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

[sawa]

I tried renting rigs at https://www.miningrigrentals.com
Many devices do not even begin to mining:
https://www.miningrigrentals.com/rigs/35897
https://www.miningrigrentals.com/rigs/35744
https://www.miningrigrentals.com/rigs/35949
In the p2pool screen, I see the messages:

Code:

Squelched JSON error:
Traceback (most recent call last):
  File "/usr/local/lib/pypy2.7/dist-packages/twisted/protocols/basic.py", line 454, in dataReceived
    self.lineReceived(line)
  File "/opt/sha256d/p2pool-btc-fork/p2pool/util/jsonrpc.py", line 164, in lineReceived
    _handle(line, self, response_handler=self._matcher.got_response).addCallback(lambda line2: self.sendLine(line2) if line2 is not None else None)
  File "/usr/local/lib/pypy2.7/dist-packages/twisted/internet/defer.py", line 1447, in unwindGenerator
    return _inlineCallbacks(None, gen, Deferred())
  File "/usr/local/lib/pypy2.7/dist-packages/twisted/internet/defer.py", line 1301, in _inlineCallbacks
    result = g.send(result)
--- <exception caught here> ---
  File "/opt/sha256d/p2pool-btc-fork/p2pool/util/jsonrpc.py", line 85, in _handle
    result = yield method_meth(*list(preargs) + list(params))
exceptions.TypeError: rpc_submit() takes exactly 6 arguments (7 given)

I last tried renting rigs on mrr half a year ago. At that time, all leased devices were well mined on the p2pools.

[jtoomim]

sawa, I didn't touch any code related to the json-rpc stuff. My guess is that this is a recently introduced bug on the miningrigrentals end. It seems that they're appending an extra parameter on their share submission messages. Have you tried the same rentals on mainnet recently? My guess is that you'd see the same issue. You could also try packet sniffing (e.g. with tcpdump -X port 9332) to see what parameters they're sending; they might just be using some stratum extension that p2pool doesn't yet support.

If you use tcpdump, you're looking for lines that look like this:

Code:

13:27:03.541709 IP tbs9-44.36404 > feather.9336: Flags [P.], seq 3992288327:3992288497, ack 1433189083, win 1304, options [nop,nop,TS val 147459895 ecr 1475188945], length 170
	0x0000:  4500 00de a5a5 4000 4006 1a46 0a00 652c  E.....@.@..F..e,
	0x0010:  0a00 0103 8e34 2478 edf5 7c47 556c badb  .....4$x..|GUl..
	0x0020:  8018 0518 bce4 0000 0101 080a 08ca 0f37  ...............7
	0x0030:  57ed 98d1 7b22 7061 7261 6d73 223a 205b  W...{"params":.[
	0x0040:  2231 4775 446e 4579 5953 4533 5261 3370  "1GuDnEyYSE3Ra3p
	0x0050:  4d61 7237 3331 3174 7835 706f 5235 5047  Mar7311tx5poR5PG
	0x0060:  5852 3322 2c20 2231 3133 3233 3031 3730  XR3",."113230170
	0x0070:  3839 3331 3033 3539 3437 3630 3432 3831  8931035947604281
	0x0080:  3630 3638 3830 3530 3633 3033 3034 222c  60688050630304",
	0x0090:  2022 3838 3530 3032 3030 222c 2022 3539  ."88500200",."59
	0x00a0:  3162 3630 3732 222c 2022 6132 3163 6163  1b6072",."a21cac
	0x00b0:  6265 225d 2c20 2269 6422 3a20 3233 3836  be"],."id":.2386
	0x00c0:  392c 2022 6d65 7468 6f64 223a 2022 6d69  9,."method":."mi
	0x00d0:  6e69 6e67 2e73 7562 6d69 7422 7d0a       ning.submit"}.

If we know what extra data they're sending, we might be able to patch p2pool to support it (or, if it's safe, to ignore it).

[sawa]

sawa, I didn't touch any code related to the json-rpc stuff. My guess is that this is a recently introduced bug on the miningrigrentals end. It seems that they're appending an extra parameter on their share submission messages. Have you tried the same rentals on mainnet recently? My guess is that you'd see the same issue.

On any nodes the same error.
I tested a DGB node http://crypto.mine.nu:5027 based on the mainnet code.
When I got the error, I tried to switch the lease to the mainnet BTC http://crypto.office-on-the.net:9332 and to the fork http://crypto.office-on-the.net:9334.
The reaction of the nodes was the same everywhere - they reported an incorrect number of arguments.

You could also try packet sniffing (e.g. with tcpdump -X port 9332) to see what parameters they're sending; they might just be using some stratum extension that p2pool doesn't yet support.
If we know what extra data they're sending, we might be able to patch p2pool to support it (or, if it's safe, to ignore it).

I sent a tcpdump to your email

[keke51]

Hey! Where did all the Petahashes go? It's 1.8 instead of recent 12-18 Ph. I mean jtoomims fork of course)

[jtoomim]

Hey! Where did all the Petahashes go? It's 1.8 instead of recent 12-18 Ph. I mean jtoomims fork of course)

Nicehash rates are probably high right now. This will give you a chance to get more of your shares into the share chain and increase the amount of reward per block you'll get should a block be found. The ≥1 PH/s of hashrate on my nodes will be staying on jtoomimnet p2pool no matter what, so we will find a block eventually even if the Nicehash renters never come back. It might take a few weeks, though.

[tubexc]

P2pool is sailing death slow !

http://imgur.com/a/ptZyP

It is better to make an agreement to join the two versions and join forces 

[windpath]

P2pool is sailing death slow !

http://imgur.com/a/ptZyP

It is better to make an agreement to join the two versions and join forces 

That's the plan (for me anyway). Just waiting on testing to compete and merged into main p2pool branch.

[nur1labs]

is there ETA for payment or wait until some week wahaha (◕‿◕)

[Dibblah]

is there ETA for payment or wait until some week wahaha (◕‿◕)

If you're mining on P2Pool, you should know the answer to this question already. When we hit a block (on one side of the fork), all miners on that side will be paid out, proportional to (as far as I understand!) the number of shares they have in the sharechain. That's IF they're currently mining when the block is hit, otherwise, they get nothing

Cheers,

Allan.

[frodocooper]

... When we hit a block (on one side of the fork), all miners on that side will be paid out, proportional to (as far as I understand!) the number of shares they have in the sharechain...

More accurately, a P2Pool miner will be paid according to the total worth of the accepted shares he or she has in the P2Pool sharechain when a block is found. For example, if Alice found one share that is of the same difficulty level as Bob's ten shares combined, their payouts according to their respective shares would be the same. P2Pool's payouts are weighted according to share difficulty, not share quantity.

... That's IF they're currently mining when the block is hit, otherwise, they get nothing...

That is incorrect. A P2Pool miner is paid according to the total worth of his or her shares remaining in the sharechain when a block is found. The sharechain contains approximately three days' worth of shares. So, for example, even if Alice were to stop mining on P2Pool two days before P2Pool finds a block, Alice still has approximately one day's worth of accepted shares that she will get paid for when the block is found.

[lightfoot]

That is incorrect. A P2Pool miner is paid according to the total worth of his or her shares remaining in the sharechain when a block is found. The sharechain contains approximately three days' worth of shares. So, for example, even if Alice were to stop mining on P2Pool two days before P2Pool finds a block, Alice still has approximately one day's worth of accepted shares that she will get paid for when the block is found.

I wondered about that: If you mine a share that exceeds the difficulty threshold do you get paid more for a share that way exceeds the threshold over one that just squeaks by? That doesn't make a whole lot of sense.

C

[jtoomim]

If you mine a share that exceeds the difficulty threshold do you get paid more for a share that way exceeds the threshold over one that just squeaks by?

The difficulty of the share is different from the hash that the share actually gets. The threshold for the share is encoded into the share itself, and is part of what gets hashed, which means it's like saying "I promised you I would make a share with difficulty X, and if you hash this sentence (which includes nonce Y) you'll see that it meets that difficulty level." It doesn't matter how far you exceed that difficulty threshold.

[in2tactics]

This seems like yet another example of people not understanding what we are actually doing.

[jtoomim]

To whomever is mining on ml.toom.im:9332 with address 3GU89xZmzJfCBXFzbB3mC2dsxDyKqrD7nz:

P2pool does not support multisig or P2SH addresses. Any worker name that is not a regular P2PKH address (starting with a 1) ends up contributing to the default address for that node. In other words, you're mining into my wallet, not your own.

I'll send you the 0.125 btc you made in the last block soon-ish (few days? I don't currently have access to that wallet), but I think everyone would be happier if you mined directly into your own address. Edit: sent.

[HashSpike]

Can someone explain the drop in hashing power pointed to p2pool?  I saw someone else mention it was up to 12PH and now it's hovering between 2-3PH.  Did I miss something? 

Everyone go back to a centralized pool?  What the hay??

[Mister S]

Can someone explain the drop in hashing power pointed to p2pool?  I saw someone else mention it was up to 12PH and now it's hovering between 2-3PH.  Did I miss something? 

Everyone go back to a centralized pool?  What the hay??

Rentals are only sustainable if they're hitting blocks on a regular basis. The price of rentals also rises with demand. Most likely scenario is the rented hash that was contributing to P2Pool ran out and the cost vs. profit was too great a ratio to justify another week of 'maybe a block'

[lightfoot]

This seems like yet another example of people not understanding what we are actually doing.

No, I thought it was the former, but it had me questioning a bit. Technically the scenario of paying more for bigger shares would not be logical as the guy/gal/dog who submits the winning share would get a larger percentage of the proverbial pie.

I don't understand every nuance, my specialty is hacking and fixing the mining hardware. That I'm good at :-)

[nur1labs]

Ok that mean fair on share wahaha. nice p2pool you are great. i see last time friday is 4 Ph/s now why become 2 Ph/s wahaha....(◕‿◕)

To whomever is mining on ml.toom.im:9332 with address 3GU89xZmzJfCBXFzbB3mC2dsxDyKqrD7nz:

P2pool does not support multisig or P2SH addresses. Any worker name that is not a regular P2PKH address (starting with a 1) ends up contributing to the default address for that node. In other words, you're mining into my wallet, not your own.

I'll send you the 0.125 btc you made in the last block soon-ish (few days? I don't currently have access to that wallet), but I think everyone would be happier if you mined directly into your own address.

is that mean p2pool adresss or something?when i used another address can't run when p2pool adress from generated tools. only that workingo(≧▽≦)o

[Mister S]

This seems like yet another example of people not understanding what we are actually doing.

No, I thought it was the former, but it had me questioning a bit. Technically the scenario of paying more for bigger shares would not be logical as the guy/gal/dog who submits the winning share would get a larger percentage of the proverbial pie.

I don't understand every nuance, my specialty is hacking and fixing the mining hardware. That I'm good at :-)

Technically, the guy/gal/dog who submits the winning share DOES get the larger percentage of the proverbial pie, or at least used to, in the form of a 0.5% incentive. I don't know if that's still coded in or not though.

From the P2Pool Wiki: A subsidy of 0.5% is sent to the node that solved the block in order to discourage not sharing solutions that qualify as a block. (A miner with the aim to harm others could withhold the block, thereby preventing anybody from getting paid. He can NOT redirect the payout to himself.) The remaining 99.5% is distributed evenly to miners based on work done recently.

[nur1labs]

This seems like yet another example of people not understanding what we are actually doing.

No, I thought it was the former, but it had me questioning a bit. Technically the scenario of paying more for bigger shares would not be logical as the guy/gal/dog who submits the winning share would get a larger percentage of the proverbial pie.

I don't understand every nuance, my specialty is hacking and fixing the mining hardware. That I'm good at :-)

Technically, the guy/gal/dog who submits the winning share DOES get the larger percentage of the proverbial pie, or at least used to, in the form of a 0.5% incentive. I don't know if that's still coded in or not though.

From the P2Pool Wiki: A subsidy of 0.5% is sent to the node that solved the block in order to discourage not sharing solutions that qualify as a block. (A miner with the aim to harm others could withhold the block, thereby preventing anybody from getting paid. He can NOT redirect the payout to himself.) The remaining 99.5% is distributed evenly to miners based on work done recently.

so that it need waiting around 13 days to payout huh?(◕‿◕)