Topic: [POOL][Scrypt][Scrypt-N][X11] Profit switching pool - wafflepool.com - page 238. (Read 465769 times)

I can't imagine that would be enough to be significant. If it is, run a script in the background to trigger it.

I had a blip in my hash rate up to 20 mh from my usual 4.8 recently reported by the server.  If you could flip the switch to do that again somehow and keep it going I'd really appreciate it

http://i.imgur.com/FZANUUj.jpg

you have no idea what you are talking about

im using 6 rigs
should i use worker name or its not neccery?
any difrence on last hour profit and end of the day?

I came here for the vardiff, and now we have this shit diff at 512, and mine worthless coins like litecoin. WTF?

Im going to try one of the other pools.

A lot of the comments seem to be due to the lack of understanding on how everything works (difficulty, variance, reported shares, etc). I've never trusted the shown hashrate (or any stats) from any pool front-end. I watch my miners directly, look at my profit over 24 hours and do the math myself. If the UI happens to match, great. If it doesn't, I'm not up the pool owner's @ss about it, I just wait a few more days and if it still sucks (or no communication has been sent), I switch pools.

poolwaffle, that explaining will continue... the only way to combat that is to actually show the 30 min or so average as "current hashrate" and probably have something like 3/6 hours average too. This way people will see their hashrate and be happy about it no matter what you do with the diff underneath.

I've only ignored people in the MC thread. All of sudden with the mass exodus of people there I now have doubled my ignore list thanks to this thread.

With a higher diff, only thing that's higher is rejected . That's about a max of 1% extra instead of choosing you're diff.

Still having the feeling, that Poolwaffle has a lot of love for the 6 GH/s guy, and not so much for the small guys, who supported him from the beginning. I know he's making you're money, and respect that. But I hope you understand, that how big or small you're rig is. We're also doing it for the money.

See a lot of messages on the forum, and hear people on IRC, with people who doesn't feel recognized by Poolwaffle.

Hope we get the better mood back in here, and that we can mine small coins again too . Their was a lot of money in them, that was one of the big reasons to join you, you had a great payout! Hope it's coming back quickly (the good payouts, and that everybody is happy again).

Hope you can think about the small guy too! and give us some features like changing diff's (low one's) and some smaller coins that makes the high payouts again!

Thank you in advance.

Daltonganger

Yep, I'll increase the timespan for hashrate calcs.  Right now its over 5min, which was fine for lower difficulty. Just bumping it higher means more DB access (storing more data, more calculations, etc).  It also means a longer time until people see their hashrate at what it should be.  Not too worried, just some side effects of the change.  I agree though, this is probably what is confusing people.

No worries.  Sorry if it came off as curt (I actually edited it almost immediately after).  Just got 30 things on my plate, and explaining variance 5x a day to people gets tiring

Haha, shhhhhhh.

I think the confusion comes from reported hash rate. People see their miners settling in at 40, 200, and 700, and expect to see, 940kh/s all the time. But what's easy to forget is that with harder work, when submitted, you get more shares. So while it takes longer to submit work, submitted work should average out to the same number of shares. I can't see my average shares over the last 2 weeks vs the last 8 hours. I'll just have to trust the math.

I think the biggest thing is this: Because shares per round is fluctuating so much, reported hash rate is also fluctuating. Shares over time should average out to being the same, thus reported hash rate should average out to being the same.

I should have just done more research before speaking. Sorry about that. :-)

You're probably getting swings that go from 6mh to 14mh, averaging out at your actual 10mh.

Getting huge swings in hashrate from 6 to 10mh and my miners are steady ~10mh.
Whats going on

This is the exact definition of variance.  Change the resolution from "5min" to "1hr" and you'll see its damn near identical to what it was before.  Your earnings will change on a short timescale, but on a long timescale will be the same (like looking at a 5min vs 1hr graph).

Since you fixed diff to 512, my reported hashrate and shares have not stopped fluctuating. I have a smaller CPU machine mining at 40kh/s, a 660ti mining at 200kh/s, and a 280x at 700kh/s. My lowest reported hashrate over the last 8 hours (of no changes on my end) was 185kh/s, and the highest was 1500kh/s. My average is DEFINITELY less (about ~830 instead of ~920) over the last eight hours than it was the previous eight before that.

As a result my shares have been all over the place, sometimes as low as 5500 / ~100000000 per round, sometimes as high as 12000 / ~100000000 per round.

Is this averaging out to what I was getting before? Because from what I'm seeing it isn't. Would it not be only slightly more difficult on the network to allow for hard-coded client-side diffs as low as 256 and as high as 1024? That should cover most people.

ETA: Just look at my hashrate over the last 8 or so hours. There is obviously a huge difference in saw-toothing going on.

I don't understand why hashrate would matter?  You'll just have a larger variance (payouts would be the same) based on size.  Mining with 1khs is the same as mining with 1000khs in terms of payouts (you'd receive 1/1000 of the payout).  There will be slight variations depending on when shares are actually found, but the miner difficulty doesn't affect it much at all.

If you're mining with multiple workers, it matters even less, all that matters is the overall hashrate (1.2MHs in this case)

Thanks for an awesome reply. What would you recommend as the minimum kh's to use your services going forward? I was using about 8 machines totaling 1.2mh. Some as low as 25 others as high as 350. Payout for me in the last two days were almost nonexistant.

Should I only point +200kh's to waffle from now on?

Thanks again. We appreciate your help.

too many coins disabled, payouts have been 0.005-0.008 per mhash after rejects/disconnects for the past 4 days mining 24/7. Pool was ruined when the botnet joined. To clevermining or hashcows?

Alright guys.

So I've been watching the impact of the bumped (hard-locked) difficulty of 512.  From my side, we're looking rediculously better.  All around load on servers is down by about 70% due to the adjustment.  This is from a few factors, first being we don't need to track every miner, and all of their previous shares (and times) for calculating and readjusting difficulty, second is the reduced amount of shares we have to log into the db (and then process into rounds, etc), third is the reduce in bandwidth required, which allows us more upstream bandwidth for outgoing block propagation.

All of these things combined makes for a _much_ better server-side.  From a miner standpoint, I can understand why it is a slight negative (smaller miners don't see shares nearly as often, may think something is broken).  But unless there is a major outcry about this, I think we need to stick with it.

The next thing is, we've boosted connectivity (a lot to do with bandwidth changes from difficulty), so we should be seeing fewer network-orphaned blocks, which is extremely good.  A bit of that was showing up earlier on the stats pages (from about 0.0080 --> 0.0086).

On top of that, I've just enabled ReddCoin, KarmaCoin, and PenguinCoin.  These are the first of their kind, in that we have to manually trade them (they're not on Cryptsy, where we auto-trade), and aren't listed on any exchanges that support API based trading.  What that means is two-fold:  1) Unexchanged balances will be significantly higher than normal (they wont move from unexchanged --> converted very often).  2) Unexchanged balance approximation will be off significantly more than it was previously, since we're exposed to market swings between the time they're earned, and the time they're converted (probably 2-3 times daily).  But even with the added risk, I think they should be more profitable on average.

Lastly, I've disabled a few more coins for the time being (under 15 difficulty), as when we switch to them, the block propagation time compared to block-find time is way too high, and results in us fighting ourselves for blocks (orphans).  As a simple example, say it takes about a full second for any block we find to make it around the world (propagate fully).  With coins under 15 difficulty, we have an average block time of about 7 seconds.  Meaning that 1/7th of the time, we might have already found a block, and not seen it yet (found at EU, hasn't gotten to USWest).  And we end up orphaning our own blocks about 1/7th of the time.

The long-term solution to this was discussed previously, which is to switch portions of our hashrate to lower difficulty blocks, in an attempt to keep our block-find time at a constant value (ex: never switch more power than would give us 20s block times).  But this is a reasonably large change to the profitability switcher, and the easier short-term solution is to just disable those coins until we can make those changes (a few days most likely).  So in the mean time we added additional coins (see above).

tldr; added new coins, disabled some coins, don't trust unexchanged balances

It is not just here. Go to any other pool thread and its the same way. There is an outbreak of entitlement.