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

[wachtwoord]

Ok I'll check next time I mine. The blue part is the part that is reporting me such awful numbers (16-23 % stales). Efficiency was 87% (plus/minus a number I don't recall)

[twmz]

Ok I'll check next time I mine. The blue part is the part that is reporting me such awful numbers (16-23 % stales). Efficiency was 87% (plus/minus a number I don't recall)

The blue and the red parts are showing the same information in different ways:

Total submitted shares: 180 valid + 8 orphan + 5 dead = 193.
Stale % = (8 orphan + 5 dead) / 193 total = 0.067 = 6.7%

But it is the red part that tells you what kind of problem you have (orphan vs dead).

Of course, these numbers are only useful after p2pool has been running for a few days.  They start out at zero everytime you restart p2pool and for the first day or two, randomness will make it hard to know what the real pattern is.

[spiccioli]

Ok I'll check next time I mine. The blue part is the part that is reporting me such awful numbers (16-23 % stales). Efficiency was 87% (plus/minus a number I don't recall)

The blue and the red parts are showing the same information in different ways:

Total submitted shares: 180 valid + 8 orphan + 5 dead = 193.
Stale % = (8 orphan + 5 dead) / 193 total = 0.067 = 6.7%

But it is the red part that tells you what kind of problem you have (orphan vs dead).

Of course, these numbers are only useful after p2pool has been running for a few days.  They start out at zero everytime you restart p2pool and for the first day or two, randomness will make it hard to know what the real pattern is.

twmz,

there is one thing in my setup which is not "standard" and this is the fact that my miner runs from an usb key so I have the miner at home and p2pool and bitcoind at work, so there is some more latency, I suspect.

I don't have my numbers at hand, but I have a lot less shares (I have less MHs) and a lot more orphans and dead shares.

spiccioli.

[Drsmite]

I am running the latest stable version of BTCMiner (https://bitcointalk.org/index.php?topic=40047.0) pointed at P2Pool.  After a few hours of running (I haven't been around when it fails) it throws an invalid length string exception and shuts down.  After this happens whenever I try to restart the miner it seems to be ineffective and constantly reports that it detects a new block via long polling.  To resolve it I have to kill and restart bitcoind, P2Pool, and BTCMiner.  Is there anyone experiencing a similar error in their miners and odd P2Pool behavior?

[forrestv]

I am running the latest stable version of BTCMiner (https://bitcointalk.org/index.php?topic=40047.0) pointed at P2Pool.  After a few hours of running (I haven't been around when it fails) it throws an invalid length string exception and shuts down.  After this happens whenever I try to restart the miner it seems to be ineffective and constantly reports that it detects a new block via long polling.  To resolve it I have to kill and restart bitcoind, P2Pool, and BTCMiner.  Is there anyone experiencing a similar error in their miners and odd P2Pool behavior?

BTCMiner's exact exception would be helpful in determining the problem - can you pastebin it somewhere? I can't test it as I have no FPGA miner..

[cheat_2_win]

Has P2pool hash rate gone down? I am seeing ~115 GH/s.

[Ranvier]

It's at 156 GH/S. If it isn't around this, you might be on the split chain and should update your client (p2pool still works on the split chain, just more variance).

[Syke]

Is it really possible that total p2pool size including both chains is 270 GH/s?

[cheat_2_win]

It's at 156 GH/S. If it isn't around this, you might be on the split chain and should update your client (p2pool still works on the split chain, just more variance).

Thanks. After updating to a newer p2pool client, I am now seeing 156 GH/s.

[someone703]

Read through this entire thread and didn't really any mention of setting up P2Pool with GUIMiner...

If I downloaded P2Pool from the original post, setup the bitcoin.conf file inside my appdata/bitcoin folder, am I pretty much good to go?

The username / pass match up with the ones I entered in the bitcon.conf:



Also had a receive address show up in my bitcoin wallet as "p2pool".  That was done automatically somewhere during this process?

[Red Emerald]

Read through this entire thread and didn't really any mention of setting up P2Pool with GUIMiner...

If I downloaded P2Pool from the original post, setup the bitcoin.conf file inside my appdata/bitcoin folder, am I pretty much good to go?

The username / pass match up with the ones I entered in the bitcon.conf:



Also had a receive address show up in my bitcoin wallet as "p2pool".  That was done automatically somewhere during this process?

What flags did you use on run_p2pool.py? If you didn't specify a pay address (with -a) then it makes one for you.

The username/password in your miner are NOT the used to access bitcoind.  The username can either be a payment address or any arbitrary name.  If you set the password to the password displayed during p2pool startup, you can track your miner on http://localhost:9332/graphs/

[someone703]

Didn't do anything with a run_p2pool.py file or saw one at all for that matter.  Seems like it did autogenerate a pay address for me.

I pretty much followed Holliday's post on the first page of this thread to a tee except I'm not sure where he got the p2pool_0.8.1_c7feb00.zip file from (step 10).

And then I kind of veered off at step 13 because he uses cgminer in his guide instead of GUIMiner.

When I view that graphs page you linked it says "Install python-rrdtool!".

Will check tomorrow and try to figure it out then.  So tired and need sleep.

[spiccioli]

twmz,

there is one thing in my setup which is not "standard" and this is the fact that my miner runs from an usb key so I have the miner at home and p2pool and bitcoind at work, so there is some more latency, I suspect.

I don't have my numbers at hand, but I have a lot less shares (I have less MHs) and a lot more orphans and dead shares.

spiccioli.

Well, I've restarted p2pool this morning at 0900 GMT+1, tomorrow I'll let you know what numbers I see.

regards.

spiccioli

[Qoheleth]

Haven't received a payout for two days. Is this just another block drought or is something wrong with my setup?
I'm seeing 156GH/s so I'm not on the split chain.

[wachtwoord]

Okay these are the result for running today for 11 hours at 251 Mhash

My miner reports: [251.34 Mhash/sec] [249 Accepted] [1 Rejected] [RPC (+LP)]

p2pool reports:

2012-02-02 18:55:01.238000 Pool: 157GH/s in 17350 shares (17354/17354 verified)
Recent: 0.13% >203MH/s Shares: 6 (0 orphan, 0 dead) Peers: 10 (0 incoming)
2012-02-02 18:55:01.238000 Average time between blocks: 0.41 days
2012-02-02 18:55:01.238000 Pool stales: 9% Own: 20±20% Own efficiency: 88±21%

This seems rather contradictionary to me:

Miner reports: 0.4%  stales (1 in 250)
p2pool reports 0% stales (0 in 6)
p2pool reports 20±20% (where does this number come from)

Also, why does it recognize only 6 of 249 submitted shares?

Can someone explain all this to me?

[Vanderbleek]

Haven't received a payout for two days. Is this just another block drought or is something wrong with my setup?
I'm seeing 156GH/s so I'm not on the split chain.

Since the transactions count as mined, you need 120 confirms for them to show up -- could that be why you're not seeing them?

[cheat_2_win]

Haven't received a payout for two days. Is this just another block drought or is something wrong with my setup?
I'm seeing 156GH/s so I'm not on the split chain.

If you check your address at blockexplorer.com, you will see if you have received any payout. If it shows balance in your p2pool address, then just wait from 120 confirmations. Otherwise 

[Qoheleth]

Haven't received a payout for two days. Is this just another block drought or is something wrong with my setup?
I'm seeing 156GH/s so I'm not on the split chain.

If you check your address at blockexplorer.com, you will see if you have received any payout. If it shows balance in your p2pool address, then just wait from 120 confirmations. Otherwise 

Nah, I'm looking at blockexplorer directly and the last Generation was Block 164634 two days ago.

[someone703]

Sorry, P2Pool has been updated and I missed that part of the instructions. Here is the new step 10.

10. Extract p2pool from p2pool_win32_1685a89.zip p2pool_win32_592ceea.zip.

All this means is unzip the P2Pool file that you've downloaded so that you can run P2Pool.

You should try cgminer, ask Rassah.  

Edit: Updated again with the new P2Pool file.

Haha, no problem.  Earlier in that post you did link the right zip file I believe.  I got P2Pool running and connected just fine.  Saw that it had some 155 Ghash/s or so in the messages that scroll by in the command prompt.

Also, I saw mention of updating Git Head or something.  What's that?  Is that necessary?

And yeah, I've had several people tell me to get cgminer cause GUIMiner is old... Is there no way of using GUIMiner with P2Pool since I've already got it set up at the moment?

[Drsmite]

I am running the latest stable version of BTCMiner (https://bitcointalk.org/index.php?topic=40047.0) pointed at P2Pool.  After a few hours of running (I haven't been around when it fails) it throws an invalid length string exception and shuts down.  After this happens whenever I try to restart the miner it seems to be ineffective and constantly reports that it detects a new block via long polling.  To resolve it I have to kill and restart bitcoind, P2Pool, and BTCMiner.  Is there anyone experiencing a similar error in their miners and odd P2Pool behavior?

BTCMiner's exact exception would be helpful in determining the problem - can you pastebin it somewhere? I can't test it as I have no FPGA miner..

Here is the last segment of BTCMiner's output:

Code:

ztex_ufm1_15d2-04A32E1A6F: f=204.00MHz,  errorRate=0.00%,  maxErrorRate=0.54%,  hash rate: 204.0MH/s,  submitted 0 new nonces,  submitted hash rate 198.5MH/s
New block detected by long polling
New block detected by long polling
ztex_ufm1_15d2-04A32E1A6F: f=204.00MHz,  errorRate=0.00%,  maxErrorRate=0.54%,  hash rate: 204.0MH/s,  submitted 4 new nonces,  submitted hash rate 199.1MH/s
New block detected by long polling
New block detected by long polling
New block detected by long polling
New block detected by long polling
ztex_ufm1_15d2-04A32E1A6F: f=204.00MHz,  errorRate=0.23%,  maxErrorRate=0.54%,  hash rate: 203.5MH/s,  submitted 0 new nonces,  submitted hash rate 199.0MH/s
New block detected by long polling
Exception in thread "Thread-0" java.lang.NumberFormatException: Invalid length of string
	at BTCMiner.hexStrToData(BTCMiner.java:695)
	at NewBlockMonitor.run(BTCMiner.java:159)
ztex_ufm1_15d2-04A32E1A6F: Error: Invalid length of string: Disabling device

  The new block detected message is pretty much constant at all times.

Edit for relevant sections from the BTCMiner code:

Line 159

Code:

BTCMiner.hexStrToData(BTCMiner.jsonParse(BTCMiner.bitcoinRequest(this, BTCMiner.longPollURL, BTCMiner.longPollUser, BTCMiner.longPollPassw, "getwork", ""), "data"), dataBuf);

Line 693-699

Code:

public static void hexStrToData( String str, byte[] buf ) throws NumberFormatException {
	if ( str.length()<buf.length*2 ) 
	    throw new NumberFormatException("Invalid length of string");
	for ( int i=0; i<buf.length; i++) {
	    buf[i] = (byte) Integer.parseInt( str.substring(i*2,i*2+2), 16);
	}
    }

JSON Parse method

Code:

 public static String jsonParse (String response, String parameter) throws ParserException {
	int lp = parameter.length();
	int i = 0;
	while ( i+lp<response.length() && !parameter.equalsIgnoreCase(response.substring(i,i+lp)) )
	    i++;
	i+=lp;
	if ( i>=response.length() )
	    throw new ParserException( "jsonParse: Parameter `"+parameter+"' not found" );
	while ( i<response.length() && response.charAt(i) != ':' )
	    i++;
	i+=1;
	while ( i<response.length() && (byte)response.charAt(i) <= 32 )
	    i++;
	if ( i>=response.length() )
	    throw new ParserException( "jsonParse: Value expected after `"+parameter+"'" );
	int j=i;
	if ( i<response.length() && response.charAt(i)=='"' ) {
	    i+=1;
	    j=i;
	    while ( j<response.length() && response.charAt(j) != '"' )
		j++;
	    if ( j>=response.length() )
		throw new ParserException( "jsonParse: No closing `\"' found for value of paramter `"+parameter+"'" );
	}
	else { 
	    while ( j<response.length() && response.charAt(j) != ',' && response.charAt(j) != /*{*/'}'  ) 
		j++;
	}
	return response.substring(i,j);
    } 

Bitcoin request call

Code:

public static String bitcoinRequest( MsgObj msgObj, String url, String user, String passw, String request, String params) throws MalformedURLException, IOException {
	bcid += 1;
	return httpGet( msgObj, url, user, passw, "{\"jsonrpc\":\"1.0\",\"id\":" + bcid + ",\"method\":\""+ request + "\",\"params\":["+ (params.equals("") ? "" : ("\""+params+"\"")) + "]}" );
    }

The source is available here http://www.ztex.de/btcminer/ and I am running release 120126
It appears the getwork call to P2Pool is returning fewer bytes than expected.