Here's what I think happens with these popular, highly anticipated coins with lots of Mhash at multiple pools.
Each of the large pools start mining based off of the current block chain. Because the difficulty is so low, block generations are occurring nearly instantly. Meanwhile, the network, users and pools are ddosing each other trying to get connections and updates. Competing block chains don't have enough time to propagate globally before N more blocks are found locally in a pool. So, each local pool has a block chain that could potentially stay longer than the updates that it is seeing from elsewhere. Other smaller pools and users are getting updates from these competing pools and either sticking with the closest/longest or switching back and forth.
As the difficulty goes up and speed of block generation finally goes down, you eventually get a more globally consistent block chain. However, potentially the vast majority of users find that they didn't get on the longest eventual chain and suddenly their coins disappear.
I would say that the solution probably lies in either much better pool communication (even under ddos type conditions) and/or raising the starting difficulty and slowing block generation to a liveable level quickly. If you don't have a globally consistent block chain then these coin launches are going to keep looking bad and anger a lot of participants.
Would love to hear other's thoughts.
Each of the large pools start mining based off of the current block chain. Because the difficulty is so low, block generations are occurring nearly instantly. Meanwhile, the network, users and pools are ddosing each other trying to get connections and updates. Competing block chains don't have enough time to propagate globally before N more blocks are found locally in a pool. So, each local pool has a block chain that could potentially stay longer than the updates that it is seeing from elsewhere. Other smaller pools and users are getting updates from these competing pools and either sticking with the closest/longest or switching back and forth.
As the difficulty goes up and speed of block generation finally goes down, you eventually get a more globally consistent block chain. However, potentially the vast majority of users find that they didn't get on the longest eventual chain and suddenly their coins disappear.
I would say that the solution probably lies in either much better pool communication (even under ddos type conditions) and/or raising the starting difficulty and slowing block generation to a liveable level quickly. If you don't have a globally consistent block chain then these coin launches are going to keep looking bad and anger a lot of participants.
Would love to hear other's thoughts.
I think you're right
I would agree, but we were finding absolutely no blocks in that time. The pool did find stale shares which were reported on the frontend, and in between the 1000% rounds there were brief 2-3 blocks around 30-300%
So, the front end was updating like it should, matching with the shares I was submitting and the shares of the pool. No blocks though...
I wonder if some of the smaller pools were constantly playing catchup with new blocks found by larger pools and by the time their participants completed all work on a block, that block was already stale. It would be like a constant live-lock. Again, another degenerate symptom of too fast block generation and insufficient blockchain syncing time.
Maybe? We will likely never know the real cause. I do run --no-submit-stale on some of my rigs, it's possible they just never got sent to the pool.
But at the time icyhash was one of the few pools up, I'd wager that it was one of the largest. I didn't take any screenshots or anything, and I wasn't paying close attention, but I think the pool hashrate got over 5gh/s for a time.
After the pool owners actions, I wouldn't mine for him again. I did manage to get coins from him, but that's because I know the wallet doesn't need to sync to get a payout address (tip for next time! Keep the wallet.dat)
