War is upon us

Steve Sokolowski wrote:We concluded that the only possible cause was if the firmware on these miners was defective and hashing incorrectly or evaluating share difficulty incorrectly. Some Internet research suggested that some cloud mining providers like NiceHash were polluted with these misconfigured miners, but we were not able to correlate the miners with any specific cloud mining service or IP address.

Just for the record - NiceHash can NOT be polluted with a misconfigured miner. We have a very rigorous per-share validation implemented in our system. This means that each share, which is sent from a miner towards NiceHash and then forwarded to any target pool, is a regular and verified share. If share is invalid, it's rejected before it is forwarded to the target pool. More details on this topic is in this FAQ.

Best regards,
NiceHash team.

nicehash wrote:

Steve Sokolowski wrote:We concluded that the only possible cause was if the firmware on these miners was defective and hashing incorrectly or evaluating share difficulty incorrectly. Some Internet research suggested that some cloud mining providers like NiceHash were polluted with these misconfigured miners, but we were not able to correlate the miners with any specific cloud mining service or IP address.

Just for the record - NiceHash can NOT be polluted with a misconfigured miner. We have a very rigorous per-share validation implemented in our system. This means that each share, which is sent from a miner towards NiceHash and then forwarded to any target pool, is a regular and verified share. If share is invalid, it's rejected before it is forwarded to the target pool. More details on this topic is in this FAQ.

Best regards,
NiceHash team.

Thanks for the input. I think you might have misunderstood what the problem is. The shares that are submitted are valid and are getting through to the target pools as expected. The problem is that shares that find blocks are not submitted, leading to lower luck at the affected pool. Cloud mining providers, which do not earn revenue from finding blocks, can't prevent this problem because they don't even know what luck is supposed to be at the target pool.

However, you are correct that we were not able to find a correlation with any specific source of these shares, so there is no evidence that NiceHash is more likely to be a source for these issues than any other miner. The inaccurate suggestions that cloud mining services are solely responsible are found in Internet forums.

Based off my brief experience in playing around w/ bfgminer source code ... I can see how easy it would be to do a simple "IF share diff > network diff THEN dont submit"
And you're right, launching an army of miners w/ that type of code against a pool ... would sure drive away other users from that pool due to the long running bad luck caused by witholding blocks.

Such a sad state crypto is in right now =( ...hopefully it will weather the storm ...

tbf I have been reading this, a way to attack pools is simply pointing lots of hashpower via PPS and simply discard any solution over the current difficulty, also called a block withholding attack. That could explain the 'luck' where these miners were concerned. You said nicehash is polluted with these types, this has been suspect where rentals are concerned from there as it would encourage people to rent more as many rent hash from there in hopes to hit on CKs solo pool.