Corey & Hugh
Although we don't have to take the Pumps down to grease them, I used the need to grease as excuse to practice taking them down.  We changed the ACC & DECelleration times from 1 to 300 seconds to allow the other running pumps to compensate for the one spinning down.  This combined with closing the outlet valve to prevent the pump from reversing was not sufficient however to eliminate glitches in the pressure and the Actuators.  Attached are trends from the Pump System and BSC2.  You can see the motor drive changes, CH2, during the transitions from 4 to 3 pumps and back; as well as, the pressure changes usually at the beginning or end of the transition but not just there.  Also, looking at the other channels on the graph, you can see glitches in the local IPS and Cartesian channels of the BS HEPI, Cartesian CPS channels and BS ISI Witness channels.   In the second attachment I've zoomed into the second transition from 3 back to 4 pumps.  This is when we had the largest glitch in the pressure and you can see it strongly make it into the M1 Witness channels.  The glitches didn't show as strongly in the velocity sensors as much so I don't include any of those although I wouldn't say they were unaffected.
The first plot is 2 hours long and we were on 3 pumps for about 45 minutes as Corey and I figured out how to make a grease gun work.  During the 3 pump running time, it seems many of the HEPI & CPS signals are noisier.  Happy to accept guidance on evaluating if this is a configuration we want to avoid, that is, 3 pumps versus 4.