Continuum Mining Pool: No fees; Client uptime monitoring via twitter and email

[martok]

Hello,

I would like to announce the availability of the Continuum mining pool. It is new and could use some testing. The pool runs with no fees and is aimed at more serious miners, though all are certainly welcome. The pool supports long-polling and should provide good connectivity to north-american miners. It also supports monitoring of miners; if your miner hasn't submitted a share within a defined period, you will get an email or Twitter to that effect.

Usage:
Simply connect your miner to continuumpool.com port 8332. Use the bitcoin address to which you would like to be paid as the username and use any password. I recommend that you use a different Bitcoin address for each worker so that balances and uptime monitoring can be set per-worker.

The pool provides an RPC interface. It does not provide a web interface, though anyone is welcome to write one. To use the RPC service, connect to http://rpc.continuumpool.com:8330/rpc and use the following methods.

balance(workeraddr) Returns confirmed unpaid balance
balancecurrent(addr) Current round balance
roundstart() Returns the time of the start of the round.
monitoradd({worker => 'workeraddr', 'contact' => 'mailto:me@mydomain.com', 'sharetime' => '5 minutes', 'name' => 'some descriptive name'}) Adds an email monitor. If the worker hasn't submitted a share in set time, an email is generated
monitordrop({id => monitorid, worker => 'workeraddr'}) Cancels a monitor. Get the ID from the generated email.

Once again, testing and feedback would be greatly appreciated.

[xenon481]

Share or Score based?

When are confirmed balances paid? At x time, threshold, etc?

[martok]

I am using the scoring algorithm described here:
http://bitcointalk.org/index.php?topic=4787.0
It was designed to prevent pool-hopping attacks.

Payout is automatic when the worker hits 1 BTC.

[xenon481]

I am using the scoring algorithm described here:
http://bitcointalk.org/index.php?topic=4787.0
It was designed to prevent pool-hopping attacks.

Payout is automatic when the worker hits 1 BTC.

Thank you, sounds great!

[martok]

For those interested, here is a little bit of perl that sets up a monitor for a worker. Replace the obvious bits.

#!/usr/bin/perl
use JSON::RPC::Client;
my $client = new JSON::RPC::Client;
my $uri = 'http://rpc.continuumpool.com:8330/rpc';
$client->prepare($uri, ['hashrate','roundstart','monitoradd','dropmonitor', 'balance','balancecurrent']);
$worker = '1NNdsxTuF1p9J9hmFKoxdM3ki4GgMBajN7';
$res = $client->monitoradd({
'worker' => $worker, 'contact' => 'mailto:me@mydomain.com',
'sharetime' => '10 minutes', 'name' => 'My 5970'}));
print "Success" if ($res == 1);

Using the framework above, you can get the pool hashrate with:
print $client->hashrate(undef);
or the worker hashrate with:
print $client->hashrate("workeraddr");

balance and balancecurrent work the same way.

[martok]

Twitter monitoring support is in. You simply use twitter:username instead of mailto:username in the contact field shown above. When your miner goes down, you'll get a DM which you can have Twitter send to your phone etc.

Note: you must be following ContinuumPool on Twitter for this to work.

If there is interest, I can add the ability to query balances, set up monitors etc via Twitter. Might be handy for those who don't want to code RPC clients.

[martok]

To the user submitting shares with a username which is not a valid bitcoin address: you can no longer use a username which is not a bitcoin address so are no longer able to submit shares. That being said, we found two quick blocks today so you are owed some BTC assuming they mature. If you can PM me with the username your client was using and a BTC address, I will have them credited to you.

[allinvain]

Cool, a Canadian based pool ! I definitely shall give your pool a try.

Do you pay out from generated blocks or directly like Luke's (Eligius) pool?

Seems like a clone of Luke's (Eligius) pool, but it doesn't really matter. The more pools the better.

Now for some quick questions. Are you running this off of a dedicated connection (data center) or a shaw cable connection ? Do you have a backup server or some sort of high availability setup?

Also where do you guys get the pool software from? I too would like to setup a similar type of pool for me and a bunch of friends. Any chance of sharing the software?

[Meni Rosenfeld]

I am using the scoring algorithm described here:
http://bitcointalk.org/index.php?topic=4787.0
It was designed to prevent pool-hopping attacks.

I'm very excited to hear this! I'm switching my 2.2 GH/s from slush as soon as I have the time. I'll be happy to further advise you about the implementation details.

By "no fees", you mean a balancing negative fixed fee?

[martok]

I'm very excited to hear this! I'm switching my 2.2 GH/s from slush as soon as I have the time. I'll be happy to further advise you about the implementation details.

I would appreciate that. The implementation is pretty simple and is in a single PLPGSQL function I would be happy to post in your thread.

By "no fees", you mean a balancing negative fixed fee?

Correct, that was the only way I could see to implement this. So c = 0.001 and f = (-0.001/0.999) or -0.001...

[Meni Rosenfeld]

By "no fees", you mean a balancing negative fixed fee?

Correct, that was the only way I could see to implement this. So c = 0.001 and f = (-0.001/0.999) or -0.001...

Right. Be mindful of the effect these parameters have on the variance of yourself and the participants'.

[martok]

Cool, a Canadian based pool ! I definitely shall give your pool a try.

I would appreciate any feedback.

Do you pay out from generated blocks or directly like Luke's (Eligius) pool?

Up until last round, the payouts were from generated blocks IE 120 confirmations later. However, I think I have direct payment working. We'll see if it generates direct payments on the next block and go from there. But that is the direction in which I would like to go.

Seems like a clone of Luke's (Eligius) pool, but it doesn't really matter. The more pools the better.

Eligius is an Excellent pool. I wanted to do more with notifications of downed miners though. IE when a miner is down, I want an SMS etc. Luke definitely has the payment system right though.

Now for some quick questions. Are you running this off of a dedicated connection (data center) or a shaw cable connection ? Do you have a backup server or some sort of high availability setup?

I have two servers, though only one is currently running the pool. I will either do a failover DNS setup or possibly a round-robin approach to balance the load between the two servers. For now though, I am more focused on seeing whether the system will stand up to some load other than that which I impose myself. I imagine most folks have failover setups of their own. IE if a pool fails, fallback.

Also where do you guys get the pool software from? I too would like to setup a similar type of pool for me and a bunch of friends. Any chance of sharing the software?

I am using pushpool for the backend. It is written by someone else but is free software. You can search the forums. The database and RPC stuff is custom. Though I would be happy to share it, it probably wouldn't be very useful.

[martok]

By "no fees", you mean a balancing negative fixed fee?

Correct, that was the only way I could see to implement this. So c = 0.001 and f = (-0.001/0.999) or -0.001...

Right. Be mindful of the effect these parameters have on the variance of yourself and the participants'.

Indeed. I'm still not commited to the current value of c but in my local testing it performed alright. It will be interesting to give it more wider testing.

[martok]

We are now hosted on a commercial softlayer server. Not that uptime has been a problem but now we can do fallback if we need to. New host is continuumpool.com for mining and rpc.continuumpool.com:8330/rpc for RPC interface.

[tiberiandusk]

Anyone want to setup a neat graph thingy like someone did for Eligius? I am lazy.

[fortyniner]

with Eligius down, just moved my 700+ Mhash miner to Continuum.  Smooth sailing so far.  Thanks.

[fortyniner]

I see miner stats swinging from 440 to 930 Mhash; must be a very short duration (instantaneous?) calculation.   And with a current Overall hashrate of 3.X Ghash/s and no block since 5/17 ... we're likely to be solving one soon.

[martok]

The hashrate calculation uses a 5 minute window so that's very strange indeed. I was doing some database work today so maybe that might explain it.
We actually found 3 blocks on 17-may.
It's worth noting that the probability of finding a new block doesn't increase with the round duration. It is the same for a 1 minute round as it is for a 4 day round.

[fortyniner]

You are, of course, correct on probability.  I look at it from the gambler's point of view, not the mathematician's.  Six days of nothing means good luck's comin'.  Is your server really in Canada?  I'd guess Texas.

And did you break the "last round" stat or me?  http://www.continuumpool.com/roundstart.php reports nothing for me.

[martok]

And did you break the "last round" stat or me?  http://www.continuumpool.com/roundstart.php reports nothing for me.

Sorry, I broke it. Should be working now. Thanks for the heads up.