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

[PatMan]

Whats with our super crappy luck the past 5 days?  Does it have some thing to do with Stratum support in V10 and v11 released 5 days ago?

PS

I was wondering the same.  A bit too coincidental.  According to my local graph, 92% are on v11, the rest are on v9.  Something up with v11, or is the mix of v9/v11 causing problems?

M

I don't believe it is coincidental - I think it's definitely related to the new releases, the stratum code is buggy for sure. I reluctantly pulled my rigs off of p2pool a couple of days ago due to this, which hurt as I completely love the p2p ethos with this pool, but at the moment it's just not working as it should be. Been running on Ozcoin (stratum enabled) for two days with zero problems & zero rejects, this is how it should be with stratum enabled I believe.

I want to get back on to p2pool, but until the new code is fixed/stable it's just not worth it I'm afraid.

Peace.

[spiccioli]

I don't believe it is coincidental - I think it's definitely related to the new releases, the stratum code is buggy for sure. I reluctantly pulled my rigs off of p2pool a couple of days ago due to this, which hurt as I completely love the p2p ethos with this pool, but at the moment it's just not working as it should be. Been running on Ozcoin (stratum enabled) for two days with zero problems & zero rejects, this is how it should be with stratum enabled I believe.

I want to get back on to p2pool, but until the new code is fixed/stable it's just not worth it I'm afraid.

Peace.

PatMan,

you can still use p2pool without stratum adding --fix-protocol to cgminer running command line.

spiccioli

[spiccioli]

About memory leaking, this is a ubuntu 64bit after a little more than 48 hours of running latest p2pool

Code:

PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND 
5409 user      20   0 1371m 1.2g 3552 S  8.3 30.9 168:18.05 python    

and I have 7 peers with 4 incoming ones.

spiccioli.

[PatMan]

I did that

I don't believe it is coincidental - I think it's definitely related to the new releases, the stratum code is buggy for sure. I reluctantly pulled my rigs off of p2pool a couple of days ago due to this, which hurt as I completely love the p2p ethos with this pool, but at the moment it's just not working as it should be. Been running on Ozcoin (stratum enabled) for two days with zero problems & zero rejects, this is how it should be with stratum enabled I believe.

I want to get back on to p2pool, but until the new code is fixed/stable it's just not worth it I'm afraid.

Peace.

PatMan,

you can still use p2pool without stratum adding --fix-protocol to cgminer running command line.

spiccioli

Yes, I did that as soon as it started when the new releases came out (see previous posts), but it seemed to defy the while idea for me. If stratum is available then why not use it? Given the choice between 1) getwork on my favorite pool with high DOA/rejects & low payout, 2) stratum on my favorite pool with medium DOA/rejects & low payout, or 3) stratum on a different pool with zero DOA/rejects & good payout - well, it's a no brainer TBH. Don't get me wrong, I'm not just in this for the payouts, but I also can't afford to spend money on wasted electricity for nothing.

I just hope that the fix comes soon, because with stratum enabled and working properly on p2pool it will simply blow the other pools away in my opinion. I'm just a small time GPU miner and still finding my way around the scene, but the numbers and graphs don't lie. I don't know nearly enough about bitcoin and it's inner workings yet, but I can do Math & read a graph.

Peace.

[mdude77]

I did that

I don't believe it is coincidental - I think it's definitely related to the new releases, the stratum code is buggy for sure. I reluctantly pulled my rigs off of p2pool a couple of days ago due to this, which hurt as I completely love the p2p ethos with this pool, but at the moment it's just not working as it should be. Been running on Ozcoin (stratum enabled) for two days with zero problems & zero rejects, this is how it should be with stratum enabled I believe.

I want to get back on to p2pool, but until the new code is fixed/stable it's just not worth it I'm afraid.

Peace.

PatMan,

you can still use p2pool without stratum adding --fix-protocol to cgminer running command line.

spiccioli

Yes, I did that as soon as it started when the new releases came out (see previous posts), but it seemed to defy the while idea for me. If stratum is available then why not use it? Given the choice between 1) getwork on my favorite pool with high DOA/rejects & low payout, 2) stratum on my favorite pool with medium DOA/rejects & low payout, or 3) stratum on a different pool with zero DOA/rejects & good payout - well, it's a no brainer TBH. Don't get me wrong, I'm not just in this for the payouts, but I also can't afford to spend money on wasted electricity for nothing.

I just hope that the fix comes soon, because with stratum enabled and working properly on p2pool it will simply blow the other pools away in my opinion. I'm just a small time GPU miner and still finding my way around the scene, but the numbers and graphs don't lie. I don't know nearly enough about bitcoin and it's inner workings yet, but I can do Math & read a graph.

Peace.

I'm on bitminter because it supports namecoin and stratum. 

M

[DonChate]

There do seem to be issues with v11 stratum + latest cgminer.

All shares are logging errors like "submitted share with hash > target" using stratum and the diffs running are mostly 1 resulting in a ton of share submissions. Using --fix-protocol in cgminer and forcing getwork seems to work as usual, with much higher diffs being given by p2pool.

[spiccioli]

About memory leaking, this is a ubuntu 64bit after a little more than 48 hours of running latest p2pool

Code:

PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND 
5409 user      20   0 1371m 1.2g 3552 S  8.3 30.9 168:18.05 python    

and I have 7 peers with 4 incoming ones.

spiccioli.

Here is a fedora 16, 32 bit, using pypy after three days, more or less

Code:

14844 user  20   0 1427m 1.3g  14m S  8.3 23.8 428:34.52 pypy

spiccioli.

[gyverlb]

About memory leaking, this is a ubuntu 64bit after a little more than 48 hours of running latest p2pool

Code:

PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND 
5409 user      20   0 1371m 1.2g 3552 S  8.3 30.9 168:18.05 python    

and I have 7 peers with 4 incoming ones.

spiccioli.

Here is a fedora 16, 32 bit, using pypy after three days, more or less

Code:

14844 user  20   0 1427m 1.3g  14m S  8.3 23.8 428:34.52 pypy

spiccioli.

Gentoo 64 bit, python 2.7.3 after 5 days:

Code:

bitcoin   9030  5.8 11.5 973680 705960 pts/12  Ssl+ Jan02 439:01 /usr/bin/python2.7 run_p2pool.py --disable-upnp --max-conns 6 [...]

I left the options that didn't leak any authentication info, I use merged mining on namecoin. It uses more RAM than it used to (some earlier version didn't need much more than 300M).

[Red Emerald]

About memory leaking, this is a ubuntu 64bit after a little more than 48 hours of running latest p2pool

Code:

PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND 
5409 user      20   0 1371m 1.2g 3552 S  8.3 30.9 168:18.05 python    

and I have 7 peers with 4 incoming ones.

spiccioli.

Here is a fedora 16, 32 bit, using pypy after three days, more or less

Code:

14844 user  20   0 1427m 1.3g  14m S  8.3 23.8 428:34.52 pypy

spiccioli.

Gentoo 64 bit, python 2.7.3 after 5 days:

Code:

bitcoin   9030  5.8 11.5 973680 705960 pts/12  Ssl+ Jan02 439:01 /usr/bin/python2.7 run_p2pool.py --disable-upnp --max-conns 6 [...]

I left the options that didn't leak any authentication info, I use merged mining on namecoin. It uses more RAM than it used to (some earlier version didn't need much more than 300M).

I'm merge mining namecoin with 0981bdfb4e3fc266463be104c8e5441720173cb8 on Ubuntu 12.04.1 LTS 64-bit with python 2.7.3.  However, I'm not actively mining on my node.

Code:

Node uptime: 2.771 days Peers: 6 out, 21 in

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                                                   
29401 coiner    20   0  635m 366m 4012 S   15  9.3 321:10.29 python    

I asked this on IRC, but didn't get much response.

I just have a small GPU rig and so I stopped mining on p2pool.  What benefits are there to p2pool if I keep my node online?  It has relatively fast hard drives and very fast and stable internet.  I'm guessing it would make a DOS attack against p2pool harder.  Would one node that isn't mining make a difference in an attack?  It might also help with block progation and could reduce orphans.  Any way to quantify how much?

[rav3n_pl]

Dear TRC users, pls upgrade your Terracoin client and p2pool. There is at least 3 nodes that are messing in TRC pool:

Code:

88.69.122.87
94.219.158.58
188.97.132.96

I have blocked this IPs because of huge wasted traffic (peers are asking for shares over and over again).

[lenny_]

Dear TRC users, pls upgrade your Terracoin client and p2pool. There is at least 3 nodes that are messing in TRC pool:

Code:

88.69.122.87
94.219.158.58
188.97.132.96

I have blocked this IPs because of huge wasted traffic (peers are asking for shares over and over again).

Maybe forrestv can do a fork? So users with version below specific one will be forked to smaller, separated p2pool./

[zvs]

Dear TRC users, pls upgrade your Terracoin client and p2pool. There is at least 3 nodes that are messing in TRC pool:

Code:

88.69.122.87
94.219.158.58
188.97.132.96

I have blocked this IPs because of huge wasted traffic (peers are asking for shares over and over again).

along these lines, I've been trying to optimize my own peer list by seeing which ones send dupe transactions, which ones never are the 'first' to send me a new share, etc.  i guess it isn't 'fair' to ppl on bad connections, but..  i've been able to up my max blocksize to 100,000 and now at 51 shares w/ 1 orphan and 1 dead.  you can expect around 3-5% just on sheer chance, really. 

[forrestv]

FYI, I've been looking at the memory leak issue for 2 days and think I've made some progress. More news soon.

[forrestv]

The leak (or at least one leak) is fixed in commit 348aba84c6. Share objects previously held a reference to the peer object the share was received from. This kept the peer object in memory for as long as the share was, even if the connection to the peer was lost. Continually losing connections and reconnecting to peers could cause these peer objects to build up in memory, potentially leaking memory quickly since the peer objects contain an incoming message buffer.

[forrestv]

P2Pool release 11.1 tag: 11.1 hash: 348aba84c627879ee229447731bbff4cd566b4bd

Windows binary: http://u.forre.st/u/kurgptuu/p2pool_win32_11.1.zip
Source zipball: https://github.com/forrestv/p2pool/zipball/11.1
Source tarball: https://github.com/forrestv/p2pool/tarball/11.1

Changes:
* Fixed memory leak of peer objects
* Added memory usage graph (doesn't yet work on Windows)

[Schleicher]

Changes:
* Fixed memory leak of peer objects
* Added memory usage graph

Is the graph supposed to look like this?

There's something missing, I think.
The other graphs are ok.

[bitpop]

I welcome the extra memory, I have 32GB to burn.

[forrestv]

Changes:
* Fixed memory leak of peer objects
* Added memory usage graph

Is the graph supposed to look like this?

There's something missing, I think.
The other graphs are ok.

Ah, I neglected to mention that that graph doesn't (yet) work on Windows. Sorry.

[spiccioli]

P2Pool release 11.1 tag: 11.1 hash: 348aba84c627879ee229447731bbff4cd566b4bd

Windows binary: http://u.forre.st/u/kurgptuu/p2pool_win32_11.1.zip
Source zipball: https://github.com/forrestv/p2pool/zipball/11.1
Source tarball: https://github.com/forrestv/p2pool/tarball/11.1

Changes:
* Fixed memory leak of peer objects
* Added memory usage graph (doesn't yet work on Windows)

Thanks forrestv, very good news!

spiccioli

[sharky112065]

Forrestv,

My memory usage went up for P2pool on the latest git version.

Currently at 824M Resident and 975 Virtual.

I would have told you in IRC but you were disconnected.

-----

Version: 11.1-1-g8c9f2ca

Pool rate: 318GH/s (14% DOA+orphan) Share difficulty: 639

Node uptime: 1.578 days Peers: 10 out, 5 in

Local rate: 6.62GH/s (1.7% DOA) Expected time to share: 0.115 hours

Shares: 358 total (27 orphaned, 9 dead) Efficiency: 104.2%

-----

Date   Memory Usage/(B)
Sat Jan 12 2013 01:38:24 GMT-0800 (Pacific Standard Time)  867M
Sat Jan 12 2013 01:04:48 GMT-0800 (Pacific Standard Time)  853M
Sat Jan 12 2013 00:31:12 GMT-0800 (Pacific Standard Time)  851M
Fri Jan 11 2013 23:57:36 GMT-0800 (Pacific Standard Time)   843M
Fri Jan 11 2013 23:24:00 GMT-0800 (Pacific Standard Time)   829M
Fri Jan 11 2013 22:50:24 GMT-0800 (Pacific Standard Time)   822M
Fri Jan 11 2013 22:16:48 GMT-0800 (Pacific Standard Time)   814M
Fri Jan 11 2013 21:43:12 GMT-0800 (Pacific Standard Time)   804M
Fri Jan 11 2013 21:09:36 GMT-0800 (Pacific Standard Time)   797M
Fri Jan 11 2013 20:36:00 GMT-0800 (Pacific Standard Time)   787M
Fri Jan 11 2013 20:02:24 GMT-0800 (Pacific Standard Time)   783M
Fri Jan 11 2013 19:28:48 GMT-0800 (Pacific Standard Time)   780M
Fri Jan 11 2013 18:55:12 GMT-0800 (Pacific Standard Time)   772M
Fri Jan 11 2013 18:21:36 GMT-0800 (Pacific Standard Time)   757M
Fri Jan 11 2013 17:48:00 GMT-0800 (Pacific Standard Time)   751M
Fri Jan 11 2013 17:14:24 GMT-0800 (Pacific Standard Time)   743M
Fri Jan 11 2013 16:40:48 GMT-0800 (Pacific Standard Time)   736M
Fri Jan 11 2013 16:07:12 GMT-0800 (Pacific Standard Time)   729M
Fri Jan 11 2013 15:33:36 GMT-0800 (Pacific Standard Time)   717M
Fri Jan 11 2013 15:00:00 GMT-0800 (Pacific Standard Time)   705M
Fri Jan 11 2013 14:26:24 GMT-0800 (Pacific Standard Time)   693M
Fri Jan 11 2013 13:52:48 GMT-0800 (Pacific Standard Time)   685M
Fri Jan 11 2013 13:19:12 GMT-0800 (Pacific Standard Time)   678M
Fri Jan 11 2013 12:45:36 GMT-0800 (Pacific Standard Time)   663M
Fri Jan 11 2013 12:12:00 GMT-0800 (Pacific Standard Time)   652M
Fri Jan 11 2013 11:38:24 GMT-0800 (Pacific Standard Time)   640M
Fri Jan 11 2013 11:04:48 GMT-0800 (Pacific Standard Time)   636M
Fri Jan 11 2013 10:31:12 GMT-0800 (Pacific Standard Time)   624M
Fri Jan 11 2013 09:57:36 GMT-0800 (Pacific Standard Time)   615M
Fri Jan 11 2013 09:24:00 GMT-0800 (Pacific Standard Time)   607M
Fri Jan 11 2013 08:50:24 GMT-0800 (Pacific Standard Time)   599M
Fri Jan 11 2013 08:16:48 GMT-0800 (Pacific Standard Time)   591M
Fri Jan 11 2013 07:43:12 GMT-0800 (Pacific Standard Time)   577M
Fri Jan 11 2013 07:09:36 GMT-0800 (Pacific Standard Time)   565M
Fri Jan 11 2013 06:36:00 GMT-0800 (Pacific Standard Time)   559M
Fri Jan 11 2013 06:02:24 GMT-0800 (Pacific Standard Time)   544M
Fri Jan 11 2013 05:28:48 GMT-0800 (Pacific Standard Time)   541M
Fri Jan 11 2013 04:55:12 GMT-0800 (Pacific Standard Time)   533M
Fri Jan 11 2013 04:21:36 GMT-0800 (Pacific Standard Time)   517M
Fri Jan 11 2013 03:48:00 GMT-0800 (Pacific Standard Time)   506M
Fri Jan 11 2013 03:14:24 GMT-0800 (Pacific Standard Time)   502M
Fri Jan 11 2013 02:40:48 GMT-0800 (Pacific Standard Time)   495M
Fri Jan 11 2013 02:07:12 GMT-0800 (Pacific Standard Time)   478M