I can only speculate that it has to do with the miners software on those things which doesn't seem to like the pool
I 100% agree. Somehow I think its linked to that "garbage string" they seem to send to stratum (from what i've read)....
I wonder if someone compiles a fresh cgminer for them if it would fix the problem.
The strange thing is it seems to increase in severity with hashrate.
My S3's rarely saw downtime, and it was always happening on any pool anyways.
Ive heard of the S5's having issues staying connected on the pool
And worse with S7 Etc...
I wonder whats so different between your stratum and kano....