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

[twmz]

Ok this is correct, I am in the second one ("1N9mMCgV1CzU3s8Kvf7BcQhBBXGgoo9Wx1": 4.43964579) and not the first one and I am a small miner (252 Mhash) so I could very well fall into the small miner lottery. Thank you

One last question and I'll stop bothering you : What does the 4.43964579 mean? (amount of shares would seem weird due to the fractional number)

The second URL was "how would 1000 BTC be divided up between p2pool miners".  4.43964579 means that you would have received 4.43964579 BTC out of the 1000 BTC.  Of course no one would likely donate 1000 BTC   I just picked a very large number to make sure that no one would fall below the 0.01 cutoff.

[P4man]

I just pointed my non dedicated miner to p2pool to give it a try.
It seems to work (hurray!), but I have at least one question: why does cgminer report so many longpoll events when mining on p2pool?

Here is recent sample:

Code:

2012-02-08 21:34:41] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:34:46] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:34:52] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:34:57] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:06] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:09] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:22] Accepted 00000000.e626df82.c92c7a5e GPU 0 thread 0 pool
2012-02-08 21:35:25] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:30] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:34] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:37] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:39] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:42] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:46] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:55] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:35:58] Accepted 00000000.8e5eab72.9e42f8e9 GPU 0 thread 0 pool
2012-02-08 21:36:04] Accepted 00000000.652b3770.be6a3762 GPU 0 thread 1 pool
2012-02-08 21:36:06] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:36:07] Accepted 00000000.8072fb91.64a7cfea GPU 0 thread 0 pool
2012-02-08 21:36:12] Accepted 00000000.4dc0dc0c.77f4e052 GPU 0 thread 0 pool
2012-02-08 21:36:14] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:36:18] Accepted 00000000.9ebbb9a6.7e98eaad GPU 0 thread 1 pool
2012-02-08 21:36:26] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:36:30] LONGPOLL requested work restart, waiting on fresh work
2012-02-08 21:36:31] Accepted 00000000.9aabb60a.99a24530 GPU 0 thread 1 pool
2012-02-08 21:36:31] LONGPOLL requested work restart, waiting on fresh work

On traditional pools, youd expect about 1 longpoll every ~10 minutes (or maybe 4 or so when merged mining).  Why does p2pool generate so many more? Or is something wrong with my config?

[cabin]

I just pointed my non dedicated miner to p2pool to give it a try.
It seems to work (hurray!), but I have at least one question: why does cgminer report so many longpoll events when mining on p2pool?
On traditional pools, youd expect about 1 longpoll every ~10 minutes (or maybe 4 or so when merged mining).  Why does p2pool generate so many more? Or is something wrong with my config?

Once every ~10 seconds is normal. That is how often p2pool shares are generated (as opposed to ~10 min for bitcoin blocks)

[Gabi]

@p4man: it's perfectly fine. That's how p2pool work



Aaand... we are over 210ghash/s

[Tittiez]

Finally recieving my payments for mining 6-7 hours the last three nights and 6 hours today. Sitting at .45 Bitcoin!

[P4man]

Good to know its normal. And Im sure there is a good reason for it, but arent you causing a lot of stales by having to force a restart every 10 seconds?

[forrestv]

Good to know its normal. And Im sure there is a good reason for it, but arent you causing a lot of stales by having to force a restart every 10 seconds?

Yes, but those stales aren't lost. They can still be block solutions, and payouts are computed by your number of shares relative to everyone else's, so if everyone gets the same amount of stales, payouts are completely fair.

However, some miners try to minimize network overhead by dropping share solutions when a long poll response comes in. This could possibly cause P2Pool to lose a true block solution (the same applies to any merged mining pool), so people should run CGminer with --submit-stales or use newer versions of CGminer that do that automatically. Poclbm also has this behaviour with no way to disable it, as of yet ... I'll put in a pull request.

[P4man]

Good to know its normal. And Im sure there is a good reason for it, but arent you causing a lot of stales by having to force a restart every 10 seconds?

Yes, but those stales aren't lost. They can still be block solutions, and payouts are computed by your number of shares relative to everyone else's, so if everyone gets the same amount of stales, payouts are completely fair.

However, some miners try to minimize network overhead by dropping share solutions when a long poll response comes in. This could possibly cause P2Pool to lose a true block solution (the same applies to any merged mining pool), so people should run CGminer with --submit-stales or use newer versions of CGminer that do that automatically. Poclbm also has this behaviour with no way to disable it, as of yet ... I'll put in a pull request.

Okay, thanks for the explanation. I guess I have some reading to do

[wachtwoord]

Does Phoenix submit stales?

[forrestv]

Does Phoenix submit stales?

Looks like "no". Going to submit a pullreq for that too. EDIT: Did, see https://github.com/jedi95/Phoenix-Miner/pull/3

[m3ta]

Or is something wrong with my config?

No, but there is something wrong with you not reading what was discussed here ad nauseam before asking.

[P4man]

Or is something wrong with my config?

No, but there is something wrong with you not reading what was discussed here ad nauseam before asking.

If its asked ad nauseum, perhaps  it should be put it in the first post. I read the first 4 or 5 pages of this thread before asking anything, I dont think you can expect everyone to read all 25

[forrestv]

If its asked ad nauseum, perhaps  it should be put it in the first post. I read the first 4 or 5 pages of this thread before asking anything, I dont think you can expect everyone to read all 25

It's on the wiki, as of an hour ago. https://en.bitcoin.it/wiki/P2Pool#Frequently_Asked_Questions

[DiabloD3]

Does Phoenix submit stales?

But DM does! I don't emit the extension name for it though.

[wachtwoord]

So if I use Phoenix and have about 1% stales in the miner this will really cost me 1% as opposed to using another miner? Damn it, I have Phoenix optimized and now I have to it for some other miner too.

Any ideas on which is best?

[forrestv]

So if I use Phoenix and have about 1% stales in the miner this will really cost me 1% as opposed to using another miner? Damn it, I have Phoenix optimized and now I have to it for some other miner too.

Any ideas on which is best?

1% is pretty normal.. I have 0.8% right now with CGminer. Other miners won't necessarily be much better. Just pull in my change from https://github.com/jedi95/Phoenix-Miner/pull/3 if you're going to use Phoenix!

[wachtwoord]

I knew 1% was okay, but if other miners submit stales I need Phoenix to do this as well (or switch miners) otherwise I lose out on 1% efficiency.

Is there some link I can trace to see if it is incorporated in the main build (of the main developers accept your commit)? (I am using Windows and usually building things yourself on none-Linux systems is a real pain,so I usually end up waiting for the windows binary like a regular user ).

[rjk]

FEATURE REQUEST: Could P2Pool load config values from a file, and accept a parameter to tell it where to put its data? This would be nice to have, instead of passing a massive long command line to it.

[Red Emerald]

FEATURE REQUEST: Could P2Pool load config values from a file, and accept a parameter to tell it where to put its data? This would be nice to have, instead of passing a massive long command line to it.

+1

[forrestv]

I knew 1% was okay, but if other miners submit stales I need Phoenix to do this as well (or switch miners) otherwise I lose out on 1% efficiency.

Is there some link I can trace to see if it is incorporated in the main build (of the main developers accept your commit)? (I am using Windows and usually building things yourself on none-Linux systems is a real pain,so I usually end up waiting for the windows binary like a regular user ).

The link I gave will tell you what they do with it.