bitHopper: Python Pool Hopper Proxy - page 185.

joulesbeef

sr. member

Activity: 476

Merit: 250

moOo

still some problems flower if you can help.

I dont seem to get any error messages(that I recognize) but cant seem to connect, did you change the ports or anything?

Code:

D:\Users\joulesbeef\Desktop\dev\dev\c00w-bitHopper-50ab019>bitHopper.py
[15:32:48] reslice 1: -1 938434.470503 0
[15:32:48] reslice 1: -1 1564057.45084 0
[15:32:48] reslice 1: -1 1094840.21559 0
[15:32:48] reslice 1: -1 1564057.45084 0
[15:32:48] reslice 1: -1 1564057.45084 0
[15:32:48] reslice 1: -1 1564057.45084 0
[15:32:48] reslice 1: -1 1564057.45084 0
[15:32:48] slicing True
[15:32:48] multiclone weighted to 2.96311510161e-16
[15:32:48] eligius weighted to 1.98466422248e-16
[15:32:48] multiclone sliced to 359.326668493
[15:32:48] eligius sliced to 240.673331507
[15:32:48] Server change to multiclone, telling client with LP
[15:32:48] bitHopper Calculating MHash
[15:32:48] MHash0
[15:32:48] nofee:   5777681 3.69403
[15:32:48] reslice 1: 359.326668493 938434.470503 938434.470503
[15:32:48] OK NOT slicing because of multiclone
[15:32:48] reslice 1: -1 1564057.45084 1564057.45084
[15:32:48] reslice 1: 240.673331507 1094840.21559 1094840.21559
[15:32:48] OK NOT slicing because of eligius
[15:32:48] reslice 1: -1 1564057.45084 1564057.45084
[15:32:48] reslice 1: -1 5777681 1564057.45084
[15:32:48] reslice 1: -1 1564057.45084 1564057.45084
[15:32:48] reslice 1: -1 1564057.45084 1564057.45084
[15:32:48] slicing False
[15:32:49] mtred:             1254179 0.80188
[15:32:49] reslice 1: 359.326668493 938434.470503 938434.470503
[15:32:49] OK NOT slicing because of multiclone
[15:32:49] reslice 1: -1 1254179 1564057.45084
[15:32:49] FORCE slicing because of mtred
[15:32:49] slicing True
[15:32:49] multiclone weighted to 2.96311510161e-16
[15:32:49] eligius weighted to 1.98466422248e-16
[15:32:49] multiclone sliced to 359.326668493
[15:32:49] mtred sliced to 169.046142328
[15:32:49] eligius sliced to 240.673331507
[15:32:49] bitclockers:       2441503 1.56101
[15:32:49] reslice 1: 359.326668493 938434.470503 938434.470503
[15:32:49] OK NOT slicing because of multiclone

It just keeps doing that... and mine guiminer just says connecting.. it looks cool I want to try if you can help.

paraipan

legendary

Activity: 924

Merit: 1004

Firstbits: 1pirata

Quote from: flower1024 on July 17, 2011, 02:25:44 PM

i think we need to be that hub itself.
because if the hub announces the new block it wouldn't tell us where it get it from.

but: the long way to go is with a central server like fasthopper.appspot.
^^ the only question is how good that service would/could be

I was just thinking all the posib. around that fasthopper. Btw, we could view and correlate some ip's with block announce in central hubs making a static list, hope pools don't stay on dynamic ip's

... just joking

flower1024

legendary

Activity: 1428

Merit: 1000

btw: does bitp.it fakes it stats?
second time today that my hopper reports > 70%/diff and then immediatly goes down to 28%

unconfirmed reward did not change. so did they even found a block?

muyoso

member

Activity: 84

Merit: 10

Quote from: flower1024 on July 17, 2011, 02:05:44 PM

you take the risk of getting banned.
its VERY easy for pool operators to caught a hopper if he stays for more- let's say - 5 rounds.

and make a new account every 3 blocks: you need much hashrate to afford that, as there is a minimum payout

edit: btw the time-slice-method makes it more difficult for pool operators will detect you. as your hashrate will slowly go down

I don't think he cares about hopping. Pretty sure his main concern was thousands of connections coming from a single IP to a single account. Me jumping in and out of rounds with 1 connection from 1 ip, I don't think he cares, but we will see. His pool has had major connection issues, with multiclone and botnets overwhelming his servers.

joulesbeef

sr. member

Activity: 476

Merit: 250

moOo

Quote

there are entries missing. my new keys: slice and slicedShares (or sth like that)

I actually came back as quick as I could to erase my last post.. but you are too quick.

I tried to add arsbitcoin as a backup and screwed something up

ahh you use multiclone instead as a backup.. cool. I tried them a little, seems like a good backup.

how does it choose to use the backups? will too many screw me up? can i still add arscoin?

flower1024

legendary

Activity: 1428

Merit: 1000

i think we need to be that hub itself.
because if the hub announces the new block it wouldn't tell us where it get it from.

but: the long way to go is with a central server like fasthopper.appspot.
^^ the only question is how good that service would/could be

paraipan

legendary

Activity: 924

Merit: 1004

Firstbits: 1pirata

Quote from: flower1024 on July 17, 2011, 02:00:36 PM

Quote from: joulesbeef on July 17, 2011, 01:57:09 PM

cool flower, I will check out your hopper as well. and let you know.

its not mine... its c00w. just for clarification

i am just playing with the maths

and it is not perfect right now!

planned:
- better handling of backup bools
- regard pools hashrate

but next i'll try to patch bitcoin itself to detect btcguild block - just because i think its possible

me too "thinks that", don't remember where i read something about a central hub patch for the pools to announce it's blocks very fast. We could connect to that hub instead of patching bitcoin, what do you think

flower1024

legendary

Activity: 1428

Merit: 1000

please recheck pool.py

there are entries missing. my new keys: slice and slicedShares (or sth like that)

joulesbeef

sr. member

Activity: 476

Merit: 250

moOo

I meant your modification of his hopper.
I do like how you cleaned out the cobwebs

but I am having problems making it work, cause I am soo noob

Code:

D:\Users\joulesbeef\Desktop\dev\dev\c00w-bitHopper-50ab019>bithopper.py
Unhandled error in Deferred:
Unhandled Error
Traceback (most recent call last):
  File "D:\Users\joulesbeef\Desktop\dev\dev\c00w-bitHopper-50ab019\bitHopper.py"
, line 391, in 
    main()
  File "D:\Users\joulesbeef\Desktop\dev\dev\c00w-bitHopper-50ab019\bitHopper.py"
, line 383, in main
    slice_call.start(2)
  File "D:\Python27\lib\site-packages\twisted\internet\task.py", line 163, in st
art
    self()
  File "D:\Python27\lib\site-packages\twisted\internet\task.py", line 194, in __
call__
    d = defer.maybeDeferred(self.f, *self.a, **self.kw)
---  ---
  File "D:\Python27\lib\site-packages\twisted\internet\defer.py", line 133, in m
aybeDeferred
    result = f(*args, **kw)
  File "D:\Users\joulesbeef\Desktop\dev\dev\c00w-bitHopper-50ab019\bitHopper.py"
, line 48, in slice_server
    select_best_server()
  File "D:\Users\joulesbeef\Desktop\dev\dev\c00w-bitHopper-50ab019\bitHopper.py"
, line 128, in select_best_server
    log_msg("reslice 1: " + str(info["slice"]) + " " + str(info["shares"]) + " "
 + str(info['slicedShares']))
exceptions.KeyError: 'slice'
[15:14:39] bitHopper Calculating MHash
[15:14:39] MHash0
[15:14:41] nofee:   5773049 3.69107
[15:14:41] Error in pool api for nofee
"[Failure instance: Traceback: : 'slice'\nD:\\Python
27\\lib\\site-packages\\twisted\\internet\\defer.py:1076:gotResult\nD:\\Python27
\\lib\\site-packages\\twisted\\internet\\defer.py:1063:_inlineCallbacks\nD:\\Pyt
hon27\\lib\\site-packages\\twisted\\internet\\defer.py:361:callback\nD:\\Python2
7\\lib\\site-packages\\twisted\\internet\\defer.py:455:_startRunCallbacks\n--- <
exception caught here> ---\nD:\\Python27\\lib\\site-packages\\twisted\\internet\
\defer.py:542:_runCallbacks\nD:\\Users\\joulesbeef\\Desktop\\dev\\dev\\c00w-bitH
opper-50ab019\\pool.py:132:selectsharesResponse\nD:\\Users\\joulesbeef\\Desktop\
\dev\\dev\\c00w-bitHopper-50ab019\\bitHopper.py:195:server_update\nD:\\Users\\jo
ulesbeef\\Desktop\\dev\\dev\\c00w-bitHopper-50ab019\\bitHopper.py:128:select_bes
t_server\n]"
[15:14:42] mtred:             1172395 0.74959
[15:14:42] Error in pool api for mtred

using python 2.7 if that makes any difference.

I copied over my password file from the cows unmodified version, erasing what i didnt need

any ideas?

flower1024

legendary

Activity: 1428

Merit: 1000

you take the risk of getting banned.
its VERY easy for pool operators to caught a hopper if he stays for more- let's say - 5 rounds.

and make a new account every 3 blocks: you need much hashrate to afford that, as there is a minimum payout

edit: btw the time-slice-method makes it more difficult for pool operators will detect you. as your hashrate will slowly go down

muyoso

member

Activity: 84

Merit: 10

Quote from: joulesbeef on July 17, 2011, 01:57:09 PM

yeah not a fan of bitcoin.lc.. how he handled the whole multiclone affair left a foul taste in my mouth. He eventually gave them back their bitcoin but doesnt change how he handled it in the first place.

I'd rather just deal with one less place to hop even if our list is rather short.

cool flower, I will check out your hopper as well. and let you know.

Yea, the way he handled that was ridiculous. I think his issue was the thousands of connections going to one account, but trying to keep the money is plain theft.

The only reason I would consider to continue mining there is because there simply isn't that many pools to hop to right now, and I feel the larger the list of hoppable pools the more I will make.

flower1024

legendary

Activity: 1428

Merit: 1000

Quote from: joulesbeef on July 17, 2011, 01:57:09 PM

cool flower, I will check out your hopper as well. and let you know.

its not mine... its c00w. just for clarification

i am just playing with the maths

and it is not perfect right now!

planned:
- better handling of backup bools
- regard pools hashrate

but next i'll try to patch bitcoin itself to detect btcguild block - just because i think its possible

joulesbeef

sr. member

Activity: 476

Merit: 250

moOo

yeah not a fan of bitcoin.lc.. how he handled the whole multiclone affair left a foul taste in my mouth. He eventually gave them back their bitcoin but doesnt change how he handled it in the first place.

I'd rather just deal with one less place to hop even if our list is rather short.

cool flower, I will check out your hopper as well. and let you know.

flower1024

legendary

Activity: 1428

Merit: 1000

bitcoins.lc is known to HATE hoppers.
and the owner is a really unkind person

i wont ever mine their again (even when hopping [hopefully someday] becomes unprofitable)

i don't work for douches!

muyoso

member

Activity: 84

Merit: 10

Damnit, bitcoins.lc. . . They routinely let their site misreport the length of rounds and then after the fact correct them. They had a "14 hour" round that suddenly turned into a 12 hour round, a 13 minute round and an hour and 45 minute round. Bithopper didn't jump in on the 13 minute or hour and 45 minute round because of the misreporting. They do this almost everytime there is a very long round. Wonder if its an attempt to stop hopping or if there is an issue with the pool itself. People complain about this regularly in their thread in the Pools section.

Clipse

hero member

Activity: 504

Merit: 502

Thanks will check it out and report back.

Appreciate it

flower1024

legendary

Activity: 1428

Merit: 1000

here you go:

www.k1024.de/dev.7z

it has a small web interface:

http://ip:mineport/current/index

i emptied password.py completely

EDIT: please check pool.py. sometimes i change keys there too.... to lazy to look at it right now

Clipse

hero member

Activity: 504

Merit: 502

Well it would be nice to get a seperate bithopper version using your method, so I can run both evenly split among my available mhash and see over 24-48hrs if there is a different in earnings.

flower1024

legendary

Activity: 1428

Merit: 1000

Quote from: Clipse on July 17, 2011, 01:14:02 PM

Quote from: flower1024 on July 17, 2011, 12:58:06 PM

earlier sharesare worth more than later shares.
you always (prop) want to go the pool which founds a block.

it gets a little tricky if two pools found a block very close. i made a patch for that - have to discussed with c00w (just see git if interested)

I will wait till c00w update bithopper if neccesary.

So what you suggest is, no matter the speed of a pool, if one pool moves above X shares in current difficulty, it is allways better to switch to another pool that has lower than X shares in current difficulty, even if X shares from first pool is less than the predefined hop percentage ie. 40% ?

this current version i published to discuss it only works with pools which has an more-or-less equal hashrate.

i don't say you should ever not switch to a pool with the most less shares.

but if there are two pools which found a block the same second you don't know which of them is better (means: who will solve this block first).

so best thing would be to divide your shares 50/50 (again: same hashrate atm only, could be improved)

i would prefer to switch pool whenever a share is found on a random basis (see my other version, not on git but in this forum - just for a look; it isnt really usable), but that just do only work if you have just one miner attached. with two or more you'll fuck up some getworks and get way more stales.

i am using my "time-slice" method since ten hours now (tweaked it a little more) and it seems good.

eg: the last mtred -> ozcoin switch. for you it switched immediatly. for me my shares has been splitted

Clipse

hero member

Activity: 504

Merit: 502

Quote from: flower1024 on July 17, 2011, 12:58:06 PM

earlier sharesare worth more than later shares.
you always (prop) want to go the pool which founds a block.

it gets a little tricky if two pools found a block very close. i made a patch for that - have to discussed with c00w (just see git if interested)

I will wait till c00w update bithopper if neccesary.

So what you suggest is, no matter the speed of a pool, if one pool moves above X shares in current difficulty, it is allways better to switch to another pool that has lower than X shares in current difficulty, even if X shares from first pool is less than the predefined hop percentage ie. 40% ?

Topic: bitHopper: Python Pool Hopper Proxy - page 185. (Read 355816 times)