I mine on zPool and noticed that a couple of my Z9 Mini's were missing from the pool. When I checked the miners, they were stuck mining on the dev pool and had been doing that for over 12 hours and not flicked back! How do I stop this from happening again? This is way more than the stated 1%.
Hmmm 1%

? not even close - I use awesome miner to run my miners and what its reports show me that zec-bj.poolin.com (I do not mine on that pool so I assume that is a dev fee pool) in the last 7 days was mined on 5.98% and 6.7% in last 24 hours.
I will provide more details on it later. There is something really strange that is going on with 2.1c. I am not 100% sure if that has to do with new awesome miner update or 2.1c I have dedicated 5 minis and 5 bigies to run some tests and get to the bottom of it and will provide mining history reports shortly. But so far I definitely noticed increased traffic to poolin.com on the firewall; Slush poll mining became really unpredictable - meaning some miners will be reported on the pool side some will not even though awesome miner and miners web gui will show them mining with 0 rejects on Slush. (may be it is what spoonwzd described- I will check on this) Also miners started to get disconnected more often due to socket error and even though they haven't been rebooted when they re-connecte they will start dev cycle again... On the side note: Sorry Jason, I am not trying to bash your work and definitely not trying to accuse you in being dishonest about 1%. For all I know it might be just my set up and I apologies in advance if it is - I will post an update with reports shortly
I had a similar issue and notified Jason a few weeks ago. The miner was connecting to his callback server and dev server more then it should. I was using the netstat on the machine to monitor its connections.
Thank you for your feedback and assumptions on methodology and similarities. When you tried to test it was when dev fee disabled other than the startup sequence. There were other status callbacks for testing and simulation but did not change the miner to a dev pool. That is what you were seeing. A single problem was found where an accidental dev fee happened after 12 hours before it was officially turned on. That was resolved and the user who found it compensated.
A second bug was found where the wrong pool could be returned to, which led to 2.1c.
I would ask you to also provide your supportID.
Ill let spoonwzd post of his own accord if he desires as to the status and any potential resolutions, but he provided a supportID and we are working through the conditions in his environment.
Jason