http://www.ligo.caltech.edu/~misi/summary/day/20150504/
I left the arms locked last night. Looks like the drift in the Y arm power is related to the Y arm control signal being much bigger than X.