The goal of the shift was to relock the ITF up to CARM offset zero, in order to start the ICS/TCS tuning of the ITF working point. Unfortunately this could not be done, as we never got past 300 mW of arm powers:

• at the beginning of the shift we had quite some hard time in locking the DRMI; we soon found that the IMC was out of place of about 800 um, as FmodErr was badly mistuned; we proceeded to move the MC_F0_Z setpoint but we soon found that the IMC cannot be unlocked anymore with the proper relay (or, equivalently, the VPM button); we moved the IMC anyway, just discharging the thermal correction every couple of steps (FIgure 1);
• after this, we could relock the DRMI quite easily, after the usual pre-alignment; at this point we started to suffer from several unlocks, all of them around 200 mW of arm powers; basically all of them were caused by the 9-10 Hz oscillation (Figures 2 to 6, UTCs in the filename);
• during these tests, we observed that the DARM gain seemed too low once the DOF is controlled with the B1p_56MHz_I_ARM_NORM signal; we saw, looking at the DARM UGF line and also the DPHI monitor, that the demodulation phase was quite mistuned (~1 rad); we proceeded to re-tune it at every step of the CARM offset reduction, as this has been seen to change; however, in the last (and best) trial we made, it was more or less optimal at 3.85 all the way (this is default in the locking script); to be rechecked in next attempts, especially at 300 mW or arm powers; the same mistuning was not apparently present in the QPD signals from the same bench;
• later the SDB2_SBE vertical control opened, and it was quickly restored; the cause of this is not clear yet;
• in the last trial, things seemed to work better keeping a higher gain for the SRCL loop w.r.t. previous attempts (manually, 0.5 @ 100 mW, 0.6 @ 200 mW, 0.8 @ 300 mW); given there is no statistics in this, it's hard to conclude this was effective, but it should be checked.

We leave the arms locked on the IR beam.