offline however) and 2 cause the difficulty drop to be longer than normal by affecting the blocks/hour ratio even if done for a day this would cause a greater than predicted difficulty drop and allow said attacker to mine more coins than prior during the next cycle

I like you conspiracy theories Deslok! Plus, less difficulty is hard to hate.
Question, is it better to point small (~2 Ghash) guns at a operating pool during times like this, or is it better to go solo? My line of thinking is it's better to be in a pool because you'll contribute to solving a block faster, and that's what effectively hinders any 51% bs. Going solo with anything less than 10 Ghash seems wasteful of precious block solving time.

BTW, I'm at BTC Guild for now.

[jjiimm_64]

I like you conspiracy theories Deslok! Plus, less difficulty is hard to hate.
Question, is it better to point small (~2 Ghash) guns at a operating pool during times like this, or is it better to go solo? My line of thinking is it's better to be in a pool because you'll contribute to solving a block faster, and that's what effectively hinders any 51% bs. Going solo with anything less than 1 Ghash seems wasteful of precious block solving time.

It is my understanding that Ghash's count towards blocks found either  solo or at a  pool equally well.

[phillipsjk]

Question, is it better to point small (~2 Ghash) guns at a operating pool during times like this, or is it better to go solo?

If the case of a widespread DDOS attack, mining solo is better so that you can relay transactions, taking load off of the pools. I am currently reading the forum instead of geting some CPU miners up. If the attack continues, I may have two (relatively power hungry) CPU miners at two separate IP addresses.

I have been planning on setting up a netbook CPU miner drawing 12Watts (for the near term), but have not set that up yet. The other CPU miners I have available probably draw 100W each. I am also concerned about bandwidth use and have not set up Bandwidth monitoring on my end yet.