As the IFO went into oscillations that have historically indicated it's headed for lockloss, I noticed a peak in MICH and some increased noise in PRCL around 15Hz.
I ran the DRMI_FOM template with 150 averages starting when the oscillations appeared until lock loss, 21:50UTC to 22:01:55UTC (Lockloss was at 22:02:24UTC).
DRMI displacements plot attached.