p2pool - Decentralized, Absolutely DoS-Proof, Pool Hopping-Proof Pool [archival]

[forrestv]

I'm mining with p2pool, running forrestv-p2pool-27ab63b / forrestv-poclbm-5590e7c on radeon 5850 on ubuntu 11.04 64bit. Box is a dedicated miner. Seems there may be some stability problems. Indicated hash rate changes dramatically. With a straight forward:

By default, poclbm looks at shares that have been generated in the last 15 minutes when calculating that estimate of the hashrate. P2Pool has a higher difficulty (currently around 19), so you'll only get a share about every four minutes, but sometimes you might not get one for longer. This leads a lot of variance there.

[fehknt]

will try out the poclbm branch...  (edit: results below)

Seems to work fine, still a significantly higher stale proportion (~.5) than median but I suppose someone has to get higher than median for it to be the median...  Set on a high target fps as recommended but haven't had time to experiment with exactly how this changes things.
cgminer 1.5.8 (the last version that I got working correctly) got ~.7, so it's an improvement.

[DeathAndTaxes]

Does any website track the hashing power of p2pool?

[forrestv]

Does any website track the hashing power of p2pool?

Every node has that data in H/s: http://forre.st:9332/rate , for example.

[jamesg]

Does any website track the hashing power of p2pool?

Every node has that data in H/s: http://forre.st:9332/rate , for example.

So this -> 9657287493 would mean that the network is running at 9.65Gh/s?

[forrestv]

So this -> 9657287493 would mean that the network is running at 9.65Gh/s?

Yep, there's a graph of the history here: http://u.forre.st/p2pool/600.png

[finway]

Does this work ?

[Eveofwar]

Does this work ?

Yes, but it doesn't put the "shit" in the blockchain.

Man, you sure do have a way to just be extremely useless to this community..

[jonathan]

Code:

$  cd ~/forrestv-p2pool-27ab63b;python run_p2pool.py user xxx
11:08:55.705643 p2pool (version 27ab63b)
11:08:55.705710 
11:08:55.705732 Testing bitcoind RPC connection to 'http://127.0.0.1:8332/' with username 'user'...
11:08:56.193601     ...success!
11:08:56.193772     Current block hash: 649e981220551b0f99208ee9ad9355b70b9b870ab703fa0a02
11:08:56.193852 
11:08:56.193933 Testing bitcoind P2P connection to '127.0.0.1:8333'...
11:08:56.417277     IP transaction denied ... falling back to sending to address.
11:08:56.483169     ...success!
11:08:56.483339     Payout script: 76a91452d75b482c2ec1c1490f02a1483b795c301ee85988ac
11:08:56.483416 
11:08:56.483491 Loading cached block headers...

...

11:08:58.376403 140000
11:09:04.915384    ...done loading 149530 cached block headers.
11:09:04.915453 
11:09:04.915673 Loading shares...
11:09:05.140885     1000

...

11:09:26.560440     87000
11:09:26.565103     ...inserting 69516 verified shares...
11:09:27.423753     ...done loading 87018 shares!
11:09:27.423832 
11:09:27.424093 Initializing work...
11:09:29.340955 Traceback (most recent call last):
11:09:29.341016   File "/usr/lib/python2.7/dist-packages/twisted/internet/defer.py", line 455, in _startRunCallbacks
11:09:29.341032     self._runCallbacks()
11:09:29.341045   File "/usr/lib/python2.7/dist-packages/twisted/internet/defer.py", line 542, in _runCallbacks
11:09:29.341058     current.result = callback(current.result, *args, **kw)
11:09:29.341071   File "/usr/lib/python2.7/dist-packages/twisted/internet/defer.py", line 1076, in gotResult
11:09:29.341083     _inlineCallbacks(r, g, deferred)
11:09:29.341095   File "/usr/lib/python2.7/dist-packages/twisted/internet/defer.py", line 1020, in _inlineCallbacks
11:09:29.341116     result = g.send(result)
11:09:29.341128 --- <exception caught here> ---
11:09:29.341140   File "/home/iktinos/forrestv-p2pool-27ab63b/p2pool/main.py", line 195, in main
11:09:29.341152     set_real_work2()
11:09:29.341163   File "/home/iktinos/forrestv-p2pool-27ab63b/p2pool/main.py", line 159, in set_real_work2
11:09:29.341175     best, desired = tracker.think(ht, current_work.value['previous_block'], time.time() - current_work2.value['clock_offset'])
11:09:29.341188   File "/home/iktinos/forrestv-p2pool-27ab63b/p2pool/data.py", line 349, in think
11:09:29.341200     desired.add((self.verified.shares[random.choice(list(self.verified.reverse_shares[last_hash]))].peer, last_last_hash))
11:09:29.341213 exceptions.KeyError: 21938799823926796951370893733906263687302461232855068765006963872814835735435L
$

Used to work, but no longer. Bitcoind is running fine.

[sd]

Has anyone got p2pool to work with merged mining?

[Eveofwar]

Why is there a 0.5% fee to the creator if this is a decentralized pool? Can't the people mining here just compile without the code that gives the fee and still make it work?

Also why so little adoption?

Proportional pools usually grow faster

[sd]

Why is there a 0.5% fee to the creator if this is a decentralized pool? Can't the people mining here just compile without the code that gives the fee and still make it work?

Also why so little adoption?

That's what I'd also like to know. A decentralized pool is an absolutely fantastic idea, but having a 0.5% fee doesn't look trustworthy. A nag message and a suggested donation just like cgminer does would be absolutely fine.

This should have taken off big time especially with the recent DDOS attacks against the big pools.

[BombaUcigasa]

Why is there a 0.5% fee to the creator if this is a decentralized pool? Can't the people mining here just compile without the code that gives the fee and still make it work?

Also why so little adoption?

That's what I'd also like to know. A decentralized pool is an absolutely fantastic idea, but having a 0.5% fee doesn't look trustworthy. A nag message and a suggested donation just like cgminer does would be absolutely fine.

This should have taken off big time especially with the recent DDOS attacks against the big pools.

Here's a cool idea. YOU pull the public project code, add new ports, remove the 0.5% fee and publish it with a different name. Can you do that?

Everyone can then mine fee free. You won't get anything for your efforts, but that's not the point, right?

It's not like the 5 BTC or so he got for his amazing work is an amazing fortune, now is it?

[sd]

Why is there a 0.5% fee to the creator if this is a decentralized pool? Can't the people mining here just compile without the code that gives the fee and still make it work?

Also why so little adoption?

That's what I'd also like to know. A decentralized pool is an absolutely fantastic idea, but having a 0.5% fee doesn't look trustworthy. A nag message and a suggested donation just like cgminer does would be absolutely fine.

This should have taken off big time especially with the recent DDOS attacks against the big pools.

Here's a cool idea. YOU pull the public project code, add new ports, remove the 0.5% fee and publish it with a different name. Can you do that?

Everyone can then mine fee free. You won't get anything for your efforts, but that's not the point, right?

It's not like the 5 BTC or so he got for his amazing work is an amazing fortune, now is it?

I don't have the spare time. Tell you what, you do it and add a non-mandatory donate option and I'll use it and I will donate.

The problem isn't that the guy wants some BTC for his excellent work, it's that the fee is mandatory not optional.

[DeathAndTaxes]

Chicken or the egg scenario.

1) Small pools have huge volatility.
2) p2pool is very very very very small pool.
3) volatility makes it difficult to grow.
4) small pool remains a small pool
5) Goto step #1

Honestly the only way I see p2pool or similar concepts taking off is if someone who has a private farm decides to point that hashing power towards it.  Someone with say 20GH/s could more than double the hashrate which would make it more attractive to others.

p2pool also currently doesn't support merged mining which as of right now is a 20% reduction in revenue.  I am interested and likely someday the bonus for merged miniing will be much lower but I can't ignore 20% free revenue right now.

[BTCurious]

p2pool also currently doesn't support merged mining which as of right now is a 20% reduction in revenue.  I am interested and likely someday the bonus for merged miniing will be much lower but I can't ignore 20% free revenue right now.

This is my main reason.
Also, there's effort involved with switching pools.

[gyverlb]

There may be some other reasons p2pool isn't more used. In my case, the roadblocks were :

• needs an alpha/beta version of bitcoind supporting a new command to work well (I had 0.3.24 and installed 0.4.1_rc1 then 0.5.0 to make it happy) <- the good version should really be mentionned on the OP first post
• this version (0.5.0) of bitcoind crashed many times during my last tests (bitcoind 0.3.24 crashed too previously so maybe p2pool should implement a fallback using other p2pool peers)
• p2pool needs large amount of memory (300MB seems the minimum)

This last point made me abort my last try at running p2pool (the OOM-killer on my 512MB VPS killed it today).

[btc_artist]

This sounds like a *really* good idea, now we just need to think of how to get it off its feet.

[Brian DeLoach]

Honestly the only way I see p2pool or similar concepts taking off is if someone who has a private farm decides to point that hashing power towards it. Someone with say 20GH/s could more than double the hashrate which would make it more attractive to others.

I'm going to point all my miners (20 gh/s) to this pool this weekend.

I love the concept and I'll throw all my weight and support behind this pool to help it get going. If we can get a couple more big time miners (anyone know of any?) to mine for this pool we can hit 100+ GH/s and this pool can really takeoff!  

[jamesg]

Honestly the only way I see p2pool or similar concepts taking off is if someone who has a private farm decides to point that hashing power towards it. Someone with say 20GH/s could more than double the hashrate which would make it more attractive to others.

I'm going to point all my miners (20 gh/s) to this pool this weekend.

I love the concept and I'll throw all my weight and support behind this pool to help it get going. If we can get a couple more big time miners (anyone know of any?) to mine for this pool we can hit 100+ GH/s and this pool can really takeoff!  

I still think some more development needs to be done to make the setup and use simpler. I also don't see a reason for the .5% fee if development work is not continuing.

What needs to happen to get this dev work done. A bounty maybe?