bitHopper: Python Pool Hopper Proxy

[c00w]

gnadget: We only hold the last 5 minutes worth of objects. Memory usage with 500 getworks in 5 minutes < 1 mb.

[paraipan]

also, has anyone got polmine to work? i just get API error and 591402.50% shares.

copy>paste whole pool.cfg.default entry to you cfg and edit only miner and pass, here work just fine

[gnaget]

gnadget: We only hold the last 5 minutes worth of objects. Memory usage with 500 getworks in 5 minutes < 1 mb.

But we hold the entire merkle root in string format (64 characters).  Assuming a 2 byte encoding, we are looking at 128 bytes per merkle root.  I would estimate I am averaging 3 getworks / s so 3*60*5 = 900.  900 * 128 = 115,200 BYTES...  I see my error in caclulations: I was thinking in kilobytes, duh computers are hard.

[gooie]

also, has anyone got polmine to work? i just get API error and 591402.50% shares.

copy>paste whole pool.cfg.default entry to you cfg and edit only miner and pass, here work just fine

i got it to work, i think it might have something to do with that there's an english version and a polish version of the site, resulting in that it doesn't find the polish text "szystkich".

So i changed it to make it friendly for both the versions.

Code:

api_key:: &nbsp; </b> <br/>([ 0-9]+)<br/> <br/></td><td><b>

[paraipan]

gnadget: We only hold the last 5 minutes worth of objects. Memory usage with 500 getworks in 5 minutes < 1 mb.

But we hold the entire merkle root in string format (64 characters).  Assuming a 2 byte encoding, we are looking at 128 bytes per merkle root.  I would estimate I am averaging 3 getworks / s so 3*60*5 = 900.  900 * 128 = 115,200 BYTES...  I see my error in caclulations: I was thinking in kilobytes, duh computers are hard.

that's 10% of 1 mb, would handle bigger loads then...

edit: rfc lags big time, let's hope they upgrade sometime in the future

[gnaget]

check the pool websites your mining at, rfc is showing 2% stales for me this round using your script but that could be for other reasons.

Good idea...  I think I can take this as confirmation (rfc):
Last 15 Mins    1359    0    0.000%

I can't believe this gets less stales than phoenix by itself. 

[c00w]

Phoenix and most other miners only have 1 thread doing submissions. If it fails they drop the work. Bithopper doesn't.

[gnaget]

Phoenix and most other miners only have 1 thread doing submissions. If it fails they drop the work. Bithopper doesn't.

kick ass, is that one of your features?  I just keep becoming more and more of a fan of this proxy

[paraipan]

Phoenix and most other miners only have 1 thread doing submissions. If it fails they drop the work. Bithopper doesn't.

hope they mention you guys when they will "borrow" code from bH if it appears to have a better approach to getwork's

[gnaget]

Of course, right after I move it over to a screen session, I get this:

Last 15 Mins    1158    21    1.813%

21 stales, and on the stats page I see 9257 / 0

This is unfortunate, there might be an issue with stales not getting reported.  I think the store is working though, rfc sucks, so it isn't a good test

[gnaget]

You know you are watching your stats screen too closely when you start seeing these:

[00:26:02] RPC request [dead1000] submitted to BTC World

[Keninishna]

lol 0 stales in last 15 min at rfc for me. 

[gnaget]

ok, I feel confident enough to leave it unattended for the night.  0 stales on ozco in the last 20 minutes

[djex]

ok, I feel confident enough to leave it unattended for the night.  0 stales on ozco in the last 20 minutes

Same here. It's working great. I got 2 stales on rfc out of 1652 shares but that's more than likely lag. Great work.

[hawks5999]

God, the night you guys break through and kill stales I'm stuck in a tent just a phone. I'll probably burn 10,000 stales before I can get to upgrade.

[djex]

God, the night you guys break through and kill stales I'm stuck in a tent just a phone. I'll probably burn 10,000 stales before I can get to upgrade.

That's why I have remote ssh into my server from my phone

[c00w]

@Hawks
How many servers and how many gh/s do you have if you make 10,000 shares in a night? Or are you on a 20 day trek?

[EskimoBob]

c00w, your script is getting better all the time. 

I still like to come back to this old issue of switching pools for no reason aka "DNS lookup failures"

Code:

[09:50:56] RPC request [2a769000] submitted to ozco.in
[09:50:59] mtred: 3165996
[09:50:59] RPC request [8289e000] submitted to ozco.in
Caught, jsonrpc_call insides
DNS lookup failed: address 'ozco.in' not found: [Errno -2] Name or service not known. <--- really? This can/will cause unnecessary hopping :(  
[09:51:04] RPC request [f65c1000] submitted to ozco.in
[09:51:05] RPC request [getwork] submitted to ozco.in
[09:51:06] triple: 8156248 <-- they really need some help about now

Cheers and thank you for all your work.

[c00w]

Eskimo:
Did it switch?

And DNS Lookup Failures is an error. I can write code to just retry though on that specific error if you are proccesing getworks. If you are submitting work it will currently never change servers although it will mark them as lagging.

[macboy80]

@c00w - I'm getting a stats page error like before. Only it's happening when I run from the python.exe. Running from idle, it works fine. I believe the offending code is:

Code:

D:\BitHopper\c00w-bitHopper\website.py, line 45 in render_GET
43   except:
44    index = 'index.html'
45   file = open(index, 'r')
46   linestring = file.read()
Locals
index	'index.html'

<type 'exceptions.IOError'>: [Errno 2] No such file or directory: 'index.html'

That's the last bit of info on the stats page.