...
Thanks Kano. I remember and appreciate your input back in August, however my concern is not entirely that my miner will completely stop. It's that I'll be operating in a reduced capacity for some time and not realize it. This is why I check the hash rate, and not the time since last share. It's my understanding that even if one blade of my miner goes down, I'm still operating at 66% capacity and will be submitting shares. Would time since last share field still be a good indicator in this situation?
Nope, in that case it wont help you with any certainty.
The pool adjusts the share rate to be 18 SPM, so the share diff changing by a large amount would be one way to identify it.
However, I'm not sure how quickly that changes, and it also has to be over a certain % before ckpool will change it.
The real issue though is that you can't tell, conclusively from the shares submitted, that a miner hash rate has dropped, for quite a while after the event.
The board failing would be known to cgminer, but that also depends on what Bitmain has is their API answers.
I guess the problem here is buying unreliable hardware and trying to make it less unreliable ...
Maybe Bitmain should consider making reliable hardware, with something like 100Million$ they make each year ...
The pool hash rate is based upon shares, so you can only detect a partial failure based on a slow to determine hash rate.
Monitoring the cgminer API, assuming Bitmain has made the appropriate information available, is your only way to identify that quickly.