[1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool

[Smoovious]

I`m very small miner, may avg power is only 150MH. When it (share diff) was ~500 i found 5-8 shares/day. Now when it is over 1000 i found nothing. Mining LTC now.

You can decrease the difficulty.  I think it'll drive up your stale rate, but I think it'll get you shares.

M

Um, no, you can't... if the difficulty you set via the /### option, is less than the current share diff, then you mine at the share diff.

-- Smoov

[Ente]

There were a lot of ideas and concepts how to solve the scaling issues (i.e. high diff drives small miners away). Ideas were several p2pools, nested p2pools, dynamic difficulty and so on.

We should really concentrate on that now. Difficulty is rising again, and soon it will all skyrocket with BFL. "We" probably means a bounty or something to get good programmers on it.
Don't forget it will take a lot of time to adjust and test! I would expect some months on testnet, before it goes productive!

Ente

[rav3n_pl]

Suggestion - when SD is rising over 500:
1. make base share diff (it will be also a minimum pool share diff) as 500
2. calculate miner hashrate in last 5/10/60 seconds and calculate miner share diff proportionally to current pool rate (make SD much higher for big miners)

This way small miners will have chance to find at least few SD500 shares per day and get any payout and bigger miners will have less shares per day than now but way powerful and get propotionally bigger (same as currrent) payout.

[Naelr]

Trying to set up a node... I think I did this correctly if anyone would like to try my Litecoin node to help me test it

http://fearmypenguin.ath.cx:9327

Let me know if I did anything wrong... I can't start the Bitcoin/namecoin node yet because my namecoind is still far behind on downloading the blockchain.... and I need to go to bed 

It might be I can't run this because of being on cable modem... but I can move it if I have it running right.

Thanks

Naelr

[RandomQ]

Trying to set up a node... I think I did this correctly if anyone would like to try my Litecoin node to help me test it

http://fearmypenguin.ath.cx:9327

Let me know if I did anything wrong... I can't start the Bitcoin/namecoin node yet because my namecoind is still far behind on downloading the blockchain.... and I need to go to bed 

It might be I can't run this because of being on cable modem... but I can move it if I have it running right.

Thanks

Naelr

I tested your LTC pool, and everything works. My miner was able to connect and I was able to see your stats page.

[rav3n_pl]

Trying to set up a node... I think I did this correctly if anyone would like to try my Litecoin node to help me test it

http://fearmypenguin.ath.cx:9327

Let me know if I did anything wrong... I can't start the Bitcoin/namecoin node yet because my namecoind is still far behind on downloading the blockchain.... and I need to go to bed 

It might be I can't run this because of being on cable modem... but I can move it if I have it running right.

Thanks

Naelr

You CAN start bitcoin merged node, it will just not work correctly till Xcoind`s are downloading blocks
LTC node works fine.

[Naelr]

Thanks for testing guys... I think I am gonna leave it up so if anyone wants to use it as fail over... feel free.... that Linux box stays on all the time anyway...

I set it up as a 0.5% fee is that too much, on par, or being an ass?  going to work on the other tonight.... please feel free to give more feedback

thanks

Naelr

[Naelr]

Ok now I have another question... when I look at my data ...

Shares: 225 total (17 orphaned, 14 dead) Efficiency: 106.1%

that doesn't look too bad but when I look at when I had my gpu in the mining I see in the graphs

http://fearmypenguin.ath.cx:9327/static/graphs.html  

while I was mining with my gpu is when I think I got most of the orphaned and dead shares... how can I tweak this , can I tweak this, is this something I have never seen before because I am usually on a PPS site?

Naelr

[rav3n_pl]

Ok now I have another question... when I look at my data ...

Shares: 225 total (17 orphaned, 14 dead) Efficiency: 106.1%

that doesn't look too bad but when I look at when I had my gpu in the mining I see in the graphs

http://fearmypenguin.ath.cx:9327/static/graphs.html  

while I was mining with my gpu is when I think I got most of the orphaned and dead shares... how can I tweak this , can I tweak this, is this something I have never seen before because I am usually on a PPS site?

Naelr

This is probably miner config issue.
If you are using cgminer make sure you have set:
"queue" : "0"
"gpu-threads" : "1"

[zvs]

Ok now I have another question... when I look at my data ...

Shares: 225 total (17 orphaned, 14 dead) Efficiency: 106.1%

that doesn't look too bad but when I look at when I had my gpu in the mining I see in the graphs

http://fearmypenguin.ath.cx:9327/static/graphs.html  

while I was mining with my gpu is when I think I got most of the orphaned and dead shares... how can I tweak this , can I tweak this, is this something I have never seen before because I am usually on a PPS site?

Naelr

I dunno, I'd say 225 shares w/ 17 orphaned and 14 dead is pretty bad...  the efficiency rating may have been high at that particular moment, but most of the time it'd be under 100%... those figures are 15% stale+dead..    Having peers with fast ping times + bitcoind w/ many connections helps

[RandomQ]

Ok now I have another question... when I look at my data ...

Shares: 225 total (17 orphaned, 14 dead) Efficiency: 106.1%

that doesn't look too bad but when I look at when I had my gpu in the mining I see in the graphs

http://fearmypenguin.ath.cx:9327/static/graphs.html  

while I was mining with my gpu is when I think I got most of the orphaned and dead shares... how can I tweak this , can I tweak this, is this something I have never seen before because I am usually on a PPS site?

Naelr

I dunno, I'd say 225 shares w/ 17 orphaned and 14 dead is pretty bad...  the efficiency rating may have been high at that particular moment, but most of the time it'd be under 100%... those figures are 15% stale+dead..    Having peers with fast ping times + bitcoind w/ many connections helps

Having bitcoind running in the above problem with 1000 or 1 connections will have no effect on efficiency  (Port 9327 AKA LTC)

But the longer you run LTC p2pool the more connections litecoind has the higher the efficiency rate goes up.
When i start up i always get sub 100% efficiency for a day or so before it goes over 100%

Just to compare stats my LTC p2pool has been running only 1 day more than yours.
I have the same number of incoming and outgoing connections.
my doa rate is .14% lower
I have alot less shares than you, 79 total (5 orphaned, 0 dead) Efficiency: 111.1%
and I'm using minerd only not cgminer

Your DOA rate seems to be proportional to you Hash rate, your hashing @ 500kHs with a doa of maybe 30kHs.

plus i'm sure cgminer has a higher DOA rate because its LTC mining is very recently added to it,and i'm sure it will get better over time.

[zvs]

Ok now I have another question... when I look at my data ...

Shares: 225 total (17 orphaned, 14 dead) Efficiency: 106.1%

that doesn't look too bad but when I look at when I had my gpu in the mining I see in the graphs

http://fearmypenguin.ath.cx:9327/static/graphs.html  

while I was mining with my gpu is when I think I got most of the orphaned and dead shares... how can I tweak this , can I tweak this, is this something I have never seen before because I am usually on a PPS site?

Naelr

I dunno, I'd say 225 shares w/ 17 orphaned and 14 dead is pretty bad...  the efficiency rating may have been high at that particular moment, but most of the time it'd be under 100%... those figures are 15% stale+dead..    Having peers with fast ping times + bitcoind w/ many connections helps

Having bitcoind running in the above problem with 1000 or 1 connections will have no effect on efficiency   (Port 9327 AKA LTC)

But the longer you run LTC p2pool the more connections litecoind has the higher the efficiency rate goes up.
When i start up i always get sub 100% efficiency for a day or so before it goes over 100%

Just to compare stats my LTC p2pool has been running only 1 day more than yours.
I have the same number of incoming and outgoing connections.
my doa rate is .14% lower
I have alot less shares than you, 79 total (5 orphaned, 0 dead) Efficiency: 111.1%
and I'm using minerd only not cgminer

Your DOA rate seems to be proportional to you Hash rate, your hashing @ 500kHs with a doa of maybe 30kHs.

plus i'm sure cgminer has a higher DOA rate because its LTC mining is very recently added to it,and i'm sure it will get better over time.

neg

running bitcoind with more connections would (in most cases) mean you'd receive a block faster, hence less DOA

or, in this case, litecoind, or whatever the hell it is

[RandomQ]

neg

running bitcoind with more connections would (in most cases) mean you'd receive a block faster, hence less DOA

or, in this case, litecoind, or whatever the hell it is

I know I was just playing around with you, I know you meant litecoind. 

The Max amount of LTC incoming connections my P2Pool has seen was 25 around a month ago. The Pool in question only has around 9, so right now its at under 50% of total LTC p2pool running.

[RandomQ]

Ok now I have another question... when I look at my data ...

Shares: 225 total (17 orphaned, 14 dead) Efficiency: 106.1%

that doesn't look too bad but when I look at when I had my gpu in the mining I see in the graphs

http://fearmypenguin.ath.cx:9327/static/graphs.html  

while I was mining with my gpu is when I think I got most of the orphaned and dead shares... how can I tweak this , can I tweak this, is this something I have never seen before because I am usually on a PPS site?

Naelr

This is probably miner config issue.
If you are using cgminer make sure you have set:
"queue" : "0"
"gpu-threads" : "1"

I just wanted to point out he was mining LTC with CGminer not BTC.

I don't think we have a consensus on cgminer miner setting for LTC on p2pool.


 --shaders
would depend on what card you are running

--intensity
Goes from 0 to 20( we use mainly 7 for BTC, what is everyone else using for this)

option settings
-g #
cgminer is a little vague on this on LTC(setting 1 to 4)

--thread-concurrency:
Overides shaders, 5x shaders for 5XXX, 4x shaders for 6xxx & 7xxxx

--lookup-gap
default setting of 2

Also memory speed matters for LTC mining with Cgminer, so don't underclock it for best performance.

What is everyone else running? Because there are at least 16 of you running GPU's on p2pool? And some of you have to be running cgminer

[Naelr]

Ok now I have another question... when I look at my data ...

Shares: 225 total (17 orphaned, 14 dead) Efficiency: 106.1%

that doesn't look too bad but when I look at when I had my gpu in the mining I see in the graphs

http://fearmypenguin.ath.cx:9327/static/graphs.html  

while I was mining with my gpu is when I think I got most of the orphaned and dead shares... how can I tweak this , can I tweak this, is this something I have never seen before because I am usually on a PPS site?

Naelr

This is probably miner config issue.
If you are using cgminer make sure you have set:
"queue" : "0"
"gpu-threads" : "1"

I just wanted to point out he was mining LTC with CGminer not BTC.

I don't think we have a consensus on cgminer miner setting for LTC on p2pool.


 --shaders
would depend on what card you are running

--intensity
Goes from 0 to 20( we use mainly 7 for BTC, what is everyone else using for this)

option settings
-g #
cgminer is a little vague on this on LTC(setting 1 to 4)

--thread-concurrency:
Overides shaders, 5x shaders for 5XXX, 4x shaders for 6xxx & 7xxxx

--lookup-gap
default setting of 2

Also memory speed matters for LTC mining with Cgminer, so don't underclock it for best performance.

What is everyone else running? Because there are at least 16 of you running GPU's on p2pool? And some of you have to be running cgminer

But I am not mining with cgminer on Litecoins... I am using reaper for GPU and minerd for CPU  does that matter?

[Naelr]

Someone please give my BTC merged with namecoin node a try and let me know.

Thanks!!!

[zvs]

Someone please give my BTC merged with namecoin node a try and let me know.

Thanks!!!

worked for me

[zvs]

mine here!   nogleg.com:9332

Local rate: 4.41GH/s (0.97% DOA)

Shares: 75 total (0 orphaned, 0 dead) Efficiency: 108.3%

i need my transactions with 0 fees to process, hoho

2012-08-16 10:30:09.399664 New work for worker! Difficulty: 0.999985 Share difficulty: 756.956939 Total block value: 50.885716 BTC including 3163 transactions

calling all ppls with low priority transactions!

still

2012-08-16 10:45:08.291954 New work for worker! Difficulty: 0.999985 Share difficulty: 673.211956 Total block value: 50.110450 BTC including 2186 transactions

aha

[Naelr]

OK when merged mining on p2pool where do the namecoins go?  

so I login with cgminer -o http://localhost:9332 -u my payout address -p x  

so how does node know where to send namecoins?  do I have to do another switch with namecoin ?  I am currently searching for this answer but nothing has turned up yet... perhaps my google-fu today is off the mark...

thanks

Naelr

[zvs]

OK when merged mining on p2pool where do the namecoins go?  

so I login with cgminer -o http://localhost:9332 -u my payout address -p x  

so how does node know where to send namecoins?  do I have to do another switch with namecoin ?  I am currently searching for this answer but nothing has turned up yet... perhaps my google-fu today is off the mark...

thanks

Naelr

it seems to just pull it directly from your namecoin client..  i was searching for the same thing, but after solving a block, i also had the namecoin merged stuff incoming as well (w/o doing anything extra)