Finally got my p2pool node setup and my rigs seem to be hashin faster than when on a pool
Slightly in reality. In p2pool, ping makes a major difference. That is why people really ought to test ping below 50ms for optimal performance. If you setup your own local node, then your ping is the best possible, and your reject rate really decreases, so your "net" hashrate is in fact nicely higher.
Mining someone's p2pool, even on low ping, your registered ping is higher, but so is the reject %, and p2pool dont "subtract" that from your "net" hashrate, so its kind of an illusion. Also, p2pool, you can have waves of higher than expected hash, but also the opposite.
Manually adjusting your diff rate makes a huge difference. So having all this into account, any form of p2pool definitely does perform better than a centralised pool.
I noticed with sgminer 5 and running darkcoin-mod kernel I was getting really high hw errors and I had used the exact miner and config on the pool and had no issues but I downloaded another build and whah-la perfection. My reject rate is also really low too and I'm running my node locally so the latency is extremely low.
I set up a node because I live outside of Atlanta and there's no nodes close to me and when I connect to the closest nodes my rejects were insanely high. I had been meaning to get around to setting it up but finally got around to it last night. The only problems I ran into was some commands had changed from the guide I was reading but other than that it was actually a lot smoother than I anticipated.
From memory, some of the SGMiner versions needed to be set-up quite differently depending on what drivers you were running. I remember one set of drivers hating dual threads but you could push it quite hard as a single thread, and visa versa for other versions of ATI Catalyst.
