[ANN][YAC][POOL]p2pool for yacoin!!

[rav3n_pl]

did I miss something?

Yes, creation of yacoin.conf :F

[fronti]

did I miss something?

Yes, creation of yacoin.conf :F

ok I add this also

[fronti]

Remember to use the latest cpuminer with the -a scrypt-jane option

which is the latest for windows? I tried here one and this is not working.
ok, maybe i need also try to setup a compile system here for this

[caish5]

Not sure, i compiled on linux. From the git at https://github.com/ali1234/cpuminer

[ntkrnl]

NO, it's a bug.
The newest protocol  conflicts with the old one, that's why I changed the server to the new version, and changed back again.
We must use a new port and a new protocol  prefix. I have update the source on github a moment ago. Please download the newest one.
https://github.com/cctvsmg/yacoin-p2pool   

the newest version use a port 8338 like:
http://pool.bitcn.org:8338

I've got all the bugs sorted on my node now, so if you want to mine some YAC on a good solid server then head over to.....

--url http://mining.bitvolcano.com:8338
--user youryacaddress
--pass x

Remember to use the latest cpuminer with the -a scrypt-jane option
There is a 1% fee.
Stats are at http://bitvolcano.com:8338

Take out that addition you don't need it, the peers are propagated though the net from the boot node.

Ok, cool. Have updated to the latest code and restarted p2pool.

http://yac.procrypto.com:8080/ - 0% fee, 0% to author.

Solid server with pro host in top data centre.

Also throwing a few (loss making) server instances at it to get the hashrate up.

The lastest version seems to be working very well, the hash "00000000060a41a3dac8ffa43c30d09372b73be0a841804e9ffdc00078d43059" can be found in the blockchain. 

[procrypto]

What would be hardware good enough to accommodate 100 - 1000 miners?

The answer is, more than I've allocated so far, because the web page for mine is starting to lag with ~50 workers!  

Will scale it up shortly and let you know   If your miners get disconnected, it shouldn't be for more than 30 seconds.

[procrypto]

What is the weakest link or bottleneck when running P2Pool? I guess RAM but I would not bet on that, LOL, I actualy don't even know how pool works behind the scene but I'll need
to learn or find someone trustable to do it for me. I would like to do it with classic pool and PPS payout system though so that miners can use YAC as soon as they submitted share.

Not sure as yet, will monitor it. Doesn't really appear to use much RAM or CPU power though (circa 55MB RAM for p2pool,  85MB for yacoind and minimal CPU use at current load).

My node is going well, hashrate is decent and we've been picking up a few blocks. Throw all of your might at it, and we could be doing well!

Would the OP consider updating his post with the current live nodes and their fee structures?

[erk]

What would be hardware good enough to accommodate 100 - 1000 miners?

The answer is, more than I've allocated so far, because the web page for mine is starting to lag with ~50 workers!  

Will scale it up shortly and let you know  If your miners get disconnected, it shouldn't be for more than 30 seconds.

What is the weakest link or bottleneck when running P2Pool? I guess RAM but I would not bet on that, LOL, I actualy don't even know how pool works behind the scene but I'll need
to learn or find someone trustable to do it for me. I would like to do it with classic pool and PPS payout system though so that miners can use YAC as soon as they submitted share.

The p2pool web extended interface doesn't scale well with lots of miners, it crashed my Firefox after a while. I replaced the web-static folder with the standard p2pool boring one. Both are crap, but the standard one doesn't crash the browser with pushes. p2pool needs a new web interface that can organize the stats for a lot of miners, the current liner graph page is a joke when there are more than 10 or so miners on the node.

[procrypto]

The p2pool web extended interface doesn't scale well with lots of miners, it crashed my Firefox after a while. I replaced the web-static folder with the standard p2pool boring one. Both are crap, but the standard one doesn't crash the browser with pushes. p2pool needs a new web interface that can organize the stats for a lot of miners, the current liner graph page is a joke when there are more than 10 or so miners on the node.

Yeah, I wondered what was slowing it down earlier, I was at work, so didn't have time to poke into it.  It's most likely the loading of jQuery from twitter.github.com (prob fixed by loading jQuery locally instead) - that's what seems to hold my browser up when traffic increases.

[ntkrnl]

What is the weakest link or bottleneck when running P2Pool? I guess RAM but I would not bet on that, LOL, I actualy don't even know how pool works behind the scene but I'll need
to learn or find someone trustable to do it for me. I would like to do it with classic pool and PPS payout system though so that miners can use YAC as soon as they submitted share.

Not sure as yet, will monitor it. Doesn't really appear to use much RAM or CPU power though (circa 55MB RAM for p2pool,  85MB for yacoind and minimal CPU use at current load).
My node is going well, hashrate is decent and we've been picking up a few blocks. Throw all of your might at it, and we could be doing well!
Would the OP consider updating his post with the current live nodes and their fee structures?

pypy may be helpful to reduce the cpu usage. There is a memory leak in P2PoolExtendedFrontEnd, and jquery & bootstrap lib should be speeded up with a CDN. This will be fixed in the next version.

The main thread is updated, please download the lastest source or binary, this time we'll be really mining in only one pool.

[DonChate]

Glad the issues are getting sorted. Will check out the new sources.

[caish5]

Anyone mining at bitvolcano please change to port 8336 so we're on the main p2pool.
I've updated to the latest git repo.

[fronti]

Anyone mining at bitvolcano please change to port 8336 so we're on the main p2pool.
I've updated to the latest git repo.

i updated my small pool also

[procrypto]

Moved my p2ool node to port 8336 and updated to the latest code. Thought we were all mining together anyway, but wanted to be sure.

Please update your miner if you were mining on my node. 

[procrypto]

Bit confused, graphs page shows "Current payout to default address" as a few fractions of a YAC, however, I've set the fees to 0% and I'm the only person mining on my own node right now. I thought this amount should show as zero.

Any help OP? Why does it show this amount?

[microxp]

wo，thank you for your great work!

[procrypto]

P2Pool does not use PHP / MySQL?  

Nope, uses python, and stores data in text format (in /yacoin-p2pool/data/yacoin/)

edit: There is a very good p2pool info post - https://bitcointalk.org/index.php?topic=153232.0 - when I get time I plan to try some of the tuning suggested there. Answers some questions about bottlenecks/latency.

Incidentally my efficiency has dropped massively since I updated code and changed ports. Hopefully a temporary issue and it will straighten out as I find more shares. Any thoughts OP?

[altsay]

Could anyone explain the difference between Pushpool and p2pool? Which one is better?

[rav3n_pl]

Could anyone explain the difference between Pushpool and p2pool? Which one is better?

In pushpool you are running OWN pool and allow other ppl mine in your pool.
In P2Pool we ALL mine in ONE pool. And other ppl can mine in "open" nodes.

[ntkrnl]

Bit confused, graphs page shows "Current payout to default address" as a few fractions of a YAC, however, I've set the fees to 0% and I'm the only person mining on my own node right now. I thought this amount should show as zero.

Any help OP? Why does it show this amount?

Yeah, here is the reason:

The broken code on github hasn't been updated in 3 days.

What's the progress?

It's OK now.

Thanks, that seems to be working.
Is there anyway to set the default payout address or is that removed?

In the POS(PROOF OF STACK) featured coin, when a block is mined, and submitted to the client, it must be signed with your public key, while bitcoin/litecoin don't have to. That' why novacoin p2pool delete "-address" option.  Because when you specify an address, it's just a hash160(pubkey), not your publish key, yacoind will throw an error "sign block failed".

In the lastest version of yacoin-p2pool, no matter what the "fee" is, the host user of p2pool will be paid at least 0.01, just for the pubkey, without this, you can never sign a block.

So, if you want to specify an individual address, use the worker-name-as-address mode.

UPDATE:
mmh, there is still something wrong with the payment system
In other words, ppcoin novacoin yacoin basiclly don't support paid-by-worker mode( -u username as address). You can only run p2pool minerd yacoind on your own machine and pay to yourselft.

So if I've been running this P2Pool on my server overnight, connected miners in the usual P2Pool way (address as username, any password), and found a couple of blocks, where would the payouts have gone to?

Also, for a complete noob like me, how do I go about changing my setup for now, so that all work goes towards the single wallet address on my server?

Thanks for your help.. will definitely be giving you a solid tip once the teething problems are fixed.

In fact, if you choose the address-as-name mode, you got nothing, the block was not successfully submitted with an error "Sign failed" in /.yacoin/debug.log  
Here is some technical details:

1.Compared with bitcoin litecoin, there is a quite different feature in ppcoin novacoin yacoin . If a block is found, the p2pool will call a  jsonprc method called "submitblock", here is the "submitblock" funtion in yacoin
's source:

Code:

// ppcoin: sign block
bool CBlock::SignBlock(const CKeyStore& keystore)
{
    vector<valtype> vSolutions;
    txnouttype whichType;

    if(!IsProofOfStake())
    {
        for(unsigned int i = 0; i < vtx[0].vout.size(); i++)
        {
            const CTxOut& txout = vtx[0].vout[i];

            if (!Solver(txout.scriptPubKey, whichType, vSolutions))
                continue;

            if (whichType == TX_PUBKEY)
            {
                // Sign
                valtype& vchPubKey = vSolutions[0];
                CKey key;

                if (!keystore.GetKey(Hash160(vchPubKey), key))
                    continue;
                if (key.GetPubKey() != vchPubKey)
                    continue;
                if(!key.Sign(GetHash(), vchBlockSig))
                    continue;

                return true;
            }
        }
    }
    else
    {
        const CTxOut& txout = vtx[1].vout[1];

        if (!Solver(txout.scriptPubKey, whichType, vSolutions))
            return false;

        if (whichType == TX_PUBKEY)
        {
            // Sign
            valtype& vchPubKey = vSolutions[0];
            CKey key;

            if (!keystore.GetKey(Hash160(vchPubKey), key))
                return false;
            if (key.GetPubKey() != vchPubKey)
                return false;

            return key.Sign(GetHash(), vchBlockSig);
        }
    }

[color=red]    printf("Sign failed\n");[/color]
    return false;
}

The variable "keystore" is the keys(pubkey pubkey_hash privatekey etc.) of your wallet. That means if you submit a block, you should sign it up with the pubkeys from your wallet,  but  username-as-address mode just only sends p2pool your pubkey_hash.  you can easily get a pubkey_hash with an address, but never get pubkey, just like you want to get the text password from md5.

I've already found a tricky way to fix this problem, just wait for the next version