Bitcoin Forum

Is there something inherently wrong with p2pool's infrastructure that should cause it to be 1% of the network hash-rate and falling at present?

Bad luck aside, which it has been recently and would explain some of the drop-off, it has never garnered a very large following.  Why?

I would expect that a p2p currency might be more supportive of a p2p pool.  Is there something faulty in my thinking that that should be true?

Because it's outside the norm and appears to be for techno-nerds only, i.e. not user-friendly.

Setting up a node is more trouble, but using a public node is no more difficult than using any other pool.  Point box at address and port, and insert public key.  What am I missing?

+1 I am curious to find the answer as well. For one I find it pretty hard to find the proper networks.py settings for altcoins. It seems to me every fork maintains his own networks.py config.

2 reasons: effort and FUD

You don't have to set up a node!

Just point your miner at a node IP address and add you bitcoin address!

The reason it's small is because it's new and it pays out per block which means you need a decent amount of hashing power to get above the dust threshold to see payments arrive. P2Pool recommends 40GH for Bitcoin, but I reckon 15-20GH should work.

One of the common arguments against using pie charts is that growth in one item makes it look like the others are shrinking.

P2Pool is somewhat lower than its all time peak hashrate, as your note— a lot of miners fall prey to gamblers fallacy and are only fair weather miners who mine only if the pool has been unusually lucky, even though there is no rational impact... but most of the percentage change has been due to large hashrate increases elsewhere...  And why is P2Pool not getting more of that increase— well, it lacks things like payment in yaun instead of Bitcoin. :)

P2Pool also works differently than any other pool. Such as:

The p2pool sharechain (that stores the PPLNS payment data) has a target speed of 30 seconds per block, which causes a lot of restart message spam to the miner.

Finding a share for a smaller miner can take a while, some miners believe this means they are getting "ripped off" because they don't understand variance.

Orphans/DOAs on the sharechain are common because of the fast target speed, some miners misunderstand this to be similar to orphaned blocks on the blockchain. As long as the orphan rate of shares between p2pool nodes is similar, then it isn't impacting miner payouts at all. (This is reflected in p2pool's Efficiency stat in the default interfaces.)

P2Pool is quite cool, as an experiment and to learn more about it I'm running some nodes on a new alt coin. I think the issues above can mostly be resolved with better miner education. Some way to reduce variance for smaller miners would be nice though.

True, but it has also dropped in absolute terms, not just relative terms.  Some weeks ago it had reached 200 TH/s and recently has fallen as low as almost 130 TH/s at the same time as the total network hashrate has increased.

Unsure about the yuan reference.  Which mining pools pay in yuan?  There are Chinese-based public nodes on the p2pool network.  I guess they like BTC?

I found that using p2pool with a BFL single was taking over 24 hours for a single share.  And the $amount per day/week was dropping like a stone.

Meanwhile, I could use the same electricity to mine more profitable altcoins.

Probably not the only one...


Also I was trying to help someone setup p2pool for mining ronpaulcoin (scrypt) on Windows 7 (64 bit).   It was near impossible, requirng the end-user to build a python C module (ltc_scrypt) except that it doesn't actually build correctly on 64 bit windows.  So sorry, SOL.

Not to mention there is really no main p2pool webpage for learning about it.  And the author doesn't seem at all proactive about adding new coins or maintaining it.   So lots of forks.

It seems like abandonware to me.   I hope the author or someone else picks it up and starts making it a lot more newbie friendly and also maybe think about ways to improve the share time for bitcoin miners because it is kinda sad when 60Ghz/sec cannot even get you 1 share per day.

my 2 cents

^ This. Pretty much sums it all up what the current issues are with p2pool. Looks like the masses is not interested in p2pool.

A what now? 

I ran it for a while but I went back to BTCGuild. 

You have to run a full blown bitcoind node.  Latences between components (bitcoind, p2pool, internet, miner) have to reduced to a bare minimum.  High latences lead to large numbers of orphans which in turn kills your profits.  Other more efficient nodes will win more shares than you.  I liked the idea of it, but I could not get my efficiences in order to make it make sense.  So I took my 580Ghs back to BTCGuild.

Thank you for your post roy7!

Added the following to the FAQ section on my nodes:
--------------
Q: How is P2Pool different?
A:The P2Pool sharechain is different than the coin blockchain. It has a target speed of 30 seconds per block (Bitcoin) which causes a lot of restart messages to the miner.

Smaller miners may take some time to find a share. This will result in an increase of variance but over time you should experience higher payouts when compared to other pools.

Orphans/DOAs on the P2Pool sharechain are common because of the fast target speed. Do not confuse this with orphaned blocks on the coin blockchain. Refrence the efficiency stat of your P2Pool node.
--------------
Hope you don't mind that I reworded it some :)

Not at all. :)