Topic: [ANN][AUTO-SWITCH] Profit-switch auto-exchange pool: CleverMining.com - page 268. (Read 554401 times)

hmmm... looks like rejects %% are really going down for me, I'm increasing my hashrate here to 5mh/s for some more tests.

btw, congrats to dev on reaching 3gh/s and hope they prepared for some more!  

This is a valid point that was discounted by noisy posters.

It doesn't affect everyone the same way. For example, if one is far away from the pool (e.g. a miner in Eastern Europe connecting to U.S. West), a pool becomes often unusable.

Of course, end profit is what matters and there's always a balance. Some suggestions are zero-sum, or include extra hassle and cost for pool operators . However, if it's good for you, leave it at that and don't speak for other people.

@Terk: I believe when pool hashrate goes very high, it cant point all of its hashrate to some small coins because it would produce reject shares or produce very high amount of coins that selling them is harder in market.

What do u think?

@terk:  Very good you tweak to reduce reject shares.

Actually i have only 1-2% reject to compare with 8-10% this morning.

very good job thanks a lot.

I've been seeing the lower reject rate in my stats for a couple of hours now. It is still a few percent but not double digits. Assuming the pools hourly stats maintain a good 24 hour performance rate, I'm happier with the lower reject rate.

this pool is getting big!  nearly 3k MH/s

While waffle had zero rejects for me. I'm getting an average of 3-4% here. Maybe try different settings guys? It's odd we have such a big gap among everyone on what their reject rate is.

If you haven't been mining for the full clock hour then last hour average will show lower value. It basically sums all shares submitted by you within this hour and divides by 60 minutes. Please take a look at real-time chart or wait for the full clock hour (and then 10-20 mins to stats regenerate and website cache refresh).


Yes, of course miners are notified with new work about every new block and on every coin changes. There wouldn't be much hashing if they weren't receiving new block notifications.

Yes.

Well I am happy with 100% rejects as well as long as I am getting paid well. I am here since 18hours for testing the pool for a couple of days, and we will see.

Btw a little explanation on rejects : Your share is only getting rejected when you send in a share which is not actual. Actually the pool needs to notify the miner that we have a new block/coin mining and stop mining the old one. When your miner is not getting notified then happens the reject. I am sure you will be able to improve the reject ratio, but probably you will need to recode some parts of the stratum-mining, and open and EU server. You need to force the stratum-mining to notify the miner every time there is new block or new coin being mined, maybe you can make make some debugs with sniffing the packets when the stratum fails to do so.

Because it's an average.

@Terk: You should put the most important information regarding rejects in the 1st post, or even in the FAQ on the website. The topic keeps coming up, and I think it does put a lot of people off.

Mining for 1 hour, stats -- 829Kh instead of ~2.4Mh.

Amen to that!

Our rejected ratio wasn't that high during this period. It was increasing daily because of strategy changes that I made and reached its peak today. Daily values for 2/14-2/17 were 5.31%, 6.30%, 7.47% and 8.90%, which gives 7% average (today it reached 11% partial-day average before I made changes to reduce this).

That should give 3.33 MH/s => 3.0969 MH/s after 7% rejects and 0.038410851 BTC/day.

I'm not trying to argue over pennies/satoshis here, just want to provide you the correct data to get your math straight.

I'm fully aware that high reject ratio is a problem at CleverMining and I'm actively investigating and working on this issue as I wrote in the post above. Last hour rejected ratio was 8.8%, compared to 14-15% in three previous hours. Current hour is 6.7%. I think the changes are working well.

I won't ever aim to get as low as 2% rejected ratio. It's not that hard to do, but it would need to be done by leaving some money on the table and net profitability is my main objective.

It's negligible compared to the statistical variance of mining luck over the time period measured. If that difference persisted for months, then for sure not negligible. But over a period of three days it is pure statistical noise and doesn't say anything conclusive one way or another.

Regarding shares rejected percentage

Coin-switch pool is always a balance between profitability and wasted hashpower. Different pools can have different rejected percentages because of different coin switching strategies and/or aggressiveness and/or mining different coins with different difficulty/speed. Our strategy is quite aggressive as I found it provides good results.

I tweaked our strategy a little bit to further discriminate high-reject coins and will see how it changes both reject ratio and profitability in the next hours. I hope we should expect lower rejected ratio from now on. It was deployed an hour ago and in the last hour pool-wide rejected percentage was 8.8% compared to 14-15% in the previous three hours. One hour is a small sample but I hope the trend will continue. It might take couple of hours with gradually lowered rejected ratio before it will settle, because of the nature of this tweak. Please let me know what your experience will be.

If you compare pools, you should compare your net profitability. Wafflepool publishes their today's (partial day) profitability as 0.0083 BTC/MHs. CleverMining profitability for the same period is 0.0108 BTC/MHs (I don't publish partial day results on the website, but I took this from database to compare exactly the same period of time). I didn't find Waffle's published pool-wide rejected ratio, but even if it's 1% (Middlecoin for example has 6.2% today), then you get 0.99 * 0.0083 = 0.0082 BTC/MHs there. Our average rejects today is 11%, so here you get 0.89 * 0.0108 = 0.0096 BTC/MHs, which is still 17% more even after wasting 11% hashpower on rejects.

We could achieve as low as 2-3% rejects but it would hurt profitability. Lowering rejects is important, but net profitability is much more important.

That said, with changes that I just made, I hope we'll be able to lower rejected ratio without compromising profitability too much. I'm tweaking and monitoring changes continuously to find the right balance.

that's a total misunderstanding of how the mining process works. every share you calculate takes (roughly) the same amount of time to calculate. your GPU's are churning through hashes as fast as they can and only submitting the ones that are "better" than the current difficulty. each one takes just as long as the one before and the one after to calculate, but only some of them are better than the current difficulty. it doesn't take any longer to calculate them - the difficulty just makes it less likely that any given calculated share will be good enough to submit.

the reason we are all so keen to have our GPU's have higher kHash rates is not because that makes them calculate a given share faster, it's because it measures how many shares they can spew out every second.

muney

Really? Maybe because you don't actually get that much?

Let's calculate it with BTC@1000 for simplicity.

That's $2.95 a day. Round that up to $3.

30 Days a month, that's $90.

If someone has a 10Mh/s setup, that's $270 difference.

Far from negligible.