I manually adjusted the IMC IN1 + IM2 gains and IMC fast gain, so that IMC IN2 gain is back to -29dB and IMC FAST GAIN is -16dB, and set the guardian to do this for our next locks in the LASER_NOISE_SUPPRESION state.

Some history:

On the 12th of September, Jim, Daniel and I redistrbuted gains in the IMC CM servo to increase the range it would have to avoid saturating at split mon during times of high ground motion 79998.  Starting of the 17th of September we started to have locklosses with large glitches visible before the lockloss in the FSS fast mon channel, tagged as FSS oscillation by the lockloss tool, which became more frequent over time, and these locklosses also have the IMC loosing lock at the same time as the IFO, which did not happen earlier in O4.  October 9th Camilla found that the gain redistribution corresponded with the start of the locklosses where the IMC loses lock at the same time as the IFO, 80561, so we reverted the change.  At first this seemed to help us have longer locks, but over time the locklosses again became more frequent.  After the change was reverted we do not see the signature of the glitches happening in the FSS channel as clearly, so we've been relying on the new IMC tag that Camilla added to the lockloss tool. 

Now that the NPRO has been swapped, we still saw 6 locklosses overnight with the IMC tag, lockloss website.  These do not have the large glitches in the FSS fast mon channel, but we wanted to try reverting the gain slider change to see if the glitches with the new NPRO appear in the same way that they appeared with the old NPRO and when the CARM loop has more actuation range.  We don't expect this to help us stay locked longer, it may make the locks shorter, but we hope to learn something from it and will probably undo the change once we get a couple of locklosses.