For the record, 10 minutes dev and 280 minutes normal is 3.4%, not 1%. I never claimed 1%. Im just getting home now and will look to investigate this shortly. It is 49.6 minutes a day.
Iceman, if you can PM me your supportID it would be great also. What 2.1c primarily changes is simply to ensure that it returns to the proper pool after the dev-fee. There were circumstances where it could return to one of your other pools instead. It was literally a one line change for that since I had some logic reversed accidentally.
Nothing else was changed that would match what you are seeing, for what it is is worth.
-jason

and I never said you did - I actually wasn't even paying attention to what it was supposed to be I just simply noticed the oddities on the z-farm and started looking into whats going on. When I mentioned 1% I simply wanted to say that I am not accusing you in dishonesty - my apologies if that didn't come out right. So..... like I mentioned in my post I set 5 big ones and 5 small ones on 2.1c to run on equihash overnight- disabled all custom scripts, rules etc. on awesome miner side. The only one I left running was a script that checks the pool after disconnect and if it is not correct one it switches it via SSH. I choose equihash just for the reason that it is the most screwy one with a lot of disconnects and rejects. The dev fee came out to be 4.11%
http://www.u-ok.net/images1/miningrep1.png Looking through some of the logs I can see that a few times miners got socket error and when reconcted they came to the pool that is specified in the first row of the set up thats is the reason you see a few other pools beside nicehash. One thing I can say for sure that the issue with not returning to the last pool mined on after dev cycle is definitely fixed and it doesn't matter if last pool was switched to using SSH command or it was just a default set up. Now it works the way it should. I will set the same bunch to run 24 hours on the luckpool mining hash this time - that should be much more stable compare to nicehash and will post results again tomorrow. I know you guys talked about temps and hash rate so here are a few graphs (2 big ones and 2 small ones on 2.1.c) that displays hash rate vs temperature. If anyone is interested i can post a larger sample which will have less spikes and easier to see correlation between two.
http://www.u-ok.net/images1/perf1.png http://www.u-ok.net/images1/perf2.png http://www.u-ok.net/images1/perf3.png http://www.u-ok.net/images1/perf4.png