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

[mdude77]

I saw this on wiki:

Requirement:
    Windows:
        Install Python 2.7: http://www.python.org/getit/
        Install Twisted: http://twistedmatrix.com/trac/wiki/Downloads
        Install Zope.Interface: http://pypi.python.org/pypi/zope.interface/3.8.0
            Unzip the files into C:\Python27\Lib\site-packages

Well I have not installed the Python 2.7, Twisted and Zope.  Will that be a problem resulting in my Cointerra's under performance???

No.  If p2pool is running, you have the right requirements OR you're using the executable which doesn't need that.

I'm not using my S2 on p2pool because of underperformance.  I tried for a few days throwing away 10% of my hash power and 5% rejects, and it was abysmal. 

M

[phelix]

Thanks JB, so it's not just me then. I'll investigate further........

Edit: I notice forrestv made a change to the p2poool repo ( !) 7 days ago, increased maximum worker difficulty 1000x, that wouldn't be the cause......would it?

Edit2: Tried running namecoin on rav3n_pl fork with the same results, so it's not a p2pool issue. Had to close the namecoin wallet as it was eating cpu as well. Oh well, that'll have to stay off then.......... Very strange.

There have been some huge blocks recently aggregating a mass of tiny inputs. This seems to lead to getauxblock taking a very long time. We are looking into it but it might take a while.

BTW: As far as I know Namecoin mining reward variance is very high with P2P. We would be happy to support any plans to improve this, maybe with a bounty.

[PatMan]

There have been some huge blocks recently aggregating a mass of tiny inputs. This seems to lead to getauxblock taking a very long time. We are looking into it but it might take a while.

BTW: As far as I know Namecoin mining reward variance is very high with P2P. We would be happy to support any plans to improve this, maybe with a bounty.

Thanks for that phelix - I thought I was going a little mad there. I'm still not sure how large block size can cause a timeout error like this though, tbh - what am I missing?  

Peace.

EDIT: Scrap that - I get it now. (I was too lazy to read up on it before  )

[jonnybravo0311]

Thanks JB, so it's not just me then. I'll investigate further........

Edit: I notice forrestv made a change to the p2poool repo ( !) 7 days ago, increased maximum worker difficulty 1000x, that wouldn't be the cause......would it?

Edit2: Tried running namecoin on rav3n_pl fork with the same results, so it's not a p2pool issue. Had to close the namecoin wallet as it was eating cpu as well. Oh well, that'll have to stay off then.......... Very strange.

There have been some huge blocks recently aggregating a mass of tiny inputs. This seems to lead to getauxblock taking a very long time. We are looking into it but it might take a while.

BTW: As far as I know Namecoin mining reward variance is very high with P2P. We would be happy to support any plans to improve this, maybe with a bounty.

The reason the variance is so high is because each node that is merge-mining is effectively trying to solo mine the blocks for the coins (NMC, DVC, etc).

It would be nice if the entire hashing power of the pool, or at least the hashing power of the nodes that are merge-mining, could be applied rather than individual nodes attempting to solve the blocks.

[bitcoinbearhk]

OK so I have given up using my Terraminer IV for p2pool ... it's just 20% less effective due to unknown reason.

I have pointed my Antminer S3s to my node, no problem, hashing good... Perhaps I will just stick with S3 with p2pool, and Cointerra with traditional pools.


Now I wish to setup merge mining.  Can someone please point me to any guide for Namecoin merge mining setup for Windows 7???  Thanks.

[Duce]

OK so I have given up using my Terraminer IV for p2pool ... it's just 20% less effective due to unknown reason.

I have pointed my Antminer S3s to my node, no problem, hashing good... Perhaps I will just stick with S3 with p2pool, and Cointerra with traditional pools.


Now I wish to setup merge mining.  Can someone please point me to any guide for Namecoin merge mining setup for Windows 7???  Thanks.

Please look at the previous post.

[wlz2011]

Eligius  9.5PH/S

P2pool  2.40PH/S

Very nice

[bryonp]

Looks like 3 blocks on the 19th so far?? my Node is only showing the first one.
What would cause it to not show me the last 2 blocks?
I am on line, and hashing away, I have active connections?

Thanks for any answers.....

Bryon

[windpath]

Looks like 3 blocks on the 19th so far?? my Node is only showing the first one.
What would cause it to not show me the last 2 blocks?
I am on line, and hashing away, I have active connections?

Thanks for any answers.....

Bryon

Stale p2pool shares that meet bitcoin difficulty are still submitted and will not show up in p2pool's standard interface.

You can see all p2pool found blocks here: http://minefast.coincadence.com/p2pool-stats.php

[bryonp]

Looks like 3 blocks on the 19th so far?? my Node is only showing the first one.
What would cause it to not show me the last 2 blocks?
I am on line, and hashing away, I have active connections?

Thanks for any answers.....

Bryon

Stale p2pool shares that meet bitcoin difficulty are still submitted and will not show up in p2pool's standard interface.

You can see all p2pool found blocks here: http://minefast.coincadence.com/p2pool-stats.php

Thanks as always, I kinda knew that but just thought it was more of an issue with me and not everyone like that.....
Thanks. You always start thinking that your own node is screwing up.........

[jonnybravo0311]

OK so I have given up using my Terraminer IV for p2pool ... it's just 20% less effective due to unknown reason.

I have pointed my Antminer S3s to my node, no problem, hashing good... Perhaps I will just stick with S3 with p2pool, and Cointerra with traditional pools.


Now I wish to setup merge mining.  Can someone please point me to any guide for Namecoin merge mining setup for Windows 7???  Thanks.

Sorry to hear that your Terraminer doesn't work with p2pool and suffers the same kind of performance degradation as the S2 does.  I know it's a long shot, but you might want to open a support ticket with them (I KNOW it's a long shot... lol).

Setting up merged-mining is pretty easy.  First, download and install the wallet (or you can use the *coind) of a supported coin (NMC, IXC, I0C, DVC, HUC, FSC... any I missed?) and get it all synced up.  Second, edit your coin's configuration file.  For example, here's my NMC one (changed the user and password):

Code:

server=1
daemon=1
listen=1
rpcuser=SOMEUSER
rpcpassword=SOMEPASSWORD
rpcport=7333
rpcallowip=10.0.1.*

Please use different values for user and password - these are just examples.  Also, for the rpcallowip, use a range that makes sense for you.  My local network is 10.0.1.*.  Yours might be 192.168.1.*.  Of course, you could always just put in * if you don't want to restrict it.

Now, when you fire up your p2pool node you just add the values.  Here's the snippet from my startup showing NMC:

Code:

./run_p2pool.py --merged http://SOMEUSER:SOMEPASSWORD@10.0.1.14:7333 ...

10.0.1.14 is the IP address of the machine where you put the NMC wallet.

That's it... you're merge mining BTC and NMC.

[kgb2mining]

Just wanted to chime back in on the Cointerra discussion, I've been a bit busy since yesterday.

I can confirm in my tests that the Cointerras do indeed hash at about a 20% loss on p2pool.  I switched my miners back over to BTCGuild to test it, and bingo, I went from 1.2-1.3 TH/s on p2pool to the full 1.6TH/s on BTCGuild.  So, right now the Cointerras have been left back on BTCGuild to mine there.

I can also confirm that fiddling with the difficulty and pseudo-share settings as I did made no improvement, and in some cases actually degraded performance.  I tried ranges from /512 through /2048 with combinations of +512 through +2048 as well.

FYI, I'm not going to do any more testing as I've just put the Cointerras up for sale.  Going to stick with my S1/S3's since they are running beautifully and I want to stick with P2Pool.

[bitcoinbearhk]

Cool thanks guys.

Is 1.7% DOA a reasonable number?  wondering if I should take step to improve it or just keep it as is.

[bryonp]

OK so I have given up using my Terraminer IV for p2pool ... it's just 20% less effective due to unknown reason.

I have pointed my Antminer S3s to my node, no problem, hashing good... Perhaps I will just stick with S3 with p2pool, and Cointerra with traditional pools.


Now I wish to setup merge mining.  Can someone please point me to any guide for Namecoin merge mining setup for Windows 7???  Thanks.

Sorry to hear that your Terraminer doesn't work with p2pool and suffers the same kind of performance degradation as the S2 does.  I know it's a long shot, but you might want to open a support ticket with them (I KNOW it's a long shot... lol).

Setting up merged-mining is pretty easy.  First, download and install the wallet (or you can use the *coind) of a supported coin (NMC, IXC, I0C, DVC, HUC, FSC... any I missed?) and get it all synced up.  Second, edit your coin's configuration file.  For example, here's my NMC one (changed the user and password):

Code:

server=1
daemon=1
listen=1
rpcuser=SOMEUSER
rpcpassword=SOMEPASSWORD
rpcport=7333
rpcallowip=10.0.1.*

Please use different values for user and password - these are just examples.  Also, for the rpcallowip, use a range that makes sense for you.  My local network is 10.0.1.*.  Yours might be 192.168.1.*.  Of course, you could always just put in * if you don't want to restrict it.

Now, when you fire up your p2pool node you just add the values.  Here's the snippet from my startup showing NMC:

Code:

./run_p2pool.py --merged http://SOMEUSER:SOMEPASSWORD@10.0.1.14:7333 ...

10.0.1.14 is the IP address of the machine where you put the NMC wallet.

That's it... you're merge mining BTC and NMC.

Can anyone suggest what is a good additional coin (profitable) to merge with p2pool???
I would assume, that as you are working on BTC you can make additional funds?

[kgb2mining]

Cool thanks guys.

Is 1.7% DOA a reasonable number?  wondering if I should take step to improve it or just keep it as is.

1.7% is very reasonable, and downright good.  I average between 2-4% DOA on my node.  The global DOA rate is usually in the low-to-mid teens.

[PatMan]

Can anyone suggest what is a good additional coin (profitable) to merge with p2pool???
I would assume, that as you are working on BTC you can make additional funds?

What's profitable today, might not be next week - so I mine them all, just in case......

BTC, DVC, IXC, I0C, GRP, HUC, FSC &, until yesterday NMC - which happens to be the most lucrative atm.......

[windpath]

Thanks guys,

I updated the p2pool wiki and added a list of hardware with confirmed issues:

https://en.bitcoin.it/wiki/P2Pool#Interoperability_table

[jonnybravo0311]

Thanks guys,

I updated the p2pool wiki and added a list of hardware with confirmed issues:

https://en.bitcoin.it/wiki/P2Pool#Interoperability_table

I swear there were reports of the KNC Neptune having issues with p2pool as well...

Also, you can add the SP10 as hardware the works properly with p2pool.  I'm not sure of the SP30 because I don't own one

[windpath]

Thanks guys,

I updated the p2pool wiki and added a list of hardware with confirmed issues:

https://en.bitcoin.it/wiki/P2Pool#Interoperability_table

I swear there were reports of the KNC Neptune having issues with p2pool as well...

Also, you can add the SP10 as hardware the works properly with p2pool.  I'm not sure of the SP30 because I don't own one

Thanks, added SP10, will ask in the spondoolies thread about 30.

[kgb2mining]

Ok, so now some questions about peering, in general.

Should you specify your own list of peers?
If so, how many should you consider putting in your config?  Can there ever be too many?
How do you know the performance of your peers, now and historically, to make sure it's good?
Would it help the community to establish a set of well-resourced hosts to act as "root peers", in the same fashion as root nameservers?

So, here's the explanation of my thinking.

If I understand it correctly, the key here is how fast the share information enters the p2p chain, right?  Whomever wins the "share war" gets their place in the chain and their reward, while the ones that get beat out get an orphan.

That being the case, logic should dictate that if your node is well resourced (CPU, RAM, etc), and you have good bandwith, you should be able to get your info to and fro quickly into the overall network, and thusly win that war more consistently.  If you're a small node or not connected well, you'd lose.  That to me seems yet another thing in favor of the larger miners and operations which could be another barrier to adoption and use.

The question then becomes - who are you talking to in the chain, and how fast are they spreading the info to the other nodes for you?

For the larger nodes running open mining for anyone to connect to, this probably isn't a huge concern.  But for those running their own node at home, or us behind firewalls where we can't open up, who you're peering with may indeed have an affect on their stales/DOAs/Orphans, right?

I currently have 6 peers that bitcoind "found" when it started.  I'm behind a firewall and not open inbound.  Doesn't seem to be hurting me having only those 6 peers with none inbound, I have about a 10% share "loss" and my efficiency averages above 100%.  But, is that the case across the board, and if not, can it be made better?

Or, as usual, am I just overly concerned about nothing that really requires that concern?