10 of my 30 L3s won't mine - since Nicehash went down.  They have a solid green light, the one that blinks when things are operating normally.   Has anyone seen this before?

These ten stopped mining when Nicehash went down and didn't switch to the next pool, instead they were stuck at 0.00 hash rate.    I have 20 other L3s that all switched to the next pool.

I reconfigured the pools (removing nicehash) and soft rebooted them.  Still would not start mining, "Socket connect failed: Connection refused".   So, I hard booted five of them - now every single one of those five has a solid green light, ping up, gui is responsive, but will not mine.   The status page says "Socket connect failed: Connection refused".

These 10 bad L3s are split between two different facilities miles apart, with different internet, power and cooling.    I have five down at each facility.  The thing they have in common is that all ten of these were received from bitmain in the past 30 days.   None of these were being overclocked.  

I also re-flashed the firmware on one of them - it didn't help.  Same thing.

I have to drive to the other facility to see if the non-hardbooted L3s also have a solid green light.

Any ideas?

Thanks!