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

[mdude77]

Correct.  You can not mix p2pool and non-p2pool entries in the same cgminer instance.  

P2Pool, the distributed network itself, is essentially never "down".  If you want a backup to your personal p2pool node being down, use someone else's public p2pool node as a backup (there is a thread here somewhere with a list of public p2pool servers) and specify your bitcoin address as the username.

Um.. I have p2pool and ozcoin in my config, and I switch from one to the other without a problem (usually manually, as both have always been up as long as I've been using both).  Yeah, a get a few rejects before it synchs up, but it always works. 

Maybe you mean something other than "can not mix".  Like load balancing maybe?

M

[kano]

Correct.  You can not mix p2pool and non-p2pool entries in the same cgminer instance.  

P2Pool, the distributed network itself, is essentially never "down".  If you want a backup to your personal p2pool node being down, use someone else's public p2pool node as a backup (there is a thread here somewhere with a list of public p2pool servers) and specify your bitcoin address as the username.

Um.. I have p2pool and ozcoin in my config, and I switch from one to the other without a problem (usually manually, as both have always been up as long as I've been using both).  Yeah, a get a few rejects before it synchs up, but it always works.  

Maybe you mean something other than "can not mix".  Like load balancing maybe?

M

You must ONLY be mining normal BTC or P2Pool at the same time.
Thus any sort of load balancing must only be the same type of pool as the "Priority 0" pool
You also need to use --failover-only or disable all pools that could be 'failed over' to, that are a different type.

Switching may or may not work as expected - YMMV

But think of it in terms of:
At any time, you MUST disable getting work from ALL pools that are not the same as the "Priority 0" pool

[Ed]

Um.. I have p2pool and ozcoin in my config, and I switch from one to the other without a problem (usually manually, as both have always been up as long as I've been using both).  Yeah, a get a few rejects before it synchs up, but it always works......

changed backup to mtred and problems gone, booth remote sites

thanks all

[Syke]

I can't conceive of a scenario where a scaling problem could be responsible for bad luck.  Scaling problems causing more stale shares or more orphaned blocks?  Sure.  Scaling problems making math stop working?  How, exactly could the size of the network have an impact on the likelihood that random numbers are below a defined target?

The only thing that would affect scaling would be some large miners purposefully mining on p2pool and witholding the blocks. That would be very costly to do so, but it is possible.

[Soros Shorts]

Are BFL FPGA Singles the only type mining hardware that is currently not compatible with p2pool?

When I upgrade mine to "SC" Singles (hopefully in 2 months) I am expecting these to work with p2pool. Is that a reasonable expectation?

[rav3n_pl]

Are BFL FPGA Singles the only type mining hardware that is currently not compatible with p2pool?

When I upgrade mine to "SC" Singles (hopefully in 2 months) I am expecting these to work with p2pool. Is that a reasonable expectation?

Tell us when you will have them All this things are still a question mark even for BFL

[Aseras]

V5 still seems to have problems with memory leaks, seems more to be an issue with how it talks to .7 bitcoin qt as when it happens, both start to consume HUGE amounts of memory. run_p2pool will start blasting lost connection messages or scrolling this message constantly

exceptions.ValueError: too many file descriptors in select()
2012-09-28 09:44:04.312000 > Unhandled Error
2012-09-28 09:44:04.312000 > Traceback (most recent call last):
2012-09-28 09:44:04.312000 >   File "run_p2pool.py", line 5, in <module>
2012-09-28 09:44:04.312000 >     
2012-09-28 09:44:04.312000 >   File "p2pool\main.pyc", line 781, in run
2012-09-28 09:44:04.312000 >     
2012-09-28 09:44:04.312000 >   File "twisted\internet\base.pyc", line 1162, in run
2012-09-28 09:44:04.312000 >     
2012-09-28 09:44:04.312000 >   File "twisted\internet\base.pyc", line 1174, in mainLoop
2012-09-28 09:44:04.312000 >     
2012-09-28 09:44:04.312000 > --- <exception caught here> ---
2012-09-28 09:44:04.312000 >   File "twisted\internet\selectreactor.pyc", line 104, in doSelect
2012-09-28 09:44:04.312000 >     
2012-09-28 09:44:04.312000 >   File "twisted\internet\selectreactor.pyc", line 40, in win32select
2012-09-28 09:44:04.312000 >     
2012-09-28 09:44:04.312000 > exceptions.ValueError: too many file descriptors in select()

eventually it'll use up all available ram and blow up.

[jiyu_shi]

I run p2pool, it returns

what's the red box means?

Thus, the miner works at normal speed, but accept is 0, p2pool's Local is also 0,
and there's no show at p2pool.info_active users
bitcoin-qt's datadir is at D:\bitcoin\

how to solve the problem?
thanks

[forrestv]

I run p2pool, it returns
...
what's the red box means?

Thus, the miner works at normal speed, but accept is 0, p2pool's Local is also 0,
and there's no show at p2pool.info_active users
bitcoin-qt's datadir is at D:\bitcoin\

how to solve the problem?
thanks

The red box isn't unusual.. It just means that querying bitcoin for something timed out, but unless it's happening all the time, it's harmless.

Local may be displayed as 0 if you're mining to a non-default address, because you've set your miner's username to a Bitcoin address. This isn't recommended - instead you should use P2Pool's -a option. If you need to use the username way, you can look at the graphs on http://127.0.0.1:9332/ to get some information.

[rav3n_pl]

V5 still seems to have problems with memory leaks, seems more to be an issue with how it talks to .7 bitcoin qt as when it happens, both start to consume HUGE amounts of memory. run_p2pool will start blasting lost connection messages or scrolling this message constantly

eventually it'll use up all available ram and blow up.

Version: 5.0-3-g7585a1c
Node uptime: 3.639 days Peers: 10 out, 1 in
Not see any excessive memory usage.

[forrestv]

V5 still seems to have problems with memory leaks, seems more to be an issue with how it talks to .7 bitcoin qt as when it happens, both start to consume HUGE amounts of memory. run_p2pool will start blasting lost connection messages or scrolling this message constantly

eventually it'll use up all available ram and blow up.

Aseras, running p2pool with the --debug option and sending me the log might help. It looks like it's leaking sockets for some reason..

[jiyu_shi]

I run p2pool, it returns
...
what's the red box means?

Thus, the miner works at normal speed, but accept is 0, p2pool's Local is also 0,
and there's no show at p2pool.info_active users
bitcoin-qt's datadir is at D:\bitcoin\

how to solve the problem?
thanks

The red box isn't unusual.. It just means that querying bitcoin for something timed out, but unless it's happening all the time, it's harmless.

Local may be displayed as 0 if you're mining to a non-default address, because you've set your miner's username to a Bitcoin address. This isn't recommended - instead you should use P2Pool's -a option. If you need to use the username way, you can look at the graphs on http://127.0.0.1:9332/ to get some information.

I'd already use -a [address], and miner's username was a norml word, not an address.
Another, my GPU is HD6470M which has only ~30Mh/s speed, is it too slowly to get an accept?
http://127.0.0.1:9332/ shows nothing--even no pool speed and other users' speed, but the graphs shows my speed and pool speed.

[Mobius]

I run p2pool, it returns
...
what's the red box means?

Thus, the miner works at normal speed, but accept is 0, p2pool's Local is also 0,
and there's no show at p2pool.info_active users
bitcoin-qt's datadir is at D:\bitcoin\

how to solve the problem?
thanks

The red box isn't unusual.. It just means that querying bitcoin for something timed out, but unless it's happening all the time, it's harmless.

Local may be displayed as 0 if you're mining to a non-default address, because you've set your miner's username to a Bitcoin address. This isn't recommended - instead you should use P2Pool's -a option. If you need to use the username way, you can look at the graphs on http://127.0.0.1:9332/ to get some information.

I'd already use -a [address], and miner's username was a norml word, not an address.
Another, my GPU is HD6470M which has only ~30Mh/s speed, is it too slowly to get an accept?
http://127.0.0.1:9332/ shows nothing--even no pool speed and other users' speed, but the graphs shows my speed and pool speed.

use http://127.0.0.1:9332/static/

[jiyu_shi]

it is a notebook with HD6470M, just works for test.
miner and p2pool seems work normally, but the accept is only 9 which has worked for 20 minutes.
CG option: -g 1 -v 2 -w 128 -I 6


and static shows 0 shares and 0 DOA


so why? is the GPU too poor to fix p2pool?

[Krak]

it is a notebook with HD6470M, just works for test.
miner and p2pool seems work normally, but the accept is only 9 which has worked for 20 minutes.
CG option: -g 1 -v 2 -w 128 -I 6
http://img.aalun.com/di/PRR5/1.png

and static shows 0 shares and 0 DOA
http://img.aalun.com/di/AAAP/2.png

so why? is the GPU too poor to fix p2pool?

As it says in the p2pool window, you'll only get a share every ~23 hours. The shares that are showing as accepted in cgminer are the diff-1 pseudo shares that p2pool uses to track your hashrate.

[jiyu_shi]

it is a notebook with HD6470M, just works for test.
miner and p2pool seems work normally, but the accept is only 9 which has worked for 20 minutes.
CG option: -g 1 -v 2 -w 128 -I 6
http://img.aalun.com/di/PRR5/1.png

and static shows 0 shares and 0 DOA
http://img.aalun.com/di/AAAP/2.png

so why? is the GPU too poor to fix p2pool?

As it says in the p2pool window, you'll only get a share every ~23 hours. The shares that are showing as accepted in cgminer are the diff-1 pseudo shares that p2pool uses to track your hashrate.

so, it means the HD6470M had not completed just only 1 true share for 20 minutes? if it works more than 23 hours, I will get gain?

[Krak]

so, it means the HD6470M had not completed just only 1 true share for 20 minutes? if it works more than 23 hours, I will get gain?

More or less, yeah.

[rav3n_pl]

so, it means the HD6470M had not completed just only 1 true share for 20 minutes? if it works more than 23 hours, I will get gain?

More or less, yeah.

Miner is reporting diff1 shares, p2pool need much higher (diff >500) share to pay.
If p2pool tells you that "expected time to share" is 23hrs it is possible to NOT get payd all day long, even two when mining 24/7. 35MH is just not good enough.

[mdude77]

Why am I seeing this repeatedly in my console? 

Code:

2012-10-01 20:31:14.959000 Sending 0 shares to 92.251.166.251:58565

Not "all the time", but I see it regularly.  Here's another set:

Code:

2012-10-01 20:32:32.820000  Pool: 374GH/s Stale rate: 9.3% Expected time to block: 9.1 hours
2012-10-01 20:32:34.529000 Sending 0 shares to 92.251.166.251:58565
2012-10-01 20:32:41.169000 Sending 0 shares to 92.251.166.251:58565
2012-10-01 20:32:43.630000 Sending 0 shares to 92.251.166.251:58565

M

[forrestv]

Why am I seeing this repeatedly in my console? 

Code:

2012-10-01 20:31:14.959000 Sending 0 shares to 92.251.166.251:58565

It's just a peer requesting a share you don't have, potentially on purpose to measure ping times: http://forre.st:9332/pings