CritterDog wrote:Minerman wrote:IS it because the restarttime?
It was not an issue before.
Can I do something on my end because what ever you have done does not help at all.
I think you have said 7 or 8 times that you have fixed the issues with bad profit, but I can not se any improvements at all.
No its far from fixed.. Yesterday my A2 was stuck on Anoncoin for 95 percent of the time and when it did find a block it came up as "Lost" That is a new one for me..Never have I seen that before.. The re-starts are all messed up. They are WAY to high. I was up all night trying diff settings to see if I could get my A2 unstuck!!
I finally got it going this morning.. What I had to do was lower the d= all the way down to d=512 and I lowerd my Mhz from the stock 1200 down to 1100 which dropped my hash power on the A2 but allowed it to free up and go after other coins....I had to set all my gridseeds down to 512 also.. Then it has to run for a hour to somewhat stabilize..
This all started after the changes that were made.. Dont know what to say.. I should not have to lower my hashrate from stock just so it will start changing coins.. I get the impression they dont have a clue. I have spent hours on this site trying to help in testing and so forth costing me money because I really liked this pool. But I am done here for a while. I make way more on litecoinpool. But see also they dont charge a fee over there. Here it is 4.99 percent so that does not help either.. Mining my 2 bucks worth of litecoin right now almost done then I am done.. Sorry guys had enough of this. I tried.. It is not are equipment that is at fault...
Chris will be releasing a new version of the mining server tonight after we made a discovery that had been significantly increasing the work restart frequency for some coins.
It turns out that previously, we computed whether or not to re-send work to miners when a merge-mined coin had a new block by using the primary coin's target block time. Instead, we should use the actual block time, which is usually much lower. That change will result in fewer work restarts for merge mining coins, which were never necessary because of the bad assumption about block times. If you are still interested in mining, Chris will point out when the issue is finished with testing and released later tonight.
As to the Anoncoin lost blocks, thanks for that report. Chris will fix the issue overnight. Anoncoin had a wallet.dat that was 92MB, because we had so many found blocks in it. We need to periodically create new wallets and delete old ones when the size of the wallet gets to be over 50MB. What happens is that the daemon server gets overloaded with CPU for some reason and takes too long to submit blocks to the network, resulting on a timeout. Once Chris finishes the change, we should expect Anoncoin to return to normal. There are also 20 other coins like this.
Hopefully, the merge mining block times issue was significant and we'll see significantly improved performance tomorrow morning!
