Topic: [ANN][LTC][Pool][PPLNS][STRATUM] - ltc.kattare.com - burnside's Mining Pool - page 21. (Read 118848 times)

Thank you for that, it got me thinking about the block boundary issues we've seen in the past and I realized that the litecoind process that is supposed to send a HUP to pushpool might only be reading the PID file at startup.

I restarted litecoind, which caused it to re-read the PID file, and the stales have dropped significantly now.  I thought it would just pick up changes to the PID file as they occur.  So now I know that anytime pushpoold is restarted... I have to also restart litecoind.  At least until I can alter the custom patch to re-read it on each block boundary.

We still have other issues.  Now that I've found this one I need to set the difficulty back where it was.  I'd changed the difficulty thinking it'd help matters, but I think in reality it made them a percent or two worse than they were before.

The other thing I can do is I can change the values around for how long work is considered current.  Right now you only get two minutes from the time you get the work to when you submit it.  If I make that 3-5 minutes then some of the cpu miners will probably see lower stales.

Cheers.

I don't know if this helps but Coinotron uses stratum on his LTC pool.  Ask them maybe?

Hello Burnside

Hopefully this information is helpful.

Something I've noticed is that most of the rejected shares I receive on your pool are not stales as everyone seems to be claiming. They may be listed as round stales under My Stats; but they are not. I have 4 rigs and I run both cgminer and bfgminer. When running on your pool all four rigs are seeing an increased unknown work rejection. I have tried running theses rigs on several different pools and never see this error. On your pool my unknown work rejects have hit as high as 25 percent in a round.

I don't claim to be knowlegable of any of these systems and software; but, I am a capable troubleshooter in my day job. As a trouble shooter I try to break problems down into sections to solve them. The fact that my rigs work else where eliminates one of the sections and puts the problem at your end.

Dividing your end into two sections to start:

1- Your server is sending my rigs work that is not legitamate work. My miner solves and sends the work back and it is rejected as unknown work as it should be. It shows up as a stale under My Stats and is rejected.

or

2- The work your server sends is legitimate work. My end solves correctly and sends in the work. Your server rejects this legitimate work. Maybe its the DDos filters. Maybe some other error. It shows up as a stale under My Stats and is rejected.

I hope the above information somehow points you in the right direction. If it was legitimate stales it wouldn't be much of an issue. I'm running on a  Stratum pool right now and get some stales there too. Spending time working on Stratum when the problem may lay elsewhere will gain very little for the time being. I'm looking forward to this problem being solved. I like the excellent work you've put into your pool.

Is not what you think it is.

It's for legacy mining software to be able to connect to a stratum pool.   

https://github.com/CryptoManiac/stratum-mining-proxy  ??

I'd love to get stratum going.  I was playing with the stratum-mining-proxy, got it all installed and running, did the ./mining-proxy.py --help and realized that it's for connecting getwork (old skool) miners to a stratum pool.  Cripes.  I need an scrypt version of stratum-mining. 

Anyone aware of any LTC mining pool software that can do stratum?

Yeah, I switched about 3 months ago.  I know burnside works really hard to keep the place running, but he needs to implement stratum and vardiff at this point because he can't compete with the larger pools.   It will be more beneficial to him in the end to use stratum anyhow, more miners, more blocks, more fees for him (if he had vardiff and stratum at 1.5% fees PPLNS I'd be back in an instant).

Just did the same.  So far I'm getting maybe 3-4% stales on my well-connected workers, and 6-10% on my bad-internet ones.

Sure beats 40%.

I'd come back if the situation changes.

Come to ltcmine.ru and get 0.27% stales, that's what I'm getting right now.  

are you using a batch file or just manually entering info every time you click on cgminer.exe?

I'm sorry but I had to move to litecoinpool.org until you can figure out what is wrong. I don't like switching teams in the middle of the game but the 40% average stale rate is killing me. I get <1% there.

I'm around for the rest of the afternoon if you need anyone to test stuff. just got off work. Increased my hash to 1.15-1.5 depending on the activity on the miners. I'm excited about that but I feel like 30-45% stales is not good. Is everyone getting as many stales?

Someone up above posted that if we are all staling out we are not losing rewards. That would be true if we werent finding less blocks. The difficulty is slated to hit 65+ in a couple days... that is going to be nuts.

cgminer --scrypt

Yeah, the change didn't help things at all.  Looking into other options.

Cheers.

Pool efficiency has dropped to about 60%.

How can I change it? I'm using CGMiner's latest build, and there is a scrypt kernel in the folder.

You installed a 6950 in a laptop, or are we talking about two separate machines here?

Either way, those new stales are normal, if your miner was working on old blockchain data, and thus every share submitted was stale (noticed how it stopped when a new block was discovered on the network). You just didn't give it enough time, or you're not using scrypt. Thats 100% stales, but only over a few minutes - next time let litecoin wallet update your blockchain prior to mining. Please make sure youre using --scrypt (as it appears to be btc hashes youre solving) Also, most of all your conf settings are way off. Please review https://bitcointalksearch.org/topic/poor-radeon-hd6950-litecoin-mining-performance-need-help-solved-131083 or other such topics. Your concurrency, lookup gap,etc are all invalid values.

seems like you are running it with sha256 instead of scrypt ? That could'nt work with LTC

Also, my buddy started mining for me with a 6870. I have 3 configurations of cgminer set up for him. Lag Miner for when he is gaming or blueraying, Active Miner for surfing and netflix, etc, and Idle Miner for when he is sleeping/working. Active Miner works at 11 aggression, lag at 7 and idle at 18.

Idle and Active are overheating his card past 100C so he isn't able to keep mining for me. what am I missing?

I used the standard values in that thread, and I'd get a hard boot every time. So, I start adjusting variables on my own accord until I was able to get a stable set of variables. That's why my intensity, thread_concurrency are set so low.