Topic: [POOL][Scrypt][Scrypt-N][X11] Profit switching pool - wafflepool.com - page 222. (Read 465769 times)

So is there any way to implement this "merged mining" I'm hearing about? Is this even really feasible or doable?

For those who haven't heard of this, it's some sciencey-mathy-magicky trick where you can mine several coins at once, all of them at your top hash rate. Not split between them, all of them at once at full hash. At the same time.

I now realize that the cpuminer does support stratum and I was using it and not realizing it. I will verify the latency to the various servers to make sure I didn't miss something. I did wipe my VPS before reconfiguring for MC, so maybe I made a mistake earlier and couldn't find it. I hope to get this sorted out and mining back at Waffle soon. Thanks!

I really don't know how I missed that after all this time. I have actually ran some instances of it without the stratum proxy and it worked and I never even put that together. Most mining pools still have instructions in the "Getting Started" or FAQ sections saying that minerd (pooler cpuminer) doesn't support stratum and that you should go get the stratum proxy. I've been mining for 2.5 years and feel like a fool to have not noticed that and just believed all those instructions because it was stated over and over. But yes, I see at the source that it does support it. Thanks for clearing that up for me.

As for the pinging, I was getting < 20ms from cpu to stratum proxy, 12ms to VPS, and 17ms to Wafflepool. I thought that it would be low enough, but maybe not?

This ^ no direct stats makes those who wanted to opt-out happy and shuts the voyeurs up that wanted to see said stats and were mad when their favourite hash master was removed .

Personally, I'm not a fan of dropping somebody's stats from the page without getting rid of the whole thing. The transparency is a prime reason why I chose this pool over/after MC. This flies right in the face of that. Now I honestly don't believe their is anything malicious/wrong about it in this instance, but it sure wasn't transparent and practically begs the conspiracy theories to start up.

Personally I think it should haven to 3 choices. Remove the page completely, add all users back in, or put a banner on that page stating that the stats are not accurate as users are allowed to opt-out by sending an email to xxx.

Noted.  I've had a few other people email me about this.  My answer thusfar has been "I have more pressing stuff to attend to".  I'd of course be happy about an added donation, but I don't feel right taking it until the pool is the best it can be.  I've got things that I'd like to get done regardless of donation factors, but they're on my list eventually (easy answer is a "donate=0.5" password for a half percent donation - done some time in the future).


Absolutely not trying to call question to this, if MC is performing better for you, thats nothing I can argue with, and you should do due diligence to make sure!  That said, minerd (cpuminer/pooler) has full support for stratum, and running it through a proxy will only negatively affect it.  Please also make sure you're connected to the closest endpoint (closest meaning ping-wise, not necessarily geographical location).


Grouping these two together.
Unfortunately, its less a matter of getting more clusters.  Although more clusters breaks us up into smaller distinguishable groups (instead of having 4 USEast servers, we might have 8-10), it doesn't get rid of the problems of the USEast servers been 200-300ms away from our EU cluster (regardless of how large that cluster is).  Although I appreciate the (assumed) sentiment of a larger fee % would afford more servers would assume slightly larger earnings, I think a larger part of it is just playing with our algorithms and getting things tweaked for where we are currently.  Paying 3% fees, for a 4% boost is obviously something everyone would agree to (yay math), but I think we can get a 5-10% boost without adding the fees yet, and I'd like to focus on that first   That said, if it comes down to it later, and I can reasonably-well promise and increase in earnings compared to an increase in fees (hiring someone, more servers, etc), I'll definitely bring it up when I have hard numbers.

The second part, throwing too much hashrate at small coins, is an issue I've mentioned in the past, and I think we have a good solution for it.  That said, I don't want to mention the whole thing here (I think its at least minorly innovative), and ask for a few more days of time to code/test it out.  The unfortunate thing here, is I can code up a new idea, wait 4-6 hours to see some results, and have the results skewed by luck pretty heavily still.  So it takes a few full days to test a new branch of code fully.  Sucks, but its the nature of the beast.  And as long as we're in line with the competition, there will be some days where our testing underperforms (bad code, bad luck, etc), and days where we overperform (good code, good luck, etc), and its unfortunately, very hard to tell the two apart.


Please report back, as this is a great piece of info.  We can obviously claim 10% higher stats, if 10% of everyones hashrate gets rejected   I'm personally too involved to give any sort of non-biased stats, and I appreciate anyone that can come back and say "from time A to B, I earned C on wafflepool, and D on ".  WP has things to learn from other pools, they have things to learn from us

I gotta admit, I never had problems with waffle. Really low reject rates too.

I also admit that I am testing out clevermining to see if their claims of profitability match their stats. My reject rate there though.... 

My average after 3 days is now exactly even with clientside cgminer and I expect to surpass it in the coming days. Best. Pool. Ever. If you're not mining Waffles, well, you're not really mining at all

Was upgrading the app. It's back up now, and has a whole bunch of new changes. :-)

The problem is throwing too much hashing power at the small coins. Adding more clusters will throw more hashing power and overwhelm it. A better option is being able to somehow fine tune how large the cluster is that is thrown at it.

Thank you for the answer and thanks for the work you are doing. I guess even different sets of coins for clusters will do the job and will let us get some profit from those low difficuly coins no matter how much more will wafflepowah grow.
And if this is the esiest way to spread gh/s around for the sake of using those <20 coins, would 1.5% or even 2% fee cover your expenses, to run 2-3 more clusters? You do a lot of math anyways, what are your estimates, will it lead to an increase in average profit for miners over today's 0.0098?

working.

Is waffles stats working for you guys?

I'm getting a "502 Bad Gateway" 

poolwaffle,
Are there any bonuses for Blocks Found or statistics on the member page Blocks Found - it's just information? Thank you.

PS regarding letters of flipped funds from an account at a dead worker - is it possible?

minerd=cpuminer? I ask because I use cpuminer and the executable is called minerd, and it has stratum.

Did you ran some network tests from your VPS to the pool servers? (ping traceroute etc.) Because it looks like your shares are "killed" by latency as they travel.

Great info, Dawg75.

Just curious if you'd share where you're located?

Thanks for your updates poolwaffle.

As an update for my 11 miners, I am now seeing an average of 28.16MH/s accepted which is almost dead on what I see from the client side.

In 83.85 hours I now see a 0.01048056 BTC/MH average (highest is 0.01217259 and lowest is 0.00996523). 

The average was pushed down a good bit over the past 24 hours. 

And even then, it never seems to be in much of an order (at least that I can see?) so it's not even really useful for that.

I say either get rid of it or actually sort the hashrates in descending order.

I'm using minerd miners (no default stratum support) connected to a central stratum proxy. That stratum proxy connects to a headless linux VPS forwarding all traffic to whatever pool I choose. All I did was change the IP on the VPS to point to Middlecoin instead of Wafflepool and all of my rejects went away. So something else is going on, because my miner configurations haven't changed at all since I started mining LTC a year ago. I guess I was just having connection issues to just Wafflepool.

Definitely sounds like something wrong with your stratum proxy.  Are you using the stratum CPU miner, or longpoll?

--

@wafflepool: I would say replace the miners page with a summary of how many miners we have in various hashrate brackets.  Something like:

51.2MH+: 4 miners
25.6MH-51.2MH: 6 miners
12.8MH-25.6MH: 10 miners
6.4MH-12.8MH: 30 miners
3.2MH-6.4MH: 20 miners
1.6MH-3.2MH: 20 miners
800KH-1.6MH: 200 miners
400KH-800KH: 130 miners
200KH-400KH: 150 miners
100KH-200KH: 100 miners
0-100KH/s: 100 miners

Provides an interesting stats summary, whereas the current miners page has grown fairly useless except as a competition to see where you sit in the pool.