Introducing CherryPicking - new Windows & Linux Pool Hopper - page 21.

MrWizard

sr. member

Activity: 252

Merit: 250

Quote from: hawks5999 on August 14, 2011, 12:19:27 AM

Quote from: Milkshanks on August 14, 2011, 12:15:35 AM

Btw, I'm wondering, what are those numbers that are used for calculating the most efficient pool on CP, like:

14-08-11 02:11:20 * Most efficient pool is: 2bitcoinlc
14-08-11 02:11:20 0.8157297647834125

And why is ArsBitcoin always stuck at

14-08-11 02:10:10 * 0arsbitcoin
14-08-11 02:10:11 * Pool update done
14-08-11 02:10:11    Type: PPS
14-08-11 02:10:11    Hash rate: 441.215 GH/s
14-08-11 02:10:11    1.0022664687283978

Is that normal?

that's some internal algorithm - bloodred's secret sauce. But the further below 1 the greater the efficiency of the pool (1 likely equals 100% efficient)

And Ars being a straight up PPS pool it is always going to produce right around 100% efficiency so it will always be right around 1.00.

1.0022664687283978 = 1 + 1/441.215

i.e. 1 + 1/hashrate.

Milkshanks

member

Activity: 100

Merit: 10

Quote from: hawks5999 on August 14, 2011, 12:19:27 AM

Quote from: Milkshanks on August 14, 2011, 12:15:35 AM

Btw, I'm wondering, what are those numbers that are used for calculating the most efficient pool on CP, like:

14-08-11 02:11:20 * Most efficient pool is: 2bitcoinlc
14-08-11 02:11:20 0.8157297647834125

And why is ArsBitcoin always stuck at

14-08-11 02:10:10 * 0arsbitcoin
14-08-11 02:10:11 * Pool update done
14-08-11 02:10:11    Type: PPS
14-08-11 02:10:11    Hash rate: 441.215 GH/s
14-08-11 02:10:11    1.0022664687283978

Is that normal?

that's some internal algorithm - bloodred's secret sauce. But the further below 1 the greater the efficiency of the pool (1 likely equals 100% efficient)

And Ars being a straight up PPS pool it is always going to produce right around 100% efficiency so it will always be right around 1.00.

I see, thanks for the heads up, I was finding it kinda weird since it was stuck at this number since like, forever.

hawks5999

full member

Activity: 168

Merit: 100

Quote from: Milkshanks on August 14, 2011, 12:15:35 AM

Btw, I'm wondering, what are those numbers that are used for calculating the most efficient pool on CP, like:

14-08-11 02:11:20 * Most efficient pool is: 2bitcoinlc
14-08-11 02:11:20 0.8157297647834125

And why is ArsBitcoin always stuck at

14-08-11 02:10:10 * 0arsbitcoin
14-08-11 02:10:11 * Pool update done
14-08-11 02:10:11    Type: PPS
14-08-11 02:10:11    Hash rate: 441.215 GH/s
14-08-11 02:10:11    1.0022664687283978

Is that normal?

that's some internal algorithm - bloodred's secret sauce. But the further below 1 the greater the efficiency of the pool (1 likely equals 100% efficient)

And Ars being a straight up PPS pool it is always going to produce right around 100% efficiency so it will always be right around 1.00.

Milkshanks

member

Activity: 100

Merit: 10

Btw, I'm wondering, what are those numbers that are used for calculating the most efficient pool on CP, like:

14-08-11 02:11:20 * Most efficient pool is: 2bitcoinlc
14-08-11 02:11:20 0.8157297647834125

And why is ArsBitcoin always stuck at

14-08-11 02:10:10 * 0arsbitcoin
14-08-11 02:10:11 * Pool update done
14-08-11 02:10:11    Type: PPS
14-08-11 02:10:11    Hash rate: 441.215 GH/s
14-08-11 02:10:11    1.0022664687283978

Is that normal?

hawks5999

full member

Activity: 168

Merit: 100

bitclockers is upgrading a server right now (and in their IRC channel it seems that it is not going well). So I don't think they are faking so much as having back-end issues. Putting the sharecount to 999999999 ensures they don't get hopped at this time, decreasing their load on the infrastructure they are trying to fix/upgrade.

muyoso

member

Activity: 84

Merit: 10

Bitclockers is absolutely messing with stats, and your config on the prior page works perfectly Bloodred.

Milkshanks

member

Activity: 100

Merit: 10

Anyone else having problems connecting to bitclockers? I'm getting a lot of I/O Errors and RPC problems.

Quote from: hawks5999 on August 12, 2011, 10:36:23 AM

what pools are you mining and have noticed any issues in the console or are shares being accepted?

How many BTC are pending at the pools you are mining?

I can see CP is hopping mostly between Mt.Red, BitHashers, Bitclockers, LowCost and most of the time mining at Ars. I'm using however a config with all the PROP pools avaliable in that config pack that Bloodred posted. I had some issues with United (up to 10% stale) and some connection issues with bitclockers as well. I thought CP was supposed to hop more, but I guess no pool got new blocks right now Sad

muyoso

member

Activity: 84

Merit: 10

Awesome. I am going to keep an eye on bitclockers and if they mess with their stats again I'll be sure to use that new config.

I don't know what the best way forward is for polmine or bitcoinpool, but both were amazing pools. bitcoinpool had like 6 blocks today alone for a massive payout. Would love to have something mining them even if its not jumping at exactly 43%.

Bloodred

member

Activity: 112

Merit: 10

I've made a .cfg that scrapes bitclockers' web page instead of JSON, you can use this if you want. Just got home and wrote this, haven't tested it but I think it's good. I'll post it here then add it to the archive as well:

Code:

;BitClockers Web scrape
Type=PROP
Host=pool.bitclockers.com
Port=8332
JSON=http://bitclockers.com/statistics
HashEx1=Hash Rate:\s*\d+\.?\d*\s*Gh/s
HashEx2=\d+\.?\d*
SharesEx1=Shares This Round:[ ,0-9]+
SharesEx2=[,0-9]+
SharesRemove=,
Div=1

Quote from: muyoso on August 13, 2011, 02:34:43 PM

Edit2: Also we need to get polmine back working. There are many ways we could do this, but i think the easiest would be to go here: https://polmine.pl/?action=statistics and use the current round duration to instigate a hop to the pool. When to hop away would be another story. . .

I'll have to think of a not so complicated system that could handle parsing different kinds of round time formats. Hopping off isn't a problem, the point would be somewhat less accurate because it depends on hash rate, I don't think anything can be done about that though. The average theoretical round duration in seconds for a miner (or pool) is this:

Code:

T(H) = 2^32 * diff/H

where H is the hash rate in hashes/s
So as long as the parsed round duration is less than 0.43 * T(H) you're in the optimal period. This isn't as accurate as share-based hopping because the hash rate fluctuates and is averaged out, but I think it should be good enough to not cause any significant efficiency decrease.

scatterbrain

member

Activity: 72

Merit: 10

how do you receive an invitation code for hoppersden?

muyoso

member

Activity: 84

Merit: 10

Quote from: ?? on ??

Quote from: muyoso on August 13, 2011, 12:55:13 PM

Bitcoinpool is faking their stats. They are resetting their JSON several times per round to keep up the hash rate. IDK if this is possible, but I wonder if we could take this page:

http://bitcoinpool.com/index.php?page=1&ipp=All&do=currentround

and add up all of the user shares for an accurate number of shares per round. Is there a way to do this automatically?

Edit: Bitclockers may be doing the same thing.

plz take a look here => http://hoppersden.info/project.php?issueid=40 and brainstorm about that basic idea Grin

That isn't a bad idea, but detecting a pool faking stats isn't a problem to me. I can check and determine that in a couple minutes. I just want to get around their faking of stats. We need to change bitclockers to scrape the website instead of using the json and we need to add up all of the user shares for bitcoinpool to get around the JSON faking stats.

Edit: Seems bitclockers is back to normal JSON stats.

Edit2: Also we need to get polmine back working. There are many ways we could do this, but i think the easiest would be to go here: https://polmine.pl/?action=statistics and use the current round duration to instigate a hop to the pool. When to hop away would be another story. . .

muyoso

member

Activity: 84

Merit: 10

Bitcoinpool is faking their stats. They are resetting their JSON several times per round to keep up the hash rate. IDK if this is possible, but I wonder if we could take this page:

http://bitcoinpool.com/index.php?page=1&ipp=All&do=currentround

and add up all of the user shares for an accurate number of shares per round. Is there a way to do this automatically?

Edit: Bitclockers may be doing the same thing.

Bloodred

member

Activity: 112

Merit: 10

See the 2nd post in this topic in order to fix the errors like the one you posted from mtred. Basically you need to add their SSL certificates so Java will trust them.
If it's telling you your username or password are wrong they probably are, that is a message coming from poclbm, not CherryPicking.

Houseonfire

full member

Activity: 129

Merit: 100

I'm getting

[Miner0] Wrong username or password.

I'm not really sure why.

And also, for many of the miners, I'm getting login errors. I'm not sure why. Might be a mistake on my part I guess but as far as I can tell I'm doing everything right.

Mtred for example.

Code:

;MtRed
Type=PROP
Host=mtred.com
Port=8337
JSON=https://mtred.com/api/user/key/17e1395c3a140715cab6755aa52c4410
HashEx1="hashrate":\d+\.?\d*
HashEx2=\d+\.?\d*
SharesEx1="roundshares":\d+
SharesEx2=\d+
Div=1
Miner=houseonfire 321g

I get this error when I try to log in

Code:

* 8mtred
* Error occured while trying to communicate or open a connection to https://mtre
d.com/api/user/key/8ad572a45e942ed5f78af1e4aeb0b9e4
* sun.security.validator.ValidatorException: PKIX path building failed: sun.secu
rity.provider.certpath.SunCertPathBuilderException: unable to find valid certifi
cation path to requested target
* Update error or pool considered invalid (lagging or down)

I'm not too sure what to do.

scatterbrain

member

Activity: 72

Merit: 10

i just got through my first period of running CP for 24 hours. i have 21 of the pools from the cfg file active and am running with the STATIC_FAST argument. 2.8Gh/s has made me 1.6 BTC.

before i was split betweeen slush, bitclockers and bitcoin.lc and was getting anywhere from .8 to 1.25 per 24h, so CP is definitely an improvement. i can't wait to see the effect when we get a few short rounds strung together.

Bloodred

member

Activity: 112

Merit: 10

That's right, hopping is completely dependent on pool luck. Short rounds bring huge benefits while long rounds won't bring you any at all.

ampirebus

full member

Activity: 672

Merit: 100

Quote from: Milkshanks on August 12, 2011, 10:18:33 AM

Seriously, I think I must be doing something dead wrong, I've been mining with CherryPicking for 5 days no with 650ish mhash/s and I've only mined 1.1 BTC so far, that's roughly less than I'd mine staying on BTCGuild. Sad

to be fair the past 2 days have been horrible luck for most pools having rounds lasting multiple days. If i would have stayed @ slush for the past 48 hours I'd have .7 BTC
with cherrypicker I have .86 BTC. (this isnt a high% increase but still an increase) compared to the first weekened I u sed CP and i was getting nearly 1BTC/24hrs. Once pools have multiple shorter rounds (a few hours long or not even) you will notice huge gains. for times like this with all pools having really long rounds the gains will be small or possibly even the same or less than if you were just mining a PPS/low variance pool.

hawks5999

full member

Activity: 168

Merit: 100

what pools are you mining and have noticed any issues in the console or are shares being accepted?

How many BTC are pending at the pools you are mining?

Milkshanks

member

Activity: 100

Merit: 10

Seriously, I think I must be doing something dead wrong, I've been mining with CherryPicking for 5 days no with 650ish mhash/s and I've only mined 1.1 BTC so far, that's roughly less than I'd mine staying on BTCGuild. Sad

hawks5999

full member

Activity: 168

Merit: 100

Quote from: max in montreal on August 11, 2011, 09:32:15 PM

Anyone here successfully mining on mainframe? From the start I keep getting update error or pool considered invalid

you don't want to mine there. Whenever I've hopped or picked there, i've ended up with 0.0000202594021739130 per share. Better to be on backup/pps.

Topic: Introducing CherryPicking - new Windows & Linux Pool Hopper - page 21. (Read 43136 times)