At 0617utc (1117pmPT), H1 dropped out of OBSERVING due to a fast SQZ sdf diff (I took it back within 45seconds). 

Tony and I were chatting at the time and I mentioned the situation and he pointed me to some recent (within the last week) similar instances where this happened with these Camilla alogs:  71653 & 71652.

Attached is a plot showing the same culprits (H1:SQZ-FIBR_SERVO_COMGAIN & H1:SQZ-FIBR_SERVO_FASTGAIN) dropping to lower value for 0.5secs and then returning to their normal value. 

Do we need to not monitor these channels?  Make an FRS?