Topic: [1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool - page 483. (Read 2591928 times)

It is when you open web page of your node when it is totally fresh and have no data to display in graphs.

Thanks, that helps.  I'd rather see actual earnings over the time period, it looks like you're extrapolating.  But it still helps give me an idea of what the difference would be.  I might actually try another pool at some point.  I've been on eligius and btcguild in the past, but not eclipse.  I liked the pools just fine, I just like the idea of p2pool.  Anyway, I gotta stay on p2pool at least until I finish my latest changes to the android widget I'm working on.

Any ideas what might cause the below?  I am git-pulled to the latest code, and I get this pretty much straight away.

This is on debian unstable, with python 2.7:

Agreed.  It's easy enough to set up, once you've spent 24 hours figuring it out.  If there was some sort of installer, or if p2pool was just an executable that didn't need all the faffing, I think a lot more folks would run it.

On the Windows end of things, all it needs is an installer that adds the bitcoin.conf files, asks a few questions, opens a few ports, job done.  But, it seems linux is the main focus of bitcoin software authors, despite a huge proportion of people running Windows.  It all smacks of elitism.   But, hey, I'm just an IT administrator from the world of Windows, who started out in the days of NT3.51, what would I know?
 

With various faffing about, my average over the past 24 hours has been 31GH/s (one of my Blades was a bit sick, needed more power to stabilise it), I got BTC0.167 on EclipseMC.  I've only just moved there to see how it goes.  Eligius was paying out around 25 hours, and I think they pay out around 0.167 too.

Taking 23 full days in September so far, that's about BTC3.841.   Ignoring variance, of course.

Edit: Actual figures according to EMC are 0.16704727620043 / 24 hours

It's all in the setup my man. I gave up using high powered & expensive gear on p2pool, it's simply not worth it due to the limit of mining gear that actually works with it, unless you got an ASIC & are knowledgeable & bothered enough to faff around with it to make it work. I opted for energy efficiency over performance with p2pool & it's now worth while actually using it. I use a Sempron 145 with unlocked core (making it a 45w dual core Athlon II X2 OC'd @ 3.4!!) with 8Gig 1600 RAM, a separate SSD for the 6 merged mined coins data & a Gold 850W PSU running Xubuntu 64bit 12.04. My miner is an old Acer One Netbook (Xununtu) running 40 usb's (it gives the lowest reject/hw error rate of all my equipment: <1%) All this goes through a UPS that shows a measly 225 watts draw. That's less than the average desktop PC. I also gave up faffing around with the settings, most of the suggestions in g's guide actually made things worse anyway (after all, it's a guide, not a bible), but everybody's setup is different of course - but my rig likes things standard.

All this talk of having to be some sort of geeky administrator to get p2pool to run right makes me laugh - if a dope-head like me can do it, anyone can. I struggle with compiling stuff FFS   A geek I am not, I chuff.

P2pool will never be a "real" pool until the compatibility issues are sorted out. When noobs can come along, plug in & play, the pool will grow to be what it deserves to be, like every other pool has. Until then, for me anyway, it's a toy to play with - a learning process - but a fun one. it might even payout once or twice. If you're lucky   .

12/1Mb is plenty. I have nearly the same: ~14.5/1.4Mb. I need to use QoS settings giving higher priority to the *coind and P2Pool traffic and I tuned them according to my guide.

Out of curiosity, what are your returns like?  I have 30GH on p2pool, you have 35GH on a larger pool, so our results should be close.  I've been on p2pool for a while now and I'm wondering if you're making enough more than me for it to be worth switching.  So far in the month of September I've made 3.567564 BTC.  If you're within 10% of that I'm staying because I like the idea of p2pool. If you need a different date range to compare let me know.

Edit:  10% after accounting for your extra 5GH of course.

I'm not sure my internet is good enough to be hosting a p2pool node, either.  I'm just on 12/1Mb ADSL.  It's not the quickest, and probably quite latent when other things happen, despite QoS.

That's not so bad. People leaving are those with the worst configurations (network issues or underpowered node), who don't have the sysadmin skills to maintain a node or who can't stand the variance (due to very low hashrates or luck misunderstanding).

Regular users of p2pool are those who know how to set it up, maintain and optimize it: they build a high quality network of nodes that should outperform any centralized pool. With lots of well configured nodes broadcasting our blocks, we should have lower orphan rate. As anyone verified/quantified this by the way?

Last time I studied the orphan rate for solo miners on alt-coins, I found a block collision rate that was equivalent to a 3 to 5 seconds window where a block would be orphaned. This should amount to a ~0.5% orphan rate on the bitcoin P2P network. This is over-simplified by assuming most pools use a single bitcoind to broadcast their block (some have severals or have direct connections to other "friend" pools).

Is there some archive of P2Pool found blocks somewhere? Looking at the last thousand and comparing to the largest pools we should have an idea of our performance.

Kinda a catch 22 there... More hash rate == Quicker Block solves == lower variance. (PPLNS payout) p2pool has less total hash rate so therefore it will have higher variance which will lead to fewer people using it which leads to lower total hash rate....

It was running 13.3.  I'll maybe have another go with it sometime, perhaps there was something strange with my install, or I missed something.  I'm pretty sure I did everything I should have, though.

I'm rejiggering my workstation in the next few days, including a fresh install, so I'll see if I can have another go.

Still, all that said, I'm still getting far better returns from a larger pool than I did on p2pool.

For tuning probably: at least I'm not aware of any other verified information that would be relevant.
For the basic installation this thread's first post and the wiki page should have you covered (it assumes basic sysadmin skills).

I will give it a try. Is ur guide the most uptodate?

When was that? Before or after the last protocol fork (which made it efficient with most ASICs) which happened in July? If it was after the fork it's surprising as it should have solved these problems.

BTW, your hashrate shouldn't have any impact on the resources used by a P2Pool node:

• Stratum offloads most of the work generation to the miner, leaving on average one computation to be done only when the node needs to update the coinbase (probably every 30 seconds on average unless it is aggressively pushing additional TX in the coinbase as they come, which would not be linked to the miners hashrate)
• variable difficulty is used so that the node only receives one share per second on average to verify

I use an i5-3350P with 16GB and a 64GB SSD with ~20 altcoins nodes, 6 p2pool nodes, a PostgreSQL database for some datamining and several other daemons with an average load at ~4.
My efficiency is most of the time >102% for the Bitcoin p2pool node with an Avalon and several FPGAs mining on it.

Before that I only had an old Core2Duo at 3GHz with 6GB and the same SSD which was more than enough. I upgraded due to flaky hardware (integrated SATA controller dying) and the additional RAM allowed me to follow more altcoins.

Edit: forgot to add CPU usage stats, on average the Bitcoin p2pool node uses 5.2% of one of my CPU cores, bitcoind uses 2.2%.

Still resource hungry, still high variance.

I was really keen to use P2Pool, and ran it for several months, but to be honest it was just too much - it needed a fast machine to run it, and payouts were pretty woeful. Half of any mining profits (if I got any at all) was wasted on power to run the i3 machine needed to make P2Pool perform anywhere near well enough.  I tried to run it on a Sempron dual core 2.5GHz CPU, and it struggled with a measly 5GH/s.  Someone said they ran P2Pool on a Celeron 847, but I think they were huffing glue.

Now I'm up to 35GH/s, I gave up on P2Pool.  I get far better returns with less bother on a 'real' pool.  

why wouldnt it?

it has 2GB RAM and a 500GB HDD

obviously you wouldnt want to make it public and have a ton of people connecting to it

i guess the only thing ive noticed is that with an i7-4930k you can get the getblocktemplate latency down to 0.7ms or so (though i guess this would be more reliant on having quality memory, eh).    if you chop out all transactions it'll go for 0.85 or 0.9ms on one of those old hetzner 2700k's

I havent kept up with P2pool for so long. Way back when i tried it with GPUs more than a year ago, its very resource hungry and variances were so high.

How is it now with all the ASICs?

What is the current status of the project? are there any forks from this project?

Why not set up a separate proxy for the Blades and put a +1 on the username?

Just to note that I am connecting my Blades to p2pool via BFGMiner's http stratum proxy.  I don't know how I can set the diff to 1 by device - I'd have to keep it diff1 for everything that's connecting through BFG.