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

[phillipsjk]

Noticed, perhaps, if you are watching the pot boil. If you expect to make 50c/day and you make 0c/day, that looks like a big deal, but in reality your actual shortfall is still only 50c. Not going to hurt anybody.

Someone with a somewhat larger rig who expects to make $1000/day and makes $0/day (certainly possible when p2pool has 4+ day rounds), that's a $1000 shortfall, and may or may not be a big deal, depending on your financial situation.

Low hash-rate miners get more dry spells because they may not have found a share within 3 days prior the pool finding a block. I was PO'd when my first two shares had a zero pay-out for that reason.

[1713578883]

1713578883

[1713578883]

1713578883

[1713578883]

1713578883

[1713578883]

1713578883

[smooth]

Again, if the pool has a dry spell you are only losing payout on one share, which is really not that much in the way of coin. Bigger miners are losing out on many shares (if the p2pool doesn't get a block in three days). They have more to lose.

But if you really can't take the variance but still want to help p2pool, split off some of your hash rate to one or more of the bigger pools where you get (tiny, tiny) earnings on every few blocks. You'll get some small steady earnings with a nice boost when p2pool has enough luck to generate payouts.

[MrWong]

p2pool.org down for anyone else? Miners seem to roll over to back up pools. Anyone have any details on whats going on?

[roy7]

p2pool.org down for anyone else? Miners seem to roll over to back up pools. Anyone have any details on whats going on?

You might try one of Jude's threads or PM him. This is the general discussion thread for the p2pool software itself. p2pool.org is just a public node someone runs, not an official site of the p2pool developers.

[cr1776]

p2pool.org down for anyone else? Miners seem to roll over to back up pools. Anyone have any details on whats going on?

P2pool itself is up, you can set up other p2pool nodes as backups.  As roy7 says, for p2pool.org, you'll have to check with jude

[smartbart24]

The P2Pool concept seems revolutionary to me so I would really like to understand it in full.

I was trying to review the source code of the P2Pool here: https://github.com/forrestv/p2pool but have slight problems "finding" into the code.
This link brought some light : https://en.bitcoin.it/wiki/P2Pool_code_documentation

Still maybe you could jump start me in two questions:
1) How does the client know which work to do (it is passed to bitcoind via get getblocktemplate right ? - but I still have to know what work has been done by the other shares before)
2) How do I append my payout adress to the script and how is ensured that script is included in the final block ?

Maybe you could hint me to some code sections ?

Thank you in advance for your help

[KyrosKrane]

p2pool.org down for anyone else? Miners seem to roll over to back up pools. Anyone have any details on whats going on?

P2pool itself is up, you can set up other p2pool nodes as backups.  As roy7 says, for p2pool.org, you'll have to check with jude

At risk of tooting my own horn, my current P2pool node (in Florida) is in my sig, and I'm bringing up another pool in Germany this weekend (I hope!).

[roy7]

Just got hit with the number of files issue myself.

2014-03-13 17:36:23.039910 > Failure: twisted.internet.error.ConnectBindError: Couldn't bind: 24: Too many open files.

I google'd how to increase the #, and checked this:

cat /proc/sys/fs/file-max
165417

But honestly that should be well more than enough? It is possible the file handles to the coin daemon aren't being closed properly? So each RPC call it slowly counts up towards the maximum?

Didn't know to try this before restarting p2pool, but

cat /proc/sys/fs/file-nr
672     0       165417

only 672 files in use after it was restarted, out of the 165K maximum.

ulimit -Hn
4096
ulimit -Sn
1024

[roy7]

I use the following in my setup script for ubuntu

Code:

# Increase open file limits
sudo sh -c 'echo "* soft nofile 16384" >> /etc/security/limits.conf'
sudo sh -c 'echo "* hard nofile 65536" >> /etc/security/limits.conf' 
sudo sh -c 'echo "* soft nproc 4096" >> /etc/security/limits.conf' 
sudo sh -c 'echo "* hard nproc 16384" >> /etc/security/limits.conf' 

for Fedora it's  /etc/security/limits.d/90-nproc.conf

Thank you for this!

[roy7]

I probably ran into the soft ulimit. Looking at my other node:

$ lsof -p 9436 | wc -l
228
$ lsof -p 9436 | grep 9171 | wc -l
142

Node has 22 incoming and 10 outgoing peer connections. There are 7 active miners. However there are 142 active file descriptors on the mining port number (vertcoin is 9171).

It appears to be all of the hostnames of the miners who have ever connected to the pool since the last restart. Are those file descriptors never closed?

Looking more closely at the actual list, there are many duplicates to the same actual miners, just on a variety of ports. Such as this guy (his IP address removed):

$ lsof -p 9436 | grep 9171 | grep xxxxxxxxxxx.dynamic.upc.nl | wc -l
29

29 file handles all for one miner.

[organofcorti]

However variance as a fraction of expectation is much greater when hashrate is lower, which I think is what roy7 meant, and what is generally noticed by most miners.

Noticed, perhaps, if you are watching the pot boil. If you expect to make 50c/day and you make 0c/day, that looks like a big deal, but in reality your actual shortfall is still only 50c. Not going to hurt anybody.

Someone with a somewhat larger rig who expects to make $1000/day and makes $0/day (certainly possible when p2pool has 4+ day rounds), that's a $1000 shortfall, and may or may not be a big deal, depending on your financial situation.

No. It's not 'noticed because someone is looking for it' (which is what I think you meant with your analogy?). It is noticed because the 95% confidence interval for income as a fraction of expected income is much wider for lower hashrates than for higher hashrates.

[irrational]

Thinking 22+ GH/s should be sufficient for p2pool :-)

At 22 GH you will still have dry spells on p2pool. As I said, though, don't worry about it, you'll have big payoff days too. Look at it every week or two and you'll be happy with the results.
 

No worries 

From crunching the numbers, I think I'd feel ok with the variance at 22 GH/s at current pool difficulty. It's just a hobby anyway, and it'll all even out in the end.

[smooth]

No. It's not 'noticed because someone is looking for it' (which is what I think you meant with your analogy?). It is noticed because the 95% confidence interval for income as a fraction of expected income is much wider for lower hashrates than for higher hashrates.

Yes, as a fraction of expected income but ask yourself why that even matters. If you expect to get 0.003 BTC and you get 0.002 BTC (or even in one instance 0 BTC for that matter) how is your life impacted in some major way? People can stare at numbers on a screen or lines on a chart all day long but at some point you have to relate those numbers and lines to reality, and when you do that shortfalls on hobby mining do not really matter.

[roy7]

Yes, as a fraction of expected income but ask yourself why that even matters. If you expect to get 0.003 BTC and you get 0.002 BTC (or even in one instance 0 BTC for that matter) how is your life impacted in some major way? People can stare at numbers on a screen or lines on a chart all day long but at some point you have to relate those numbers and lines to reality, and when you do that shortfalls on hobby mining do not really matter.

You've never noticed that people using public p2pool nodes are far more likely to complain that they are getting no payouts, than that the payouts they are getting a smaller than they expected? You can argue single GPU miners shouldn't care much because their earnings are so small in absolute terms. However, they seem to care quite a bit and check their stats often. They are also very noisy when they aren't finding a share often enough to get paid on every block. Usually that degrades into saying p2pool rips off small miners since they don't understand how it'll average out long-term. Larger miners who also have variance but get paid on every block don't raise as much fuss.

[smooth]

You've never noticed that people using public p2pool nodes are far more likely to complain that they are getting no payouts, than that the payouts they are getting a smaller than they expected?

Yes I've noticed a lot of complaints. I'm suggesting to those people to chill out and have fun with it, not get so upset about things that don't really matter.

I also see a lot of theorizing about how "bad" the variance is for small miners from people I suspect are not even themselves small hobby miners. They are either not miners at all or are larger miners who are making unnecessary noise saying things that don't make real sense when you dig into the practical reality of it. They should just stop.

Don't get me wrong, if there were some way to flip a switch tomorrow and reduce variance for smaller miners I would suggest we do it, but there isn't. Mining is an exercise in tradeoffs. For hobby miners I suggest making tradeoffs in favor of having fun and helping the network rather than wanting to get a "steady paycheck" of a few dollars a day.

[jedimstr]

You've never noticed that people using public p2pool nodes are far more likely to complain that they are getting no payouts, than that the payouts they are getting a smaller than they expected?

Yes I've noticed a lot of complaints. I'm suggesting to those people to chill out and have fun with it, not get so upset about things that don't really matter.

I also see a lot of theorizing about how "bad" the variance is for small miners from people I suspect are not even themselves small hobby miners. They are either not miners at all or are larger miners who are making unnecessary noise saying things that don't make real sense when you dig into the practical reality of it. They should just stop.

Don't get me wrong, if there were some way to flip a switch tomorrow and reduce variance for smaller miners I would suggest we do it, but there isn't. Mining is an exercise in tradeoffs. For hobby miners I suggest making tradeoffs in favor of having fun and helping the network rather than wanting to get a "steady paycheck" of a few dollars a day.

Oh, but you're forgetting... There is some way to flip a switch. Not quite tomorrow, but hopefully soon enough....

Anybody have any ideas how to increase the hash rate? We are sinking to a smaller and smaller percentage of the network, and round times are growing. We're having 4 day rounds now, a few more difficulty jumps and it may be weeks or longer.

Something must be done.

I'm working on a revamp of P2Pool that will give us a few key features that will help:

• Running a "lite" P2Pool instance - one without a full Bitcoin node - will be possible without losing any of the promises P2Pool makes about its contribution to Bitcoin network health
• P2Pool will have a non-linear sharechain so that we can have shares more often than 30 seconds, allowing small miners to use P2Pool

Both of these make P2Pool easier to use; the eventual goal is being able to run P2Pool on the same hardware miners use, such as a Raspberry Pi, though that would likely require a laborious rewrite in C++ or another compiled language.

My current progress on this: I have a piece of software talking to bitcoind and augmenting blocks with UTXO merkle branches, feeding that to another piece of lightweight software that can trustlessly verify the blockchain, while holding O(1) state. I have a good idea about how the non-linear sharechain will work, but implementing that hasn't started yet.

[roy7]

Don't get me wrong, if there were some way to flip a switch tomorrow and reduce variance for smaller miners I would suggest we do it, but there isn't.

The only things I can think of are the patches I've already put in pull requests for, and for larger miners to take higher share diff targets so the share diff network wide would go lower. I wrote some thoughts on that a while back but got no replies. Maybe the revamp forrestv is working on will solve it.

[roy7]

BTW it's kinda a bummer the max share chain is only 3 days. The SPREAD is 3 blocks worth of work, so that would normally mean any share someone finds should get one or more payouts. Is the reason we have a max share chain length at all to prevent abuse? I'd think the share chain should expand as much as needed automatically to store SPREAD worth of payouts...

[organofcorti]

No. It's not 'noticed because someone is looking for it' (which is what I think you meant with your analogy?). It is noticed because the 95% confidence interval for income as a fraction of expected income is much wider for lower hashrates than for higher hashrates.

Yes, as a fraction of expected income but ask yourself why that even matters. If you expect to get 0.003 BTC and you get 0.002 BTC (or even in one instance 0 BTC for that matter) how is your life impacted in some major way?

Of course it matters. It matters to people who are hobby miners who try to optimise their equipment. It's hard to optmise if you only see a share every few days.

You're also assuming everyone is a hobby miner with no concern for profit. Just because someone can only afford a low hashrate miner doesn't mean they can treat the investment as a "hobby cost". Lots of countries are poorer than the US.

[smooth]

No. It's not 'noticed because someone is looking for it' (which is what I think you meant with your analogy?). It is noticed because the 95% confidence interval for income as a fraction of expected income is much wider for lower hashrates than for higher hashrates.

Yes, as a fraction of expected income but ask yourself why that even matters. If you expect to get 0.003 BTC and you get 0.002 BTC (or even in one instance 0 BTC for that matter) how is your life impacted in some major way?

Of course it matters. It matters to people who are hobby miners who try to optimise their equipment. It's hard to optmise if you only see a share every few days.

You're also assuming everyone is a hobby miner with no concern for profit. Just because someone can only afford a low hashrate miner doesn't mean they can treat the investment as a "hobby cost". Lots of countries are poorer than the US.

Daily variance does not reduce your profit. People are constantly on here talking about how pools with significant fees are "better" for small miners. That is false. In fact they reduce profit.

Even in the case of looking at mining as a "paycheck" most people getting actual paychecks get paid once a week or every two weeks. Looking at your earnings every single day and expecting them to be some nice flat number every single day is stupid and counterproductive.