[1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool - page 558.

Mogumodz

sr. member

Activity: 290

Merit: 250

Quote from: bitpop on April 29, 2013, 09:49:06 PM

Yes that's exactly the procedure in the readme if I remember correctly.

Thanks

bitpop

legendary

Activity: 2912

Merit: 1060

Yes that's exactly the procedure in the readme if I remember correctly.

Mogumodz

sr. member

Activity: 290

Merit: 250

Anyone see any problem running and updating 11.4 along side my working 11.3 node while it verifies all the shares?

I seem to recall this being asked a few pages back but can't seem to find it so might have dreamt it.

The output says it can't bind to the p2p port or worker port so my current mining, website stats, graphs etc are unaffected but it's verifying shares so thought could this work?

Just wanted to know If this is ok before I assume anything.

Richy_T

legendary

Activity: 2576

Merit: 2267

1RichyTrEwPYjZSeAYxeiFBNnKC9UjC5k

Quote from: forrestv on April 29, 2013, 07:55:12 PM

Quote from: Richy_T on April 29, 2013, 07:49:35 PM

Some kind of DOS attack?

Got three "GOT INCOMPLETE BLOCK FROM PEER" messages and then p2pool stopped. Peer IP is 192.203.228.65

Now just got another one.

What do you mean by "p2pool stopped"?

Miner got disconnected, p2pool not doing anything apparently though since a fourth message came up, it looks like it was not completely dead. I did a simple restart of the process and things seem fine so far. I guess I waited about 5 minutes before restarting it (did not have to reboot).

Richy_T

legendary

Activity: 2576

Merit: 2267

1RichyTrEwPYjZSeAYxeiFBNnKC9UjC5k

Quote from: evilpete on April 29, 2013, 08:11:48 PM

Quote from: Richy_T on April 29, 2013, 07:49:35 PM

Some kind of DOS attack?

Got three "GOT INCOMPLETE BLOCK FROM PEER" messages and then p2pool stopped. Peer IP is 192.203.228.65

Now just got another one.

Is this you by any chance?

2013-04-29 16:39:08.307039 Incoming connection to peer 74.100.64.248:56708 established. p2pool version: 1100 '11.2'
2013-04-29 16:39:08.472251 Sending 1 shares to 74.100.64.248:56708
2013-04-29 16:39:08.474522 Lost peer 74.100.64.248:56708 - Connection was closed cleanly.
2013-04-29 16:39:08.475089 > in handle_share_hashes:
2013-04-29 16:39:08.475163 > Traceback (most recent call last):
2013-04-29 16:39:08.475245 > Failure: twisted.internet.error.ConnectionDone: Connection was closed cleanly.

That particular peer is having a lot of trouble:

$ grep 'Lost peer' log | awk '{print $5}' | sed -e 's/:.*//' | sort | uniq -c | sort -nr
168 74.100.64.248
5 123.159.6.63
2 216.153.20.29
.. nothing else more than one..

The #2 on the list is doing things like this:

2013-04-29 08:50:54.110763 Connection timed out, disconnecting from 123.159.6.63:54137
2013-04-29 08:50:54.111417 Lost peer 123.159.6.63:54137 -
2013-04-29 08:50:54.111508 Connection was aborted locally, using
2013-04-29 08:50:54.111569 L{twisted.internet.interfaces.ITCPTransport.abortConnection}.

Are either of those two yours?

No. But see PM.

evilpete

member

Activity: 77

Merit: 10

Quote from: Richy_T on April 29, 2013, 07:49:35 PM

Some kind of DOS attack?

Got three "GOT INCOMPLETE BLOCK FROM PEER" messages and then p2pool stopped. Peer IP is 192.203.228.65

Now just got another one.

Is this you by any chance?

2013-04-29 16:39:08.307039 Incoming connection to peer 74.100.64.248:56708 established. p2pool version: 1100 '11.2'
2013-04-29 16:39:08.472251 Sending 1 shares to 74.100.64.248:56708
2013-04-29 16:39:08.474522 Lost peer 74.100.64.248:56708 - Connection was closed cleanly.
2013-04-29 16:39:08.475089 > in handle_share_hashes:
2013-04-29 16:39:08.475163 > Traceback (most recent call last):
2013-04-29 16:39:08.475245 > Failure: twisted.internet.error.ConnectionDone: Connection was closed cleanly.

That particular peer is having a lot of trouble:

$ grep 'Lost peer' log | awk '{print $5}' | sed -e 's/:.*//' | sort | uniq -c | sort -nr
168 74.100.64.248
5 123.159.6.63
2 216.153.20.29
.. nothing else more than one..

The #2 on the list is doing things like this:

2013-04-29 08:50:54.110763 Connection timed out, disconnecting from 123.159.6.63:54137
2013-04-29 08:50:54.111417 Lost peer 123.159.6.63:54137 -
2013-04-29 08:50:54.111508 Connection was aborted locally, using
2013-04-29 08:50:54.111569 L{twisted.internet.interfaces.ITCPTransport.abortConnection}.

Are either of those two yours?

forrestv

hero member

Activity: 516

Merit: 643

Quote from: Richy_T on April 29, 2013, 07:49:35 PM

Some kind of DOS attack?

Got three "GOT INCOMPLETE BLOCK FROM PEER" messages and then p2pool stopped. Peer IP is 192.203.228.65

Now just got another one.

What do you mean by "p2pool stopped"?

mdude77

legendary

Activity: 1540

Merit: 1001

Quote from: Richy_T on April 29, 2013, 07:49:35 PM

Some kind of DOS attack?

Got three "GOT INCOMPLETE BLOCK FROM PEER" messages and then p2pool stopped. Peer IP is 192.203.228.65

Now just got another one.

Mine's been acting funny the last 24 hours too. I should be between .25 and .30 per block, been below .20 the last 5.

M

Richy_T

legendary

Activity: 2576

Merit: 2267

1RichyTrEwPYjZSeAYxeiFBNnKC9UjC5k

Some kind of DOS attack?

Got three "GOT INCOMPLETE BLOCK FROM PEER" messages and then p2pool stopped. Peer IP is 192.203.228.65

Now just got another one.

forrestv

hero member

Activity: 516

Merit: 643

Quote from: gyverlb on April 29, 2013, 02:46:44 PM

forrestv: you may want to update the subject of the thread, the pool seems to stabilize at more than 650GH/s now.

Good point. Used 700GH/s, as it is now

gyverlb

hero member

Activity: 896

Merit: 1000

forrestv: you may want to update the subject of the thread, the pool seems to stabilize at more than 650GH/s now.

forrestv

hero member

Activity: 516

Merit: 643

P2Pool release 11.4 commit hash: 0cb07df5623645bc1082f87d3380f6d085a609be

Windows binary: http://u.forre.st/u/jlowtmzq/p2pool_win32_11.4.zip
Windows binary signature: http://u.forre.st/u/flasczol/p2pool_win32_11.4.zip.sig
Source zipball: https://github.com/forrestv/p2pool/zipball/11.4
Source tarball: https://github.com/forrestv/p2pool/tarball/11.4

Changes:
* Fixed a few more potential memory leaks
* Memory usage graph now works in Windows, thanks to David Kassa

mdude77

legendary

Activity: 1540

Merit: 1001

Quote from: -ck on April 29, 2013, 01:50:25 AM

Quote from: zvs on April 29, 2013, 01:01:45 AM

Quote from: -ck on April 28, 2013, 05:34:09 PM

So I tried the current version of p2pool on a single 5.7GHz Jalapeño with cgminer connecting via stratum. This works out to about 80 diff1 shares per minute. It "works" in a sense of the word, without causing any weird errors or duplicating work that happen with the Avalon but there are huge lags between share submission and responses from p2pool after a while. Watching the CPU usage, I see a spike to 100% of one core whenever a block of transactions comes in and that coincides with the lag to respond to share submission, leading to more stale rejects if that occurs at the wrong time. The CPU in question is a 12 thread 3.4GHz CPU so it is clearly not underpowered. I suspect what happens on the Avalon is simply a grossly exaggerated form of this leading to misbehaviour.

the same thing happens with GPUs while using cgminer (not a cgminer problem, but stratum I would think)

i think i commented on that some months ago

minerd picks up the new work much quicker, and it is better to use phoenix for p2pool, or not use stratum

Yes using getwork with the jalapeño is actually better than using stratum, which is counter-intuitive and adds weight to the argument that there is a problem with p2pool's stratum implementation. Having said that, if forrestv can get one as a donation from BFL, that will benefit everyone because I'm sure there will be incentive on his part to investigate and develop further. Many people have approached me about rewriting p2pool from scratch in c, and as much fun as that sounds, I seriously don't have the time to support another project of this magnitude and would rather see the original author continue it.

You know, funny you should say that. I was considering doing the same thing ... rewriting in C. It would be quite a project indeed.

M

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: zvs on April 29, 2013, 01:01:45 AM

Quote from: -ck on April 28, 2013, 05:34:09 PM

So I tried the current version of p2pool on a single 5.7GHz Jalapeño with cgminer connecting via stratum. This works out to about 80 diff1 shares per minute. It "works" in a sense of the word, without causing any weird errors or duplicating work that happen with the Avalon but there are huge lags between share submission and responses from p2pool after a while. Watching the CPU usage, I see a spike to 100% of one core whenever a block of transactions comes in and that coincides with the lag to respond to share submission, leading to more stale rejects if that occurs at the wrong time. The CPU in question is a 12 thread 3.4GHz CPU so it is clearly not underpowered. I suspect what happens on the Avalon is simply a grossly exaggerated form of this leading to misbehaviour.

the same thing happens with GPUs while using cgminer (not a cgminer problem, but stratum I would think)

i think i commented on that some months ago

minerd picks up the new work much quicker, and it is better to use phoenix for p2pool, or not use stratum

Yes using getwork with the jalapeño is actually better than using stratum, which is counter-intuitive and adds weight to the argument that there is a problem with p2pool's stratum implementation. Having said that, if forrestv can get one as a donation from BFL, that will benefit everyone because I'm sure there will be incentive on his part to investigate and develop further. Many people have approached me about rewriting p2pool from scratch in c, and as much fun as that sounds, I seriously don't have the time to support another project of this magnitude and would rather see the original author continue it.

zvs

legendary

Activity: 1680

Merit: 1000

https://web.archive.org/web/*/nogleg.com

Quote from: TurdHurdur on April 28, 2013, 10:31:55 PM

Quote from: -ck on April 28, 2013, 05:34:09 PM

So I tried the current version of p2pool on a single 5.7GHz Jalapeño with cgminer connecting via stratum. This works out to about 80 diff1 shares per minute. It "works" in a sense of the word, without causing any weird errors or duplicating work that happen with the Avalon but there are huge lags between share submission and responses from p2pool after a while. Watching the CPU usage, I see a spike to 100% of one core whenever a block of transactions comes in and that coincides with the lag to respond to share submission, leading to more stale rejects if that occurs at the wrong time. The CPU in question is a 12 thread 3.4GHz CPU so it is clearly not underpowered. I suspect what happens on the Avalon is simply a grossly exaggerated form of this leading to misbehaviour.

Does a username like ckolivas/2000+10 help?

no, because you'd still get the 6 new works every minute

the ratio of rejects to accepts would be the same

though, it would be nice if everyone over 5ghash or so did 2000 shares, so people at lower rates may actually get something at some point

zvs

legendary

Activity: 1680

Merit: 1000

https://web.archive.org/web/*/nogleg.com

Quote from: -ck on April 28, 2013, 05:34:09 PM

So I tried the current version of p2pool on a single 5.7GHz Jalapeño with cgminer connecting via stratum. This works out to about 80 diff1 shares per minute. It "works" in a sense of the word, without causing any weird errors or duplicating work that happen with the Avalon but there are huge lags between share submission and responses from p2pool after a while. Watching the CPU usage, I see a spike to 100% of one core whenever a block of transactions comes in and that coincides with the lag to respond to share submission, leading to more stale rejects if that occurs at the wrong time. The CPU in question is a 12 thread 3.4GHz CPU so it is clearly not underpowered. I suspect what happens on the Avalon is simply a grossly exaggerated form of this leading to misbehaviour.

the same thing happens with GPUs while using cgminer (not a cgminer problem, but stratum I would think)

i think i commented on that some months ago

minerd picks up the new work much quicker, and it is better to use phoenix for p2pool, or not use stratum

xhabit

full member

Activity: 209

Merit: 100

wow, very nice from bfl ...

iLike

forrestv

hero member

Activity: 516

Merit: 643

Quote from: maqifrnswa on April 28, 2013, 09:45:38 PM

Hey forrestv, do you want a Jalepeno? BFL_Josh will send you one if you'd like. You can find him here:
https://forums.butterflylabs.com/bfl-forum-miscellaneous/2164-demo-unit-forrestv-p2pool.html

I contacted him. Thanks for the tip!

TurdHurdur

full member

Activity: 216

Merit: 100

Quote from: -ck on April 28, 2013, 05:34:09 PM

So I tried the current version of p2pool on a single 5.7GHz Jalapeño with cgminer connecting via stratum. This works out to about 80 diff1 shares per minute. It "works" in a sense of the word, without causing any weird errors or duplicating work that happen with the Avalon but there are huge lags between share submission and responses from p2pool after a while. Watching the CPU usage, I see a spike to 100% of one core whenever a block of transactions comes in and that coincides with the lag to respond to share submission, leading to more stale rejects if that occurs at the wrong time. The CPU in question is a 12 thread 3.4GHz CPU so it is clearly not underpowered. I suspect what happens on the Avalon is simply a grossly exaggerated form of this leading to misbehaviour.

Does a username like ckolivas/2000+10 help?

maqifrnswa

sr. member

Activity: 454

Merit: 252

Quote from: -ck on April 28, 2013, 05:34:09 PM

So I tried the current version of p2pool on a single 5.7GHz Jalapeño with cgminer connecting via stratum. This works out to about 80 diff1 shares per minute. It "works" in a sense of the word, without causing any weird errors or duplicating work that happen with the Avalon but there are huge lags between share submission and responses from p2pool after a while. Watching the CPU usage, I see a spike to 100% of one core whenever a block of transactions comes in and that coincides with the lag to respond to share submission, leading to more stale rejects if that occurs at the wrong time. The CPU in question is a 12 thread 3.4GHz CPU so it is clearly not underpowered. I suspect what happens on the Avalon is simply a grossly exaggerated form of this leading to misbehaviour.

Hey forrestv, do you want a Jalepeno? BFL_Josh will send you one if you'd like. You can find him here:
https://forums.butterflylabs.com/bfl-forum-miscellaneous/2164-demo-unit-forrestv-p2pool.html

Topic: [1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool - page 558. (Read 2591920 times)