anisoptera
Member
Offline
Activity: 308
Merit: 10
|
|
May 27, 2011, 11:08:39 PM |
|
Once you get multiple miners to single address working (and perhaps a bit more hash power) I might move to your pool Lack of registration requirement is great.
|
|
|
|
brocktice
|
|
May 27, 2011, 11:43:55 PM |
|
I uh, invited some friends. How's the server holding up?
Indeed, and they found a block. Server load is pretty much where it was when we started. Sitting at 0.3 with most of that being consumed by bitcoind itself. Looks to me like it's having issues, the miners are going idle a lot, and the rpc is down.
|
|
|
|
martok (OP)
|
|
May 28, 2011, 01:00:57 AM |
|
I uh, invited some friends. How's the server holding up?
Indeed, and they found a block. Server load is pretty much where it was when we started. Sitting at 0.3 with most of that being consumed by bitcoind itself. Looks to me like it's having issues, the miners are going idle a lot, and the rpc is down. Is this still the case? I have been implementing the payout aggregation stuff so had to restart the pool a couple times. If it's still happening, let me know and I'll start through the logs.
|
|
|
|
martok (OP)
|
|
May 28, 2011, 01:05:32 AM |
|
You can now aggregate your worker payments by logging in with workeraddr;paymentaddr as your username. workeraddr is still the primary address to which notification monitors are tied etc, paymentaddr is the address to which payments can be sent. So for example, if you have multiple workers, you can set the same paymentaddr for all of them to get to the 1 BTC payment threshold faster.
|
|
|
|
organofcorti
Donator
Legendary
Offline
Activity: 2058
Merit: 1007
Poor impulse control.
|
|
May 28, 2011, 03:00:14 AM |
|
Thanks for the work on this pool - was hoping someone would implement Meni Rosenfeld's algo. Couple of comments - my hashrate is reported at continuumpool as 10 - 30% higher than it actually is ( I'm at an average of 350 MH/s as reported by poclbm and R tools, I'm seeing 400 to 450MH/s from the website). I'm not complaining, and it's probably normal variability, but I thought you'd want to know. also, on www.continuumpool.com you've written: The RPC interface can be accessed via rpc.pl and is documented in the Bitcoin forum. I searched for RPC in the forums and got wayyyyy too many hits - do you have a link I've missed in this thread? Cheers
|
|
|
|
martok (OP)
|
|
May 28, 2011, 06:39:34 AM |
|
Thanks for the work on this pool - was hoping someone would implement Meni Rosenfeld's algo. Couple of comments -
Well, it's far from perfect but we're getting there. my hashrate is reported at continuumpool as 10 - 30% higher than it actually is ( I'm at an average of 350 MH/s as reported by poclbm and R tools, I'm seeing 400 to 450MH/s from the website). I'm not complaining, and it's probably normal variability, but I thought you'd want to know.
I'm pretty sure it's just variability. It's just reporting the shares it gets back within 5 minutes so you may have been just hitting a lot of shares. There is a report above that hashrate was 15% below poclbm so maybe the formula could use some smoothing. I don't really pay much attention to it personally. also, on www.continuumpool.com you've written: The RPC interface can be accessed via rpc.pl and is documented in the Bitcoin forum. I searched for RPC in the forums and got wayyyyy too many hits - do you have a link I've missed in this thread? Check out the first few posts in this thread for RPC documentation and PERL code samples. They should get you on your way. I'll link the source code to those little scripts on wwww.continuumpool.com as well as that might help people write their own monitors.
|
|
|
|
fortyniner
Newbie
Offline
Activity: 27
Merit: 0
|
|
May 28, 2011, 01:40:50 PM Last edit: May 28, 2011, 03:40:06 PM by fortyniner |
|
I do not see payment from last block solve of 05-27. That's the only issue I've seen. EDIT: now I see it... 5/28/11 11:00
|
|
|
|
anisoptera
Member
Offline
Activity: 308
Merit: 10
|
|
May 28, 2011, 05:24:18 PM |
|
You can now aggregate your worker payments by logging in with workeraddr;paymentaddr as your username. workeraddr is still the primary address to which notification monitors are tied etc, paymentaddr is the address to which payments can be sent. So for example, if you have multiple workers, you can set the same paymentaddr for all of them to get to the 1 BTC payment threshold faster.
Does workeraddr have to be a valid Bitcoin address when I do this, or can I generate my own arbitrary worker name and follow it with an actual address? If not, is there any reason you can't make it this way? It would really make my life easy
|
|
|
|
martok (OP)
|
|
May 28, 2011, 07:31:56 PM |
|
Does workeraddr have to be a valid Bitcoin address when I do this, or can I generate my own arbitrary worker name and follow it with an actual address? If not, is there any reason you can't make it this way? It would really make my life easy It needs to be a valid address. The reason for this is because monitors need a unique value per worker and creating your own name would have no guarantee of uniqueness. The other reason is a problem of previous shares. Let's say I switch my worker to the a payto address mid-round. My worker address has a balance between 0 and 1 but now will never reach 1 because new shares are being credited to the new payto address. Having the same valid worker address still present will allow the pool to pay out the previous worker address once the payto address + worker hits 1 BTC. Er hope that makes some sense.
|
|
|
|
anisoptera
Member
Offline
Activity: 308
Merit: 10
|
|
May 29, 2011, 02:47:33 AM |
|
Does workeraddr have to be a valid Bitcoin address when I do this, or can I generate my own arbitrary worker name and follow it with an actual address? If not, is there any reason you can't make it this way? It would really make my life easy It needs to be a valid address. The reason for this is because monitors need a unique value per worker and creating your own name would have no guarantee of uniqueness. The other reason is a problem of previous shares. Let's say I switch my worker to the a payto address mid-round. My worker address has a balance between 0 and 1 but now will never reach 1 because new shares are being credited to the new payto address. Having the same valid worker address still present will allow the pool to pay out the previous worker address once the payto address + worker hits 1 BTC. Er hope that makes some sense. Sure, but the combination of payment address and worker name is almost guaranteed to be unique, even if it's not a bitcoin address. If I am bringing a new miner here, I don't need the association with old mined BTC, and it'd be more meaningful for me to just name them something reasonable. Of course, I could just write fake bitcoin addresses that contain the name I want, but that seems inefficient.
|
|
|
|
martok (OP)
|
|
May 29, 2011, 03:07:37 AM |
|
Sure, but the combination of payment address and worker name is almost guaranteed to be unique, even if it's not a bitcoin address. If I am bringing a new miner here, I don't need the association with old mined BTC, and it'd be more meaningful for me to just name them something reasonable. Of course, I could just write fake bitcoin addresses that contain the name I want, but that seems inefficient. I see what you're saying. But if we did that, we break the whole monitor API as that takes a worker address. Though you can associate a meaningful name in a monitor for things like Twitter and email alerts. If the pool switched to potentially non-unique worker names, we would need to redo monitors, of which there are currently quite a few. Feel free to put in dummy BTC addresses though.
|
|
|
|
martok (OP)
|
|
May 29, 2011, 07:57:20 PM |
|
I have applied an async RPC patch to bitcoind. Hopefully it will drastically reduce RPC errors and idle miners. I had no idea bitcoind blocked the entire process for a getwork request. http://forum.bitcoin.org/index.php?topic=7757.0Let me know if you're still getting RPC idle messages.
|
|
|
|
Transcendental
Newbie
Offline
Activity: 14
Merit: 0
|
|
May 30, 2011, 05:36:19 AM |
|
Three questions: 1) I seem to be getting a TON of invalid/stale exceptions now - did something drastic change? Or have I messed something up with my settings? invalid/stales are current around 30% of what I am generating.
2) I have been unable to login to the pool with poclbm using "-u wokeraddr;paymentaddr" where workeraddr and paymentaddr, could you perhaps give an example?
3) I am using multiple computers over a lan, is there someway I can set this up to connect to your server from only one of my computers?
Thanks.
|
|
|
|
martok (OP)
|
|
May 30, 2011, 07:25:46 AM Last edit: May 30, 2011, 09:53:52 PM by martok |
|
Three questions: 1) I seem to be getting a TON of invalid/stale exceptions now - did something drastic change? Or have I messed something up with my settings? invalid/stales are current around 30% of what I am generating.
I broke LP. Fixed now but I'll watch the logs for stales for the next hour and see if that fixes it. 2) I have been unable to login to the pool with poclbm using "-u wokeraddr;paymentaddr" where workeraddr and paymentaddr, could you perhaps give an example?
Try -u "worker;payto" or to avoid the quotes -u worker0payto 0 also works as a seperator because 0 cannot appear in a bitcoin address. So: poclbm -u "16d1jmkPHy26aCTA9TzCD5LbByMVHb4Wu1;1AkUQgtqHbiN4hpYX7VbJVR5hhhb3SAck9" works for me. Feel free to post your command line though and I'll give it a test over here. 3) I am using multiple computers over a lan, is there someway I can set this up to connect to your server from only one of my computers?
You could use the mining proxy discussed on the forum elsewhere. That proxy can switch over to another pool if the primary is down also. You would have all your lan computers connect to the proxy and have the proxy connect to the pool. (edit) added 0 seperator.
|
|
|
|
Meni Rosenfeld
Donator
Legendary
Offline
Activity: 2058
Merit: 1054
|
|
May 30, 2011, 10:59:16 AM |
|
Do you aggregate payouts from the same IP? I have 6 poclbm instances running on two machines, each with a different address, and it seems I got to one of the addresses the rewards for all.
|
|
|
|
martok (OP)
|
|
May 30, 2011, 05:34:53 PM |
|
Do you aggregate payouts from the same IP? I have 6 poclbm instances running on two machines, each with a different address, and it seems I got to one of the addresses the rewards for all.
No. The IP is not used in any payout calculations so that is strange indeed. I am using the sendmany RPC call to batch send transactions into a single TX but your bitcoind should still show you receiving payments on different addresses.
|
|
|
|
martok (OP)
|
|
May 31, 2011, 12:48:48 AM |
|
For those wanting PPS payment, it is now live. Simply use yourpaymentaddress;pps or yourpaymentaddress0pps as the username. PPS miners will not affect regular miners and PPS carries a 5% expected fee. Regular mining is of course still at no fee. However, PPS was requested so there it is. edit: PPS payout rate: http://www.continuumpool.com/ppsrate.phpor preferably through the ppsrate RPC call. edit: You can now pass a boolean pps (null by default) argument to hashrate. pps => null = overall hashrate pps => false regular miners only pps => true only pps miners
|
|
|
|
brocktice
|
|
May 31, 2011, 07:33:58 PM |
|
monitordrop({id => monitorid, worker => 'workeraddr'}) Cancels a monitor. Get the ID from the generated email.
How can I get the monitor ID if I'm using twitter but not email?
|
|
|
|
martok (OP)
|
|
May 31, 2011, 07:41:14 PM |
|
monitordrop({id => monitorid, worker => 'workeraddr'}) Cancels a monitor. Get the ID from the generated email.
How can I get the monitor ID if I'm using twitter but not email? Yeah that's a problem. I'll adjust the twitter msg to include a disable url sometime today. If you're removing a worker though, you'll only get alerted once when it goes off. Point taken nonetheless.
|
|
|
|
Transcendental
Newbie
Offline
Activity: 14
Merit: 0
|
|
May 31, 2011, 09:01:59 PM |
|
I have been able to login using "worker;payto" now for some reason.. Although I had tried quotes before and it was not working. My problem now is how do I get an accurate hashrate for the workers combined, as well as unconfirmed balance? If you do an rpc-call to the payment address, it does not properly display either clienthashrate(paytoaddr) or currentbalance(paytoaddr)
|
|
|
|
|