So, I made my p2pool log public at
www.nogleg.com/log (times displayed are in CST, -5 UTC/GMT. it gets reset periodically).. no sensitive information there, only IPs shown are other p2pool nodes that anyone could see themselves by running p2pool. The addresses shown getting shares can be seen by going to my pool anyway & the amount they are receiving is displayed publically in the 'payouts if block is found now'.
Anyway,
There are a few incoming connections that are incredibly annoying that I've considered firewalling, but then I think it's possible they may fix their pool at some point, so..
I assume the stuff about 'peer sent entire transaction xxxxxxxxx that was already received' means that I got person X's share before person Y's (and sometimes person Z, Z2, etc)?.. so orphan is incoming? (ed: well, for my share that was just orphaned, I didn't see anything about that... just from timestamps that the Mav one was about 150ms earlier)
What do the transactions pulled from latency queue signify?
Also, wouldn't a p2pool node w/ many people connected have problems with the 'new work to worker' stuff? It looks like each takes about 5ms or so. I usually run my p2pool on ramfs, but after some recent shenanigans where I lost about 12hrs of data, I stopped... does that make that new work process quicker?
anyone know about any of those?
(ed2: i just changed incoming connections max to 1. maybe some superstition about those IPs that just connect and timeout over and over, but didnt like the start)
(ed3: this is info from a pool other than my own, a share that I had orphaned)
mine:
Time first seen: Thu May 02 2013 08:42:42 GMT-0500 (Central Daylight Time) (1367502162.504647)
1CTgYxMTY5j6SLytKeMsBWAXuUc6yNKcAe:
Time first seen: Thu May 02 2013 08:42:44 GMT-0500 (Central Daylight Time) (1367502164.585231)
a second opinion at another pool:
mine:
Time first seen: Thu May 02 2013 08:42:42 GMT-0500 (Central Daylight Time) (1367502162.457166)
1CTgYxMTY5j6SLytKeMsBWAXuUc6yNKcAe:
Time first seen: Thu May 02 2013 08:42:44 GMT-0500 (Central Daylight Time) (1367502164.538657)
... wouldn't the base client have rendered that share DOA (as it was over 2000ms late), even though 1CTgYxMTY5j6SLytKeMsBWAXuUc6yNKcAe did happen to get the following share about 10 seconds later..... or dies it resurrect it from the dead to send out both to everyone?
... and why does someone with 60ghash+ have 2+ second lag time?