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

[Mr. Jinx]

OK, am I to presume that nobody on the p2pool thread knows what this means either? Surely someone must have knowledge of the p2pool workings on the p2pool thread.......   Any devs out there?  Why is everyone so quiet all of a sudden?

Good question, I don't know where the guru's are?
I'm having some memoryleak issue's with running a public p2pool. Would love to see some help to tackle this but no reactions.
I guess these kind of problems are hard to solve. So you'll have to wait patiently or try to fix it yourself.

[smoothrunnings]

OK, am I to presume that nobody on the p2pool thread knows what this means either? Surely someone must have knowledge of the p2pool workings on the p2pool thread.......   Any devs out there?  Why is everyone so quiet all of a sudden?

Good question, I don't know where the guru's are?
I'm having some memoryleak issue's with running a public p2pool. Would love to see some help to tackle this but no reactions.
I guess these kind of problems are hard to solve. So you'll have to wait patiently or try to fix it yourself.

If you are having leaky memory problems and can quantify it I recommend you post an issue at: https://github.com/forrestv/p2pool and see if Forrestv can fix the issue.

[KyrosKrane]

Yeah if ESXi was having a problem I would have been sent a PRTG alert and my other guests would also be complaining, but they aren't. The CPU on my P2Pool barely goes over 270Mhz even though it has 5.3Ghz to play with.

5.3GHz?! What the heck kind of CPU are you running on that monster? Or are you adding up the speeds of multiple cores/processors?

[bitpop]

Yeah if ESXi was having a problem I would have been sent a PRTG alert and my other guests would also be complaining, but they aren't. The CPU on my P2Pool barely goes over 270Mhz even though it has 5.3Ghz to play with.

5.3GHz?! What the heck kind of CPU are you running on that monster? Or are you adding up the speeds of multiple cores/processors?

My i7 3770 easily goes 4.3 ghz and that's not the 3770k

[smoothrunnings]

Yeah if ESXi was having a problem I would have been sent a PRTG alert and my other guests would also be complaining, but they aren't. The CPU on my P2Pool barely goes over 270Mhz even though it has 5.3Ghz to play with.

5.3GHz?! What the heck kind of CPU are you running on that monster? Or are you adding up the speeds of multiple cores/processors?

I have an IBM System x3400 M3 with two 6 core Xeon CPUs.

I am not adding anything, this is what ESXi tells me under the performance tab for my P2Pool server. If you must know I gave it 1 CPU 2 cores.

[smoothrunnings]

Hey guys I have having trouble with my bitcoin-qt on my P2Pool server. I started the transfer process of 0.5 BTC but it looks like its gotten stuck. It has been showing 0 confirms for the last couple of days. So I looked on Google and someone mentions running getrawtransaction and sendrawtransaction which helps get it moving accept I don't know how to do this in the linux world.

Any suggestions?

[smoothrunnings]

Has anyone ever seen this in their p2pool console?

2014-02-23 15:15:47.591287 Peer 192.168.75.xxx:42898 misbehaving, will drop and ban. Reason: was connected to self
2014-02-23 15:15:47.591393 Bad peer banned: ('192.168.75.xxx', 42898)
2014-02-23 15:15:47.592063 Peer 24.246.85.xxx:9333 misbehaving, will drop and ban. Reason: was connected to self
2014-02-23 15:15:47.592147 Bad peer banned: (u'24.246.85.xxx', 9333)

The 192.168.75.xxx was my LAN gateway and the 24.246.85.xxx was my internet IP.

[KyrosKrane]

I've been having problems with my p2pool.  Every hour or two, for some reason, it crashes.  I was able to capture the log from my latest crash.  Anyone know what's going on or how to fix it?  Thanks in advance.

Code:

2014-02-24 00:36:39.686737 P2Pool: 17661 shares in chain (17665 verified/17665 total) Peers: 7 (0 incoming)
2014-02-24 00:36:39.686859  Local: 8563MH/s in last 10.0 minutes Local dead on arrival: ~3.1% (2-5%) Expected time to share: 3.7 days
2014-02-24 00:36:39.686895  Shares: 0 (0 orphan, 0 dead) Stale rate: ??? Efficiency: ??? Current payout: 0.0000 BTC
2014-02-24 00:36:39.686934  Pool: 132TH/s Stale rate: 11.8% Expected time to block: 1.2 days
2014-02-24 00:37:24.954876 New work for worker! Difficulty: 1.697100 Share difficulty: 635654.548354 Total block value: 25.171297 BTC including 908 transactions
2014-02-24 00:37:25.361745 > Watchdog timer went off at:
2014-02-24 00:37:25.361880 >   File "/home/USER/p2pool/run_p2pool.py", line 5, in <module>
2014-02-24 00:37:25.361999 >     main.run()
2014-02-24 00:37:25.362028 >   File "/home/USER/p2pool/p2pool/main.py", line 578, in run
2014-02-24 00:37:25.362053 >     reactor.run()
2014-02-24 00:37:25.362078 >   File "/usr/lib/python2.7/dist-packages/twisted/internet/base.py", line 1169, in run
2014-02-24 00:37:25.362112 >     self.mainLoop()
2014-02-24 00:37:25.362137 >   File "/usr/lib/python2.7/dist-packages/twisted/internet/base.py", line 1181, in mainLoop
2014-02-24 00:37:25.362161 >     self.doIteration(t)
2014-02-24 00:37:25.362184 >   File "/usr/lib/python2.7/dist-packages/twisted/internet/pollreactor.py", line 167, in doPoll
2014-02-24 00:37:25.362208 >     log.callWithLogger(selectable, _drdw, selectable, fd, event)
2014-02-24 00:37:25.362232 >   File "/usr/lib/python2.7/dist-packages/twisted/python/log.py", line 84, in callWithLogger
2014-02-24 00:37:25.362256 >     return callWithContext({"system": lp}, func, *args, **kw)
2014-02-24 00:37:25.362279 >   File "/usr/lib/python2.7/dist-packages/twisted/python/log.py", line 69, in callWithContext
2014-02-24 00:37:25.362302 >     return context.call({ILogContext: newCtx}, func, *args, **kw)
2014-02-24 00:37:25.362326 >   File "/usr/lib/python2.7/dist-packages/twisted/python/context.py", line 118, in callWithContext
2014-02-24 00:37:25.362350 >     return self.currentContext().callWithContext(ctx, func, *args, **kw)
2014-02-24 00:37:25.374928 >   File "/usr/lib/python2.7/dist-packages/twisted/python/context.py", line 81, in callWithContext
2014-02-24 00:37:25.374963 >     return func(*args,**kw)
2014-02-24 00:37:25.374988 >   File "/usr/lib/python2.7/dist-packages/twisted/internet/posixbase.py", line 586, in _doReadOrWrite
2014-02-24 00:37:25.375010 >     why = selectable.doRead()
2014-02-24 00:37:25.375033 >   File "/usr/lib/python2.7/dist-packages/twisted/internet/tcp.py", line 199, in doRead
2014-02-24 00:37:25.434122 >     rval = self.protocol.dataReceived(data)
2014-02-24 00:37:25.434252 >   File "/home/USER/p2pool/p2pool/p2p.py", line 145, in new_dataReceived
2014-02-24 00:37:25.434315 >     old_dataReceived(data)
2014-02-24 00:37:25.434373 >   File "/home/USER/p2pool/p2pool/util/p2protocol.py", line 27, in dataReceived
2014-02-24 00:37:25.575921 >     self.dataReceived2(data)
2014-02-24 00:37:25.652144 >   File "/home/USER/p2pool/p2pool/util/datachunker.py", line 40, in _DataChunker
2014-02-24 00:37:25.652217 >     wants = receiver.send(buf.get(wants))
2014-02-24 00:37:25.652244 >   File "/home/USER/p2pool/p2pool/util/p2protocol.py", line 57, in dataReceiver
2014-02-24 00:37:25.652269 >     self.packetReceived(command, type_.unpack(payload, self.ignore_trailing_payload))
2014-02-24 00:37:25.652293 >   File "/home/USER/p2pool/p2pool/p2p.py", line 91, in packetReceived
2014-02-24 00:37:25.652328 >     p2protocol.Protocol.packetReceived(self, command, payload2)
2014-02-24 00:37:25.652352 >   File "/home/USER/p2pool/p2pool/util/p2protocol.py", line 71, in packetReceived
2014-02-24 00:37:25.652375 >     handler(**payload2)
2014-02-24 00:37:25.652407 >   File "/home/USER/p2pool/p2pool/p2p.py", line 282, in handle_shares
2014-02-24 00:37:25.652430 >     self.node.handle_shares(result, self)
2014-02-24 00:37:25.652451 >   File "/home/USER/p2pool/p2pool/node.py", line 48, in handle_shares
2014-02-24 00:37:25.652484 >     self.node.set_best_share()
2014-02-24 00:37:25.652506 >   File "/home/USER/p2pool/p2pool/node.py", line 295, in set_best_share
2014-02-24 00:37:25.652529 >     best, desired, decorated_heads, bad_peer_addresses = self.tracker.think(self.get_height_rel_highest, self.bitcoind_work.value['previous_block'], self.bitcoind_work.value['bits'], self.known_txs_var.value)
2014-02-24 00:37:25.652559 >   File "/home/USER/p2pool/p2pool/data.py", line 498, in think
2014-02-24 00:37:25.652583 >     decorated_tails = sorted((self.score(max(self.verified.tails[tail_hash], key=self.verified.get_work), block_rel_height_func), tail_hash) for tail_hash in self.verified.tails)
2014-02-24 00:37:25.652609 >   File "/home/USER/p2pool/p2pool/data.py", line 498, in <genexpr>
2014-02-24 00:37:25.652632 >     decorated_tails = sorted((self.score(max(self.verified.tails[tail_hash], key=self.verified.get_work), block_rel_height_func), tail_hash) for tail_hash in self.verified.tails)
2014-02-24 00:37:25.652656 >   File "/home/USER/p2pool/p2pool/data.py", line 548, in score
2014-02-24 00:37:25.652679 >     self.verified.get_chain(end_point, self.net.CHAIN_LENGTH//16))
2014-02-24 00:37:25.652701 >   File "/home/USER/p2pool/p2pool/data.py", line 547, in <genexpr>
2014-02-24 00:37:25.652724 >     block_height = max(block_rel_height_func(share.header['previous_block']) for share in
2014-02-24 00:37:25.652747 >   File "/home/USER/p2pool/p2pool/util/forest.py", line 333, in get_chain
2014-02-24 00:37:25.652770 >     assert length <= self.get_height(start_hash)
2014-02-24 00:37:25.652805 >   File "/home/USER/p2pool/p2pool/util/forest.py", line 163, in get_height
2014-02-24 00:37:25.652828 >     return self.get_delta_to_last(item_hash).height
2014-02-24 00:37:25.652850 >   File "/home/USER/p2pool/p2pool/util/forest.py", line 217, in get_delta_to_last
2014-02-24 00:37:25.652872 >     self._set_delta(update_hash, delta - delta_then)
2014-02-24 00:37:25.652894 >   File "/home/USER/p2pool/p2pool/util/forest.py", line 205, in _set_delta
2014-02-24 00:37:25.652916 >     self._deltas[item_hash] = delta - ref_delta, ref
2014-02-24 00:37:25.652948 >   File "/home/USER/p2pool/p2pool/main.py", line 245, in <lambda>
2014-02-24 00:37:25.652972 >     sys.stderr.write, 'Watchdog timer went off at:\n' + ''.join(traceback.format_stack())

[IYFTech]

OK, am I to presume that nobody on the p2pool thread knows what this means either? Surely someone must have knowledge of the p2pool workings on the p2pool thread.......   Any devs out there?  Why is everyone so quiet all of a sudden?

Good question, I don't know where the guru's are?
I'm having some memoryleak issue's with running a public p2pool. Would love to see some help to tackle this but no reactions.
I guess these kind of problems are hard to solve. So you'll have to wait patiently or try to fix it yourself.

Well, seems I'm not the only one here who can't get any answers to p2pool questions. I'd fix it myself if I knew what the problem was of course, but I don't, that's why I came here. Seems p2pool has become abandonware.....I'll try elsewhere instead then. Hope you all have better luck than me in getting some help for your problems 

I think I know now why p2pool is not attracting new users though........

[forrestv]

OK, am I to presume that nobody on the p2pool thread knows what this means either? Surely someone must have knowledge of the p2pool workings on the p2pool thread.......   Any devs out there?  Why is everyone so quiet all of a sudden?

Good question, I don't know where the guru's are?
I'm having some memoryleak issue's with running a public p2pool. Would love to see some help to tackle this but no reactions.
I guess these kind of problems are hard to solve. So you'll have to wait patiently or try to fix it yourself.

Well, seems I'm not the only one here who can't get any answers to p2pool questions. I'd fix it myself if I knew what the problem was of course, but I don't, that's why I came here. Seems p2pool has become abandonware.....I'll try elsewhere instead then. Hope you all have better luck than me in getting some help for your problems 

I think I know now why p2pool is not attracting new users though........

IYFTech, your

Code:

"/usr/lib/python2.7/dist-packages/zope/__init__.py:3: UserWarning: Module twisted was already imported from /usr/lib/python2.7/dist-packages/twisted/__init__.pyc, but /usr/local/lib/python2.7/dist-packages/autobahn-0.7.4-py2.7.egg is being added to sys.path"

error is from some random other package on your system that is misconfigured (it looks like it's "autobahn"). If P2Pool is running fine, why do you care so much? Try removing it to get rid of the error.

[forrestv]

OK, am I to presume that nobody on the p2pool thread knows what this means either? Surely someone must have knowledge of the p2pool workings on the p2pool thread.......   Any devs out there?  Why is everyone so quiet all of a sudden?

Good question, I don't know where the guru's are?
I'm having some memoryleak issue's with running a public p2pool. Would love to see some help to tackle this but no reactions.
I guess these kind of problems are hard to solve. So you'll have to wait patiently or try to fix it yourself.

Mr. Jinx, occasional memory leaks have plagued P2Pool nodes for a while. Python/Twisted makes it pretty easy to leave a reference to something or leave a task running in the background... several of them have been patched, though.

Try looking through P2Pool's log file for lines that contain a ">" character, which indicates an uncaught exception/traceback. A task dying that periodically removes old shares was the cause of a previous memory leak; maybe this is something similar.

[IYFTech]

OK, am I to presume that nobody on the p2pool thread knows what this means either? Surely someone must have knowledge of the p2pool workings on the p2pool thread.......  Any devs out there?  Why is everyone so quiet all of a sudden?

Good question, I don't know where the guru's are?
I'm having some memoryleak issue's with running a public p2pool. Would love to see some help to tackle this but no reactions.
I guess these kind of problems are hard to solve. So you'll have to wait patiently or try to fix it yourself.

Well, seems I'm not the only one here who can't get any answers to p2pool questions. I'd fix it myself if I knew what the problem was of course, but I don't, that's why I came here. Seems p2pool has become abandonware.....I'll try elsewhere instead then. Hope you all have better luck than me in getting some help for your problems  

I think I know now why p2pool is not attracting new users though........

IYFTech, your

Code:

"/usr/lib/python2.7/dist-packages/zope/__init__.py:3: UserWarning: Module twisted was already imported from /usr/lib/python2.7/dist-packages/twisted/__init__.pyc, but /usr/local/lib/python2.7/dist-packages/autobahn-0.7.4-py2.7.egg is being added to sys.path"

error is from some random other package on your system that is misconfigured (it looks like it's "autobahn"). If P2Pool is running fine, why do you care so much? Try removing it to get rid of the error.

AH!! Hello forrestv,

Well, I said it "seems" to be working fine, but I wasn't sure - and as it had only just started happening I wanted to make sure, so thanks anyway. I presume I should remove it with apt-get then autoremove?

Thanks.

EDIT: It's not that, tried sudo apt-get remove python-autobahn & it's not installed apparently - any other suggestions would be appreciated......

[IYFTech]

OK, I've noticed that autobahn-0.7.4-py2.7.egg is in the same path as slush's stratum proxy, so I'm presuming it was installed when I built that from git - which makes sense as that is what I last installed as I was trying to get it to work with p2pool (failing, of course). Do you think it is causing conflicts with p2pool? I don't want to uninstall the stratum proxy as I use it......maybe that's why it doesn't work with p2pool? I dunno.......

Thanks.

[roy7]

It's just a warning, not an error. And importing a library twice shouldn't hurt anything.

[phillipsjk]

Hey guys, what bad things come from having a high getwork latency? I'm under the impression that stratum makes getwork irrelevant, but I'm a bit worried that mine is quite high (sometimes a few seconds, you can see it here http://www.blisterpool.com/stats in the graphs bit). Is this a symptom of anything in particular? The cpu doesn't seem to be under any enormous stress, but it does peak every now and then...is this the cause?

Also, I helped a miner get configured to run on my p2pool node, and he has some ASICMiner blades (~10.7GH each). Google search results (most seem to be from a year ago) have all told me that they naturally have high DOA rate with p2pool, without any real way to fix it. Is this still the case? His dead rate is around 40-50%. I also noticed the server was getting absolutely hammered with hash > target spam, and I suggested to the miner to use his bitcoin address+1 for his username. It reduced the server spam drastically (from 100/sec to several/sec), and reduced his DOA a little bit, but it also reduced his mean hashing power by about 10%. Could anyone explain to me what's going on here? I'd like to help him get better results.

Is this miner running a stratum proxy locally? each of the 32 chips seem to mostly ask for work independently.
My blade gets work 150 times per minute, meaning that the latency should be at most 12.8 seconds (150/32*60).

I don't actually have P2Pool working yet, but moved the midstate calculations to my Stratum proxy on the assumption it can calculate mid-state faster than the blade.

Edit: Since AFAIK, the getwork protocol does not allow workers to be interrupted with new work, we should be able to estimate the expected stales given a specific block frequency. If we assume a 13s worst case latency, that works out to at most 43% stales with a 30 second target. If we assume a 6.5s average, that works out to 21.7% stale. -- that does seem high.

Edit: Apparently Longpolling works around HTTP limitations by having the miner request new work immediately. The server then does not respond until new work is ready. Testing time.

[IYFTech]

Hey guys, what bad things come from having a high getwork latency? I'm under the impression that stratum makes getwork irrelevant, but I'm a bit worried that mine is quite high (sometimes a few seconds, you can see it here http://www.blisterpool.com/stats in the graphs bit). Is this a symptom of anything in particular? The cpu doesn't seem to be under any enormous stress, but it does peak every now and then...is this the cause?

Also, I helped a miner get configured to run on my p2pool node, and he has some ASICMiner blades (~10.7GH each). Google search results (most seem to be from a year ago) have all told me that they naturally have high DOA rate with p2pool, without any real way to fix it. Is this still the case? His dead rate is around 40-50%. I also noticed the server was getting absolutely hammered with hash > target spam, and I suggested to the miner to use his bitcoin address+1 for his username. It reduced the server spam drastically (from 100/sec to several/sec), and reduced his DOA a little bit, but it also reduced his mean hashing power by about 10%. Could anyone explain to me what's going on here? I'd like to help him get better results.

Is this miner running a stratum proxy locally? each of the 32 chips seem to mostly ask for work independently.
My blade gets work 150 times per minute, meaning that the latency should be at most 12.8 seconds (150/32*60).

I don't actually have P2Pool working yet, but moved the midstate calculations to my Stratum proxy on the assumption it can calculate mid-state faster than the blade.

See this:

https://bitcointalk.org/index.php?topic=18313.msg4556966#msg4556966

I don't know of anyone who has got slush's proxy to work with p2pool, I've tried countless times myself - it just doesn't work, in fact I have a feeling that that's what is now causing my p2pool start up error/warning. Your only choice with a blade is to use +1 at the end of your user name/addy which, tbh, is a bit of a waste.....

[forrestv]

Hey guys, what bad things come from having a high getwork latency? I'm under the impression that stratum makes getwork irrelevant, but I'm a bit worried that mine is quite high (sometimes a few seconds, you can see it here http://www.blisterpool.com/stats in the graphs bit). Is this a symptom of anything in particular? The cpu doesn't seem to be under any enormous stress, but it does peak every now and then...is this the cause?

Also, I helped a miner get configured to run on my p2pool node, and he has some ASICMiner blades (~10.7GH each). Google search results (most seem to be from a year ago) have all told me that they naturally have high DOA rate with p2pool, without any real way to fix it. Is this still the case? His dead rate is around 40-50%. I also noticed the server was getting absolutely hammered with hash > target spam, and I suggested to the miner to use his bitcoin address+1 for his username. It reduced the server spam drastically (from 100/sec to several/sec), and reduced his DOA a little bit, but it also reduced his mean hashing power by about 10%. Could anyone explain to me what's going on here? I'd like to help him get better results.

Is this miner running a stratum proxy locally? each of the 32 chips seem to mostly ask for work independently.
My blade gets work 150 times per minute, meaning that the latency should be at most 12.8 seconds (150/32*60).

I don't actually have P2Pool working yet, but moved the midstate calculations to my Stratum proxy on the assumption it can calculate mid-state faster than the blade.

See this:

https://bitcointalk.org/index.php?topic=18313.msg4556966#msg4556966

I don't know of anyone who has got slush's proxy to work with p2pool, I've tried countless times myself - it just doesn't work, in fact I have a feeling that that's what is now causing my p2pool start up error/warning. Your only choice with a blade is to use +1 at the end of your user name/addy which, tbh, is a bit of a waste.....

The ASICMiner blades only do getwork requests, which already contain a midstate that P2Pool generated. Using the Stratum proxy has no or very little advantage over connecting directly to P2Pool. (As far as I know; don't have hardware to test with.)

[smooth]

The ASICMiner blades only do getwork requests, which already contain a midstate that P2Pool generated. Using the Stratum proxy has no or very little advantage over connecting directly to P2Pool. (As far as I know; don't have hardware to test with.)

The only advantage would be bandwidth if you are using a p2pool public node.  But this doesn't seem to work very well anyway.

I haven't tried blades with a local p2pool node and I've now sold them off so I can't.

[roy7]

I don't know of anyone who has got slush's proxy to work with p2pool, I've tried countless times myself - it just doesn't work, in fact I have a feeling that that's what is now causing my p2pool start up error/warning. Your only choice with a blade is to use +1 at the end of your user name/addy which, tbh, is a bit of a waste.....

Provided you are running your own node and not flooding network traffic to a public node, what's wrong with using the +1 solution?

[Mr. Jinx]

Mr. Jinx, occasional memory leaks have plagued P2Pool nodes for a while. Python/Twisted makes it pretty easy to leave a reference to something or leave a task running in the background... several of them have been patched, though.

Try looking through P2Pool's log file for lines that contain a ">" character, which indicates an uncaught exception/traceback. A task dying that periodically removes old shares was the cause of a previous memory leak; maybe this is something similar.

Ah, thanks. That makes it easier to troubleshoot!

This one happens alot. Always the same peers that gives this error (2).
What could be causing this?

Code:

2014-02-24 05:00:39.106919 > in handle_share_hashes:
2014-02-24 05:00:39.107328 > Traceback (most recent call last):
2014-02-24 05:00:39.107653 > Failure: twisted.internet.defer.TimeoutError: in GenericDeferrer
2014-02-24 05:00:39.109908 Lost peer x.x.x.x:36098 -
2014-02-24 05:00:39.110112     Connection was aborted locally, using
2014-02-24 05:00:39.110291     L{twisted.internet.interfaces.ITCPTransport.abortConnection}.
2014-02-24 05:00:39.110502
2014-02-24 05:00:39.110721     @since: 11.1
2014-02-24 05:00:39.110918     .[

This one a few times, but I don't think this is cause memory problems:

Code:

2014-02-24 05:05:18.137532 Worker xxxxxxxxxxx submitted share with hash > target:
2014-02-24 05:05:18.137946     Hash:   xxxxxxxxxxxxxxx
2014-02-24 05:05:18.138166     Target: xxxxxxxxxx

And this one a few times, while the pool hasn't been restarted for days:

Code:

2014-02-24 05:34:43.706844 > Couldn't link returned work's job id with its handler. This should only happen if this process was recently restarted!

Anyone who can shed some light on this?