Sitting down to start cavity measurements, I found both ITMs tripped.  It must have happened a while ago (I didn't bother to check dataviewer trends) because both had rms levels of <5 counts, so they've had a while to sit and quiet down.