[4+ EH] Slush Pool (slushpool.com); Overt AsicBoost; World First Mining Pool

[kano]

Well, the correct action would be to deduct the rewards paid to a miner who was withholding, and distribute them to the miners who were not paid what they should have been ...

It doesn't matter if it was accidental or not, the miner caused it and received payments that they should not have, and should have gone to the other miners.

[krisgt30]

Come on, does anyone really believe it was a "random" miner?

[bittalc1]

That message is just full of shit. Yea right miner was makeing problems.

[Prelude]

This reeks of bullshit.

[biggbox]

he posted this

Dear miners,
we would like to inform you that we have detected and resolved an unintentional block withholding attack.

The issue has been discovered a week ago and we have immediately taken an action by contacting the particular miner. The cause turned out to be a bug in a custom mining firmware, which has been promptly fixed by the miner. We have no indication that there was any bad intention. The fixed firmware solved two blocks since then so we can consider this issue as resolved.

A recent time period of worse luck can be at least partly attributed to this incident.

Block withholding attack is a well-known weakness of the whole pool mining principle and no public pool is immune against it.

As a by-product of the bad luck investigation, we have implemented a new method how to mathematically prove that the pool does not cheat on miners. This feature will be released after the public interface has been tweaked. The release is planned towards the end of next week.

The correct thing to do is to simply inform the public:

(1) What is the make and model of the miner?
(2) What is the version of the custom firmware?
(3) How to detect such "rogue" miners?

Why did the pool operator not release such public service announcement?

[bitsolutions]

Since nobody's posted it yet this is the announcement page of the block-withholding.

https://mining.bitcoin.cz/news/recent-low-luck-information/

[kano]

Well, the correct action would be to deduct the rewards paid to a miner who was withholding, and distribute them to the miners who were not paid what they should have been ...

It doesn't matter if it was accidental or not, the miner caused it and received payments that they should not have, and should have gone to the other miners.

The miners here should also be aware that for a pool to detect, with marginal certainly, that a miner is withholding, the miner would have had to mine more than the equivalent of 10 blocks and failed to find any blocks.

i.e. the miner would have been paid over 250BTC, while withholding, probably a lot more, that needs to be returned to the pool and distributed to the other miners.

I suggest you bring this up with the pool operator, since he was underpaying everyone by giving BTC to someone who was block withholding and he should be taking it back from them and returning it to the rest of the miners.

Also he should provide the bitcoin community with the details of what the miner was and the modifications made, to help stop it happening anywhere else ... unless he doesn't care?

--

I've sent support@bitcoin.cz an email saying they need to deal with this properly and a link to this post.

[d57heinz]

Well, the correct action would be to deduct the rewards paid to a miner who was withholding, and distribute them to the miners who were not paid what they should have been ...

It doesn't matter if it was accidental or not, the miner caused it and received payments that they should not have, and should have gone to the other miners.

The miners here should also be aware that for a pool to detect, with marginal certainly, that a miner is withholding, the miner would have had to mine more than the equivalent of 10 blocks and failed to find any blocks.

i.e. the miner would have been paid over 250BTC, while withholding, probably a lot more, that needs to be returned to the pool and distributed to the other miners.

I suggest you bring this up with the pool operator, since he was underpaying everyone by giving BTC to someone who was block withholding and he should be taking it back from them and returning it to the rest of the miners.

Also he should provide the bitcoin community with the details of what the miner was and the modifications made, to help stop it happening anywhere else ... unless he doesn't care?

--

I've sent support@bitcoin.cz an email saying they need to deal with this properly and a link to this post.

AS i have aswell.. on their facebook page since that is the only place they check for support related shit.. I know crazy right!?? anyway we will see whats up since we were on slush about this since back at least as far as november 2015 from what ive seen posted on fb.  Looking forward to see how this is handled!

Best Regards
d57heinz

[welshy82]

Well, the correct action would be to deduct the rewards paid to a miner who was withholding, and distribute them to the miners who were not paid what they should have been ...

It doesn't matter if it was accidental or not, the miner caused it and received payments that they should not have, and should have gone to the other miners.

The miners here should also be aware that for a pool to detect, with marginal certainly, that a miner is withholding, the miner would have had to mine more than the equivalent of 10 blocks and failed to find any blocks.

i.e. the miner would have been paid over 250BTC, while withholding, probably a lot more, that needs to be returned to the pool and distributed to the other miners.

I suggest you bring this up with the pool operator, since he was underpaying everyone by giving BTC to someone who was block withholding and he should be taking it back from them and returning it to the rest of the miners.

Also he should provide the bitcoin community with the details of what the miner was and the modifications made, to help stop it happening anywhere else ... unless he doesn't care?

--

I've sent support@bitcoin.cz an email saying they need to deal with this properly and a link to this post.

lets hope he does pay us some sort of compo lol or hes gonna lose a lot oif unhappy miners

[btiAndy]

Good to see Slush is trying to come clean, but of course the damage is done and I think it will take much for him to rebuild trust; agree there needs to be more disclosure on who this miner was and what/how the block withholding attack was accomplished.

[ohmygod21]

just wasted 4 Ph with incredible bad luck , 2 blocks > 23 h

3.80 btc loss =) hihaaaaa

[hoosier_13]

Slush is not coming clean.  Far from it.  This issue was known about for weeks and the answer was always bad luck.  They knew there was an issue and did not do the right thing and inform their loyal miners that they could be subject to lose a lot of revenue.

I lost many BTC from this debacle and they have refused any compensation.

[d57heinz]

Slush is not coming clean.  Far from it.  This issue was known about for weeks and the answer was always bad luck.  They knew there was an issue and did not do the right thing and inform their loyal miners that they could be subject to lose a lot of revenue.                                                                                                             MONTHS

I lost many BTC from this debacle and they have refused any compensation.

Best Regards
d57heinz

[krisgt30]

I mean I have only been mining since the end of December, and right away I knew something was up. I thought to myself, how can 40ph find only one block a day when at the time you had 4ph pools average 1.2 blocks a day. After about 3 weeks on slush pool I noticed this and pulled out and haven't looked back. I am frankly amazed people are still mining on this pool.

[btiAndy]

Yeah in hindsight now and thinking about this more clearly something really stinks about this explanation from Slush. He is really not coming clean at all here.

For months we have been watching this cycle and have made repeated attempts to communicate this pattern to him for explanation. Nothing. Now he is basically saying that this particular miner was "accidentally" withholding blocks and that the problem has been fixed...and in fact he knows it has been fixed as that same miner has since the repair cracked two blocks... just not right and not near enough.

Slush should have banned that miner immediately (there must be a way to do this) and made an attempt to recover lost funds to return to those of us who stuck with him for so long. I have been gone from Slush for a few weeks now (actually stopped mining all together as there is no suitable outlet for a small scale miner like me anymore), Slush has definitely soured me on BTC mining as a result though.

Really rotten.

[notabeliever]

not happy

[jonnybravo0311]

Yeah in hindsight now and thinking about this more clearly something really stinks about this explanation from Slush. He is really not coming clean at all here.

For months we have been watching this cycle and have made repeated attempts to communicate this pattern to him for explanation. Nothing. Now he is basically saying that this particular miner was "accidentally" withholding blocks and that the problem has been fixed...and in fact he knows it has been fixed as that same miner has since the repair cracked two blocks... just not right and not near enough.

Slush should have banned that miner immediately (there must be a way to do this) and made an attempt to recover lost funds to return to those of us who stuck with him for so long. I have been gone from Slush for a few weeks now (actually stopped mining all together as there is no suitable outlet for a small scale miner like me anymore), Slush has definitely soured me on BTC mining as a result though.

Really rotten.

As was explained earlier, you cannot simply ban somebody immediately.  You don't really know someone is performing a block withholding attack until a large number of hashes have been produced without a block solution.  Plenty of pools regularly suffer through long rounds - it is the nature of mining.  It goes to follow, then, that an individual miner also suffers the same.  Kano suggested 10 blocks earlier.  At that point, the chances of it being "luck" are exceptionally minimal.  Unfortunately, 250BTC have been lost to the miner.

[btiAndy]

Yeah in hindsight now and thinking about this more clearly something really stinks about this explanation from Slush. He is really not coming clean at all here.

For months we have been watching this cycle and have made repeated attempts to communicate this pattern to him for explanation. Nothing. Now he is basically saying that this particular miner was "accidentally" withholding blocks and that the problem has been fixed...and in fact he knows it has been fixed as that same miner has since the repair cracked two blocks... just not right and not near enough.

Slush should have banned that miner immediately (there must be a way to do this) and made an attempt to recover lost funds to return to those of us who stuck with him for so long. I have been gone from Slush for a few weeks now (actually stopped mining all together as there is no suitable outlet for a small scale miner like me anymore), Slush has definitely soured me on BTC mining as a result though.

Really rotten.

As was explained earlier, you cannot simply ban somebody immediately.  You don't really know someone is performing a block withholding attack until a large number of hashes have been produced without a block solution.  Plenty of pools regularly suffer through long rounds - it is the nature of mining.  It goes to follow, then, that an individual miner also suffers the same.  Kano suggested 10 blocks earlier.  At that point, the chances of it being "luck" are exceptionally minimal.  Unfortunately, 250BTC have been lost to the miner.

OK Fair enough, thank you for the explanation. I am not sure on the time frame though required for analysis with regard to Kano's explanation.I am sure I am sounding too simplistic and like a "Noob" but my question still stands, ..how long should it take to expose a withholding attack under these circumstances? At Slush this has been going on for at least 3 months in my practical opinion...we saw it repeatedly that when Slush dipped down to ca. 33 PH  the blocks started hitting, the minute we returned to 40PH+ all of a sudden performance went south. In simple terms, it seemed that when one big miner left the pool things would return to "normal", when he returned it went to hell.   just from a standpoint of analyzing a pattern of events should Slush not have been able to pick up on this pattern and figured out where the problem was?

Still curious...

[btiAndy]

And one last thing, I must say I really enjoy reading your posts Jonny and I have learned a lot over the past few months from your posts...I just want to clarify my statement in my previous post regarding banning the offending miner; I did not mean ban the miner in advance...I think the bad part here is that Slush has stated he contacted this miner after discovering the issue and the solution was simply to "fix" the "unintentional" problem and let this miner continue mining without penalty...no matter if the act was intentional or non-intentional I think those of us who were affected by this problem should have been respected a bit more.

Of course I am speaking from a "blue-sky" customer service standpoint and nothing more, I am sure that legally Slush owes his miners no real duty of service and is not obligated  in any way to compensate miners for wasted time and hash on his pool...it is gambling in the end and we should all know that. I am just saying that for the common good this offending miner who has been supposedly identified by Slush as the cause of the withholding attack should have been removed from the mining pool as consideration to the other miners still there.

[jonnybravo0311]

To address the concerns in your first reply... it's a tough bit of analysis.  In an ideal world, if the pool at 40PH was experiencing the same luck as the pool at 33PH, then it would be apparent that 7PH was withholding blocks.  If all 7PH belonged to a single miner, you found your culprit.  Unfortunately, the problem remains: you must wait some period of time before you are relatively certain that the miner isn't just unlucky.  I'm not sure how many, if any, pool operators have written this kind of analytics into their code.  It's a pretty simple bit of analysis: count shares submitted by miner vs number of block solutions.  If number of shares is some percentage over expectations, flag as potential block withholding.

That doesn't solve it, but merely alerts the pool operator something might be amiss.  Now the pool operator has to take action.  What action the operator takes is really the basis of your second reply.  Personally, I don't agree with Slush's decision to allow the miner to continue on his pool without any kind of repercussions.  "Sorry, my bad!" just doesn't cut it.  Intentional or not, if a miner is going to write his own custom firmware, that miner should be testing it on a simulated environment, not forcing the public to assume the risks of untested code.  I'm glad for the miners on the pool that the problem was found and addressed; however, if I were a miner on the pool, I'd be seriously considering going elsewhere if neither the offending miner, nor the pool operator were held accountable in some fashion.