Bitcoin Forum
December 04, 2016, 04:33:43 PM *
News: Latest stable version of Bitcoin Core: 0.13.1  [Torrent].
 
   Home   Help Search Donate Login Register  
Pages: « 1 ... 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 [363] 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 ... 744 »
  Print  
Author Topic: [1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool  (Read 2029499 times)
kano
Legendary
*
Offline Offline

Activity: 1918


Linux since 1997 RedHat 4


View Profile
December 20, 2013, 02:48:57 AM
 #7241

.......
The well known bitcoin term here is called ... variance. Read up about it Smiley

Yeah, I know.
No.

You are missing the issue of variance.

If the difficulty drops it will also go up. It averages out to the same share difficulty.

So statistically you still get the same number of shares over time.

kano, I think TierNolan is referring to the fact that some large miners can set their own difficulty much higher than the automatic amount set by the system. If I have that fact right, then this means that the average share difficulty is reduced for the remainder of miners.
Unless there's some weird calculation in the p2pool code to push the base difficulty down, it doesn't matter if someone is submitting 'double' diff shares - they are just the same as submitting 2 'single' diff shares other than an increase in variance.

Pool: https://kano.is BTC: 1KanoiBupPiZfkwqB7rfLXAzPnoTshAVmb
CKPool and CGMiner developer, IRC FreeNode #ckpool and #cgminer kanoi
Help keep Bitcoin secure by mining on pools with Stratum, the best protocol to mine Bitcoins with ASIC hardware
1480869223
Hero Member
*
Offline Offline

Posts: 1480869223

View Profile Personal Message (Offline)

Ignore
1480869223
Reply with quote  #2

1480869223
Report to moderator
1480869223
Hero Member
*
Offline Offline

Posts: 1480869223

View Profile Personal Message (Offline)

Ignore
1480869223
Reply with quote  #2

1480869223
Report to moderator
1480869223
Hero Member
*
Offline Offline

Posts: 1480869223

View Profile Personal Message (Offline)

Ignore
1480869223
Reply with quote  #2

1480869223
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1480869223
Hero Member
*
Offline Offline

Posts: 1480869223

View Profile Personal Message (Offline)

Ignore
1480869223
Reply with quote  #2

1480869223
Report to moderator
1480869223
Hero Member
*
Offline Offline

Posts: 1480869223

View Profile Personal Message (Offline)

Ignore
1480869223
Reply with quote  #2

1480869223
Report to moderator
kjj
Legendary
*
Offline Offline

Activity: 1302



View Profile
December 20, 2013, 02:58:42 AM
 #7242

Unless there's some weird calculation in the p2pool code to push the base difficulty down

There is.

p2pcoin: a USB/CD/PXE p2pool miner - 1N8ZXx2cuMzqBYSK72X4DAy1UdDbZQNPLf - todo
I routinely ignore posters with paid advertising in their sigs.  You should too.
lenny_
Legendary
*
Offline Offline

Activity: 953



View Profile
December 20, 2013, 08:45:59 AM
 #7243

Unless there's some weird calculation in the p2pool code to push the base difficulty down

There is.

There is, but it doesnt' work very well. Here is my proposal to implement alternative function:
https://bitcointalk.org/index.php?topic=18313.msg3929150#msg3929150
TierNolan
Legendary
*
Offline Offline

Activity: 1036


View Profile
December 20, 2013, 12:05:32 PM
 #7244

There is, but it doesnt' work very well. Here is my proposal to implement alternative function:
https://bitcointalk.org/index.php?topic=18313.msg3929150#msg3929150

Each worker sets its own difficulty right?

So,

- adjust minimum difficulty per miner so that shares are received at most once every 60 minutes
- all workers to the same address count as a single miner for these purposes

I think there is at least one p2pool backed pool out there that supports smaller miners.  However, it works by setting the fee to 100% and has a separate process to handle payouts to miners.

1LxbG5cKXzTwZg9mjL3gaRE835uNQEteWF
kano
Legendary
*
Offline Offline

Activity: 1918


Linux since 1997 RedHat 4


View Profile
December 20, 2013, 12:59:29 PM
 #7245

Unless there's some weird calculation in the p2pool code to push the base difficulty down

There is.
Well if there is than the statement on the main p2pool page is wrong also:

Quote
P2Pool creates a new block chain in which the difficulty is adjusted so a new block is found every 1030 seconds.

I of course updated the 10 to 30.

That statement would be false if what people are implying here is true.

To make a share average every 30s, it must use the submitted 1diff share rate to attempt to make that happen.
There is no other way to do it.

So again ...

Quote
If the difficulty drops it will also go up. It averages out to the same share difficulty.

So statistically you still get the same number of shares over time.

Pool: https://kano.is BTC: 1KanoiBupPiZfkwqB7rfLXAzPnoTshAVmb
CKPool and CGMiner developer, IRC FreeNode #ckpool and #cgminer kanoi
Help keep Bitcoin secure by mining on pools with Stratum, the best protocol to mine Bitcoins with ASIC hardware
twmz
Hero Member
*****
Offline Offline

Activity: 737



View Profile
December 20, 2013, 01:41:37 PM
 #7246

Unless there's some weird calculation in the p2pool code to push the base difficulty down

There is.
Well if there is than the statement on the main p2pool page is wrong also:

Quote
P2Pool creates a new block chain in which the difficulty is adjusted so a new block is found every 1030 seconds.

I of course updated the 10 to 30.

That statement would be false if what people are implying here is true.

To make a share average every 30s, it must use the submitted 1diff share rate to attempt to make that happen.
There is no other way to do it.

So again ...

Quote
If the difficulty drops it will also go up. It averages out to the same share difficulty.

So statistically you still get the same number of shares over time.

It's really not that complicated.  p2pool adjusts the minimum required share difficulty for shares so that a share will be found once every 30 seconds.  Now, if a bunch of people start to opt-in to a share difficulty much higher minimum requirement (doing so is a feature baked into p2pool), then obviously shares are going to start to be found less often.  In order for shares to continue to be found about once per 30 seconds, everyone else has to start finding them more often.  The only way they can find shares more often (with the same hashrate) is if p2pool lowers the minimum required share difficulty, and that's exactly what it does.

Was I helpful?  1TwmzX1wBxNF2qtAJRhdKmi2WyLZ5VHRs
WoT, GPG

Bitrated user: ewal.
TierNolan
Legendary
*
Offline Offline

Activity: 1036


View Profile
December 20, 2013, 02:49:04 PM
 #7247

The only way they can find shares more often (with the same hashrate) is if p2pool lowers the minimum required share difficulty, and that's exactly what it does.

Right, the key is that the pool can set minimum difficulty.  Rather than adjusting difficulty for all miners, it adjusts difficulty for miners who have their difficulty set to the lowest possible.

Another way would be to adjust all miner's difficulty.  A miner might set their difficulty to 100X the default value.  The pool would then adjust the default value.

1LxbG5cKXzTwZg9mjL3gaRE835uNQEteWF
crunchy
Newbie
*
Offline Offline

Activity: 19


View Profile
December 20, 2013, 05:28:56 PM
 #7248

so whats with this liteco.in not found messages at startup ??

i've seen whats happened with liteco.in. very questionable all.

so why has that adress been hardcoded into p2pool ??
kano
Legendary
*
Offline Offline

Activity: 1918


Linux since 1997 RedHat 4


View Profile
December 20, 2013, 10:21:40 PM
 #7249

...
It's really not that complicated.  p2pool adjusts the minimum required share difficulty for shares so that a share will be found once every 30 seconds.  Now, if a bunch of people start to opt-in to a share difficulty much higher minimum requirement (doing so is a feature baked into p2pool), then obviously shares are going to start to be found less often.  In order for shares to continue to be found about once per 30 seconds, everyone else has to start finding them more often.  The only way they can find shares more often (with the same hashrate) is if p2pool lowers the minimum required share difficulty, and that's exactly what it does.
That misses Stats 101

Since the shares are of higher value, then when one is submitted it will mean the hash rate will appear higher.

So back again to what I said ...

...
No.

You are missing the issue of variance.

If the difficulty drops it will also go up. It averages out to the same share difficulty.

So statistically you still get the same number of shares over time.

...
Unless there's some weird calculation in the p2pool code to push the base difficulty down, it doesn't matter if someone is submitting 'double' diff shares - they are just the same as submitting 2 'single' diff shares other than an increase in variance.
i.e. no 'weird' calculation.

If you (on average) submit a double diff share every share you submit, but do it half as often, then the average share difficulty you submit will still be the same.
I'll try a little simple maths: 2 + 0 = 2 ... average ... 1 Tongue

Since the highest user is only around 11% of the pool, the variance it causes is also pretty small.

Pool: https://kano.is BTC: 1KanoiBupPiZfkwqB7rfLXAzPnoTshAVmb
CKPool and CGMiner developer, IRC FreeNode #ckpool and #cgminer kanoi
Help keep Bitcoin secure by mining on pools with Stratum, the best protocol to mine Bitcoins with ASIC hardware
kjj
Legendary
*
Offline Offline

Activity: 1302



View Profile
December 20, 2013, 11:09:12 PM
 #7250

...
It's really not that complicated.  p2pool adjusts the minimum required share difficulty for shares so that a share will be found once every 30 seconds.  Now, if a bunch of people start to opt-in to a share difficulty much higher minimum requirement (doing so is a feature baked into p2pool), then obviously shares are going to start to be found less often.  In order for shares to continue to be found about once per 30 seconds, everyone else has to start finding them more often.  The only way they can find shares more often (with the same hashrate) is if p2pool lowers the minimum required share difficulty, and that's exactly what it does.
That misses Stats 101

Since the shares are of higher value, then when one is submitted it will mean the hash rate will appear higher.

So back again to what I said ...

...
No.

You are missing the issue of variance.

If the difficulty drops it will also go up. It averages out to the same share difficulty.

So statistically you still get the same number of shares over time.

...
Unless there's some weird calculation in the p2pool code to push the base difficulty down, it doesn't matter if someone is submitting 'double' diff shares - they are just the same as submitting 2 'single' diff shares other than an increase in variance.
i.e. no 'weird' calculation.

If you (on average) submit a double diff share every share you submit, but do it half as often, then the average share difficulty you submit will still be the same.
I'll try a little simple maths: 2 + 0 = 2 ... average ... 1 Tongue

Since the highest user is only around 11% of the pool, the variance it causes is also pretty small.

Are you being dense on purpose, or do you really not get it?

The network targets a share rate, not a hash rate, not a difficulty rate, and not a (share times value) rate.  It adjusts so that the correct number of shares is found over time, which shares can be of any difficulty equal to or greater than the minimum.  Shares of higher than the minimum difficulty are worth more, but still only take up one "slot" in the calculation.

p2pcoin: a USB/CD/PXE p2pool miner - 1N8ZXx2cuMzqBYSK72X4DAy1UdDbZQNPLf - todo
I routinely ignore posters with paid advertising in their sigs.  You should too.
twmz
Hero Member
*****
Offline Offline

Activity: 737



View Profile
December 21, 2013, 01:25:56 AM
 #7251

That misses Stats 101

You're welcome to believe what you want, but I'm done interacting with you.  It's an observable fact that if high hashrate miners opt in to a higher share difficulty/target, the minimum share difficulty/target will come down for everyone else. Think of it as "magic" if that makes you more comfortable.

Was I helpful?  1TwmzX1wBxNF2qtAJRhdKmi2WyLZ5VHRs
WoT, GPG

Bitrated user: ewal.
powersync
Member
**
Offline Offline

Activity: 74


View Profile
December 21, 2013, 04:34:49 AM
 #7252

hey guys,

I just started my ltc p2pool server again after taking a break for 6 months.  I upgraded to 13.4-4 and seeing my memory usage creep up to 1gb. I restart P2pool and all is fine. I'm running Python 2.7.2+ on Ubuntu Linux server 11.10.  I cant find anyone lately having the same problem.  Where should I start to solve this?

PS
Raulnsh
Newbie
*
Offline Offline

Activity: 22


View Profile
December 21, 2013, 10:54:35 AM
 #7253

hey guys,

I just started my ltc p2pool server again after taking a break for 6 months.  I upgraded to 13.4-4 and seeing my memory usage creep up to 1gb. I restart P2pool and all is fine. I'm running Python 2.7.2+ on Ubuntu Linux server 11.10.  I cant find anyone lately having the same problem.  Where should I start to solve this?

PS

You should use p2pool 13.3. I've tried to upgrade 13.3 to 13.4 but the latter did not got any shares.
powersync
Member
**
Offline Offline

Activity: 74


View Profile
December 21, 2013, 04:12:33 PM
 #7254

Well went back to 13.3 and all is fine....  What's up with that?  Anyone having luck with 13.4?

PS
maqifrnswa
Sr. Member
****
Offline Offline

Activity: 454


View Profile
December 21, 2013, 11:24:28 PM
 #7255

Are you being dense on purpose, or do you really not get it?

The network targets a share rate, not a hash rate, not a
difficulty rate, and not a (share times value) rate.  It adjusts so
that the correct number of shares is found over time, which
shares can be of any difficulty equal to or greater than the minimum.
Shares of higher than the minimum difficulty are worth more, but still
only take up one "slot" in the calculation.

I'm surprised too that Kano doesn't know that. p2pool works this way:

Imagine there are two miners on p2pool, miner A has 10x the hashrate
of miner B. Miner A and B both use the default difficulty, which is
adjusted so that miner A gets 10x the shares of miner B and a share is
found every 30 seconds. In one hour (on average), Miner A finds ~109
shares and Miner B finds ~11 shares.

Now miner A manually sets his difficulty to 5x higher than the
default. He now finds shares 5x slower (on average). So in an hour,
miner A only finds 22 shares, and miner B still finds 11 shares. The
p2pool network now must decrease difficulty so that 120 shares
are found in an hour. Difficulty is thus decreased to the point where
miner B (the miner mining on default minimum difficulty) finds
120-22=98 shares in an hour. That means:

Miner A increased difficulty 5x
the p2pool minimum difficulty decreases 9x

both miner A and B get the same payout on average, miner A has higher variance than default and miner B has lower variance than default.
kano
Legendary
*
Offline Offline

Activity: 1918


Linux since 1997 RedHat 4


View Profile
December 22, 2013, 01:01:50 AM
 #7256

...
Are you being dense on purpose, or do you really not get it?

The network targets a share rate, not a hash rate, not a difficulty rate, and not a (share times value) rate.  It adjusts so that the correct number of shares is found over time, which shares can be of any difficulty equal to or greater than the minimum.  Shares of higher than the minimum difficulty are worth more, but still only take up one "slot" in the calculation.
Sigh - OK so what you are telling me is that the function in p2pool is flawed and not as specified on the p2pool.info web site.

The hash rate of the pool is simply the accepted difficulty of all the shares over a given period of time divided by that time.
The shorter the time frame, the more unreliable it is i.e. the greater the variance.

The way bitcoin works is it does indeed do what you are suggesting - however it does it over a set of data that is consistent - a set of blocks that have the same difficulty requirement.

Doing it over a set of data that isn't consistent, is, well, saying it's done incorrectly.
Maybe that is part of the reason why it jumps all over the place like crap.

What you are implying is that indeed it has nothing to do with the p2pool hash rate, but rather a set of shares with somewhat random difficulty.

I will also point out that your wording is vague and missing a detail:
The difficulty of the share submitted is only the difficulty that it was requested at, not the actual difficulty of the share itself.
All shares on all pools have varying difficulty, but it is only the difficulty of the work request that can affect anything on any pool payment scheme other than PoT

Edit: this also means the payout scheme could be flawed.

Pool: https://kano.is BTC: 1KanoiBupPiZfkwqB7rfLXAzPnoTshAVmb
CKPool and CGMiner developer, IRC FreeNode #ckpool and #cgminer kanoi
Help keep Bitcoin secure by mining on pools with Stratum, the best protocol to mine Bitcoins with ASIC hardware
kjj
Legendary
*
Offline Offline

Activity: 1302



View Profile
December 22, 2013, 05:32:53 AM
 #7257

Sigh - OK so what you are telling me is that the function in p2pool is flawed and not as specified on the p2pool.info web site.

The hash rate of the pool is simply the accepted difficulty of all the shares over a given period of time divided by that time.
The shorter the time frame, the more unreliable it is i.e. the greater the variance.

The way bitcoin works is it does indeed do what you are suggesting - however it does it over a set of data that is consistent - a set of blocks that have the same difficulty requirement.

Doing it over a set of data that isn't consistent, is, well, saying it's done incorrectly.
Maybe that is part of the reason why it jumps all over the place like crap.

What you are implying is that indeed it has nothing to do with the p2pool hash rate, but rather a set of shares with somewhat random difficulty.

I will also point out that your wording is vague and missing a detail:
The difficulty of the share submitted is only the difficulty that it was requested at, not the actual difficulty of the share itself.
All shares on all pools have varying difficulty, but it is only the difficulty of the work request that can affect anything on any pool payment scheme other than PoT

Edit: this also means the payout scheme could be flawed.

Yes, you should always read difficulty in this discussion as "difficulty of target" rather than "highest difficulty that target could have been and still been satisfied by this block".  My apologies for thinking that detail was obvious enough to warrant ommission.

I can't really imagine what makes you think that doing something other than the way you think it should be done makes it "wrong".  Hubris much?

Bitcoin uses long stretches of blocks with static target, and very strict rules for changing that target.  Those rules were chosen based on the needs of bitcoin.

p2pool has different needs.  The rules that make sense for bitcoin do not make sense for p2pool, so p2pool has not blindly copied them.  For example, in bitcoin, a large hashing attacker could profit by whipping the difficulty up and down, so we have strict rules about when and how difficulty can change which mitigate those attacks.  No such profit is to be had in p2pool.

In p2pool, when a block is found, the last N shares are paid out, each share based on that share's fraction of the total.  The total is easy to find.  The fraction of the total embedded in each share is easy to find.  I'm not sure what is "incorrect" about any of this.

p2pcoin: a USB/CD/PXE p2pool miner - 1N8ZXx2cuMzqBYSK72X4DAy1UdDbZQNPLf - todo
I routinely ignore posters with paid advertising in their sigs.  You should too.
TierNolan
Legendary
*
Offline Offline

Activity: 1036


View Profile
December 22, 2013, 07:28:03 PM
 #7258

Also, how is a more difficult share treated with respect to finding the longest blockchain? The same, or is a more difficult share given priority over a less difficult one?

All the same, I think.  It would give large miners a lower orphan rate otherwise.

Quote
I assume a miner who finds a block is also considered to have found a share (at whatever her difficulty level is set at), in addition to the 0.5% bonus, correct?

Weirdly, no.  You could find a block and still not get the share accepted.  You still get your 0.5% bonus though, since you would have put that in your coinbase.

1LxbG5cKXzTwZg9mjL3gaRE835uNQEteWF
TierNolan
Legendary
*
Offline Offline

Activity: 1036


View Profile
December 22, 2013, 07:31:58 PM
 #7259

I can't really imagine what makes you think that doing something other than the way you think it should be done makes it "wrong".  Hubris much?

The point is that it could lead to instability.  If the larger miners set their difficulty high for 1 re-target period and then all switched to low difficulty later, then the share rate would be much lower than 30 seconds.

Ofc, it would auto-correct after the next difficulty period. 

Also, if all large miners suddenly went from default difficulty to high difficulty, then it would slow down the chain.

Slightly off topic, the N in p2pool is the lower of, 3 blocks worth of shares and 3 days, right?

1LxbG5cKXzTwZg9mjL3gaRE835uNQEteWF
anti76
Sr. Member
****
Offline Offline

Activity: 343


View Profile
December 23, 2013, 09:11:34 AM
 #7260

AntMiner gives a lot of errors.
tell me how to adjust?

Total 204G
Dead 45G
Pages: « 1 ... 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 [363] 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 ... 744 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!