Quote from: Bitmong on May 29, 2013, 02:07:40 AM

Quote from: gyverlb on May 29, 2013, 01:18:40 AM

Quote from: Bitmong on May 29, 2013, 12:44:30 AM

When the getblocktemplate latency started to appear, my efficiency was still between 110-115%. My getblocktemplate latency was about 30 seconds at that time.

This is the worst latency I've seen reported so far (by nearly 3x, the worst I can rember was ~12s). With <0.8.2rc3 getblocktemplate depends heavily on you CPU speed and number of transactions in your memory pool. As you seem to have an adequate CPU, it probably doesn't explain the difference. Do you use non-default values in your bitcoin.conf?

I didn't change any block size / tx fee settings in bitcoin.conf, only rpc settings and server settings.

Actually the latency was around 30 seconds when I used the bitcoind on that computer or when I set P2Pool to use the bitcoin-qt on my Windows computer with Core i7-3820 processor. So, the latency was definitely not caused by the CPU.

Maybe I just upgraded soon enough to not see these kinds of latencies (it was just below 10s at the worst).

Quote from: Bitmong on May 29, 2013, 02:07:40 AM

Quote from: gyverlb

You have relatively high bandwidth, but what is your link latency? If you use traceroute/mtr or similar tools, what is the time-to-hop for your ISP main routers and addresses in North America/Europe/China (where most nodes and probably miners are according to http://blockchain.info/fr/nodes-globe).

I live in Finland btw.

MTR shows these results for some nodes:

First ISP router 2ms
Last ISP router 4ms
US (Sayreville) 111ms
Japan 294ms
Australia 430ms
Hong Kong 340ms
Netherlands 33ms
China 330ms
Switzerland 44ms
Ukraine 57ms

These looks good, I have better European and China results (most of the time <30ms in Europe and ~220ms with China) but I'm puzzled (I don't think it can explain our differences in efficiency).

Did you restart your p2pool node recently and let it run 24h to have a more accurate stat (efficiency isn't accurate with long runs : it compares your own rates of DOA/orphans since the p2pool process start with the average over 24h of the network).