Quote from: Krak on January 24, 2014, 08:09:14 AM

Quote from: ZachB on January 24, 2014, 08:00:28 AM

I've searched for an answer to this, but I haven't found it...

I'm running a P2Pool node. I realize I can set difficulty using address+XXXX where XXXX is a user specified difficulty.

Is there any reason other than seeing a 'smooth' graph to set difficulty any lower than the share difficulty? My Pseudoshares at difficulty 128 or whatever really mean nothing other than it updates the graphs showing my miner is doing work.

If I were to set the pseudoshare difficulty to the share difficulty [currently 873000], is there any downside? I'd save in some bandwidth, seems like it would save all the trouble of sending shares in when they don't really mean anything. I'd edit the code so as share difficulty fluctuates the difficulty for pseudoshares would fluctuate as well sent to my miners...Is this a crazy thing to do?

Thanks,

Zach

If you don't mind losing the stats, you can certainly set it that high. There are no other downsides. Are your miners on a separate network from the p2pool node? I don't see why you'd need to save bandwidth if they were on the same network.

They're actually not on the same network, the P2Pool node is hosted at a large datacenter, my miner is at home on a business broadband connection. I'm about 64ms rtt away. I'm mining with about 450 GH/s. I was thinking the ASICs would be more busy mining and less busy reporting to cgminer 'results' -- IE, I'd keep them heads-down mining by not having them report proof of work as often. But I suspect the reporting every few seconds probably isn't much of a slowdown, with new work every ~30sec, they have to pause a bit anyway...

Thanks for the input, I was mainly trying to verify my understanding -- mostly academic, I kinda like knowing from the graphs it *is* doing something by reporting

--Zach