I checked the OBSERVATION.snap of LSC, ALSPDH, ALSPLL, PSL, IMC, IMCASC, and OMC models and changed the following channels to NOT_MON channels in OBSERVATION SDFs.

LSC model:

K1:LSC-AS_TRNORM_PDA1_RF17_Q_OFFSET:

This channel is managed at the ENGAGE_DARM_OFFSET state in the LSC_LOCK guardian.
Since values are calculated from IMC output power during the lock acquisition, this channel should change in each lock, so it should be NOT_MON channels.

K1:LSC-PD_DOF_{MTRX,MTRX_SETTING}_1_29:

These channels are managed at the HANDOVER_TRAS_TO_OMC state in the LSC_LOCK guardian.
Since values are calculated from IMC output during the lock acquisition, these channels should change in each lock, so they should be NOT_MON channels.

K1:LSC-PSL_PWR_SCALE_OFFSET:

This channel is managed at the SET_LASER_POWER_NORM state in the LSC_LOCK guardian.
Since values are calculated from IMC output power during the lock acquisition, this channel should change in each lock, so it should be NOT_MON channels.

ALSPDH model:

K1:ALS-{X,Y}_PDH_OFS_SLOWOUT_CALI_OFFSET:

This channel is managed at the FIND_IR_RESONANCES state in the LSC_LOCK guardian.
Since values are decided so that both IRX and IRY are flashed during lock acquisition, this channel should change in each lock, so they should be NOT_MON channels.

ALSPLL model:

K1:ALS-SUM_OFS_SLOWOUT_CALI_OFFSET:

This channel is managed at the FIND_IR_RESONANCES state in the LSC_LOCK guardian.
Since values are decided so that both IRX and IRY are flashed during lock acquisition, this channel should change in each lock, so it should be NOT_MON channels.

K1:ALS-{X,Y}_LASER_TEMP_BIAS_OFFSET:

These channels are managed at the SCAN_TEMP_BIAS state in the PLL{X,Y} guardians.
Since values are decided so that PLL can be locked during the lock acquisition, these channels should change in each lock, so they should be NOT_MON channels.

PSL model:

K1:PSL-PMC_{COMMON,FAST}_GAIN:

These channels are managed at the DOWN state in the PMC guardian.
Since values are calculated from the output power of laser amplifier during the lock acquisition, these channels should change in each lock, so they should be NOT_MON channels.

IMC model:

K1:IMC-CAV_{REFL,TRANS}_POW_NORM and K1:IMC-SERVO_IN1GAIN:

These channels are managed at the LOCK_PREP state in the IMC guardian.
Since values are calculated from the output power from PSL room during the lock acquisition, these channels should change in each lock, so they should be NOT_MON channels.

IMCASC model:

K1:IMC-PSLOUT_POW:

This channel is managed at the LOCK_PREP state in the IMC guardian.
Since values are calculated from the output power from PSL room during the lock acquisition, this channel should change in each lock, so it should be NOT_MON channels.

OMC model:

K1:OMC-PZT_HV1_OFFSET:

This channel is managed at the  FIND_RESONANCE_FOR_10W state in the OMC_LSC guardian.
Since values are decided so that OMC can be locked with TEM00 during the lock acquisition, this channel should change in each lock, so it should be NOT_MON channels.

I also checked the OBSERVATION.snap of VIS models and changed the following channels to NOT_MON channels in OBSERVATION SDFs.

IMMT1, IMMT2, PR2, PR3, PRM, BS, SR3, ITMX, ITMY, ETMX, ETMY models (fig1-11):

K1:VIS-{IMMT1,IMMT2,PR2,PR3,PRM,BS,SR3,ITMX,ITMY,ETMX,ETMY}_TM_SET_{P,Y}_OFFSET:

These channels are managed at the RESET_ASC_FEEDBACK state in the LSC_LOCK guardian.
At this state, the channel values are overwritten with the GOOD values, which continues to be updated after PRFPMI was locked.
So, these channels might change in each lock, and therefore they should be NOT_MON channels.

I checked OBSERVATION.snap of PSL and GRDCONFIG models and chenged the following channels to NOT_MON channels in OBSERVATION SDFs.

PSL model (fig1):

K1:PSL-PMC_{TRANS,REFL}_DC_POW_NORM and PSL-REFCAV_{TRANS,REFL}_POW_NORM:

These channels are managed at the DOWN state in the IO guardian.
Since Values are calculated from the current fiber laser output during the lock acquisition, these channels should change in each lock, so they should be NOT_MON channels.

GRDCONFIG model (fig2):

K1:SYS-HWP_{PSL,IFO_REFL}_STEP_TOTAL_DEG:

Since values recoarded in these channels are the current HWP angles measured with the encoders, these channel might change in each HWP rotation.
So, they should be NOT_MON channels.

K1:SYS-HWP_REFL_STEP_DEG:

This channel is managed by the INCREASING_LAS_POWER in the LSC_LOCK guardian.
Since values are calculted from the target angle (fixed value) and current angle (measured value) during lock acquisition, these channels might change in each lock, so they should be NOT_MON channels.

k1grdconfig (Fig.1)

K1:SYS-LOCKLOSS_{LATEST,PREVIOUS}_{DURATION,GPS,JST,KNOWN_ISSUE,PREV_STATE,RELATED_NODES,UTC}

K1:SYS-LOCKLOSS_ENTER_TIME

k1detdq0 (Fig.2, 3)

K1:DET-IFO_LOCK_DURATION_RECORD

K1:DET-{CFC_LATCH,OBS_INTENT}

K1:PEM-{TEMPERATURE,HUMIDITY,BATT,RSSI}_$(LOCATION)

K1:VIS-$(OPTIC)_TEMPERATURE_{CHAMBER,HEATER}_PV (Fig.5)

K1:CAL-PCAL_{X,Y}_LASER_.* (Fig.6)

k1calex (Fig.1)

K1:CAL-PCAL_EX_*_SET, K1:CAL-PCAL_EX_CURRENT_TIME

k1caley (Fig.2)

K1:CAL-PCAL_EY_TCAM_*

k1visetmxlsc (Fig.3)

K1:VIS-ETMX_TM_CAL_LINE_FREQ

Indeterminable changes

K1:VIS-ETMX_TM_CAL_CLKGAIN

Original post is written in klog33985.

---------Copy from the original post----------

I checked the SDF status about the PEM channels for the observation state.

1. K1:PEM-*_SHINDO
Fig.1.
in the model of the k1pemmanage, they monitored the shindo channels.
But those values changed during observation, so I removed from monitor channel.

I also changed confirmed the change of the channel status by the current work of the klog33950 and so on
Fig.2. - Fig.6. showed the SDF before accept

Due to the modification of LSC_LOCK guardian (klog34002), OMMT2 and OSTM alignment was changed in each lock.
So, I changed the following channels to NOT_MON channels.

OMMT2 model:

K1:VIS-OMMT2_TM_OPTICALIGN_{P,Y}_OFFSET:

These channels are managed at the RECORD_GOOD_VALUES_OMC state in the LSC_LOCK guardian.
Since values are calculated from K1:VIS-OMMT2_TM_SUMOUT_{P,Y}_OUT16 during the lock acquisition, these channels should change in each lock, so they should be NOT_MON channels.

OSTM model:

K1:VIS-OSTM_TM_OPTICALIGN_{P,Y}_OFFSET

These channels are managed at the RECORD_GOOD_VALUES_OMC state in the LSC_LOCK guardian.
Since values are calculated from K1:VIS-OSTM_TM_SUMOUT_{P,Y}_OUT16 during the lock acquisition, these channels should change in each lock, so they should be NOT_MON channels.

I changed following channels to NOT_MON channels.

guardconfig model (fig1):

K1:MIF-FINESSE_*:

These channels are used for finesse measurement and updated by finesse measurement scipts.
Though these channels are not updated without finesse measurement, I changed them to NOT_MON channels because they doesn't affect the OBSERVATION state.
In addition, since finesse measurement is done when IFO is down, so update of SDF might be done for long after finesse measurement, and it increases the risk to forget to make OBSERVATION snap accept.

sdfmanager model (fig2):

K1:CRY-*_TEMPERATURE_LASER1_PV:

These values are the readback from the thermometer installed at MN OpLev lasers, so they varies in time and should be NOT_MON channels.
 

k1sdfmanage

K1:AIR-*_{M1,M2} (Fig.1)

They are temperature readback channels of precision air conditioners. They removed from monitored list in observation.snap. K1:AIR-.*_S1 are the set point of temperature control and they are changed in human activity. So they are kept in monitored channel list. Though remaining channels (K1:AIR-.*_{J0,ER}) are readback channels, they are operation status and error status of air conditioner system. Because change in these channel should be detected, they kept in monitored list.

Additional unreported differences on k1calex, k1caley, k1visetmxlsc Fig.2-4

They cannot be confirmed because there is no klog report. But we couldn't check the operation above IFO LOCKED if these SDF differences remains before starting full calibration measurements from tomorrow morning. So we temporarily accepted them with snapshots of SDF screens (Fig.2-4) in order to verify the behavior of IFO status flag. Changes in Fig.2-4 must be much more carefully checked later than another channels which are properly reported in klog.

I apologize for the lack of klog entries regarding the values shown in Fig.2 to Fig.4.
The values that were temporarily accepted were set by me, and I confirm that all of them are correct.

k1sdfmanage (Fig.1)

K1:STEPPER-ITMY_GAS_0_{ACTUAL,TARGET}_POSITION

k1visitmyt (reverted)

I accepted the changes in k1ascbpc models (fig1).
These changes are related to the work reported in klog34021.

I found that the values of K1:LSC-CARM_SERVO_OFS_COM_CALI_OFFSET in OBSERBATION snap is different from the one defined at the DOWN state in the VERTEX guardian.
So, I accepted the value as shown in fig1.

I accepted following SDF differences.
 

k1calcs

K1:CAL-MEAS_{LATEST,CURRENT} (Fig.1)

They are the record for the history of the full and weekly calibration measurements. They were updated due to today's calibration works in klog#34033.
 

K1:CAL-CS_TDEP_{PCAL,SUS}_LINE*_{ERR,EXT,SUS,PCAL}_DEMOD_SIG (Fig.2)

These are the change in the band-pass filters for demodulating calibration lines. Filter design had been already updated in klog#33954. But I forgot to enable these filters and to disable old (O4a) filters. Today, I switched and accepted them.
 

K1:CAL-CS_TDEP_* (Fig.3)

Sorry I don't listed up one-by-one because it's so many. But all of them are listed up in JGW-L2314962, which are planned changes by periodical calibration works during observing run. These changes are the reference transfer function at line frequencies in order to use estimating online time-dependent coefficients.

k1pemmanage

K1:PEM-SNOW_*, K1:PEM-*_SHINDO, K1:PEM-WATER_* (Fig.1)

Following updates are acceppted on observation.snap.
 

k1calcs

K1:CAL-CS_TDEP_{SUS,PCAL}_LINE?_REF_* (Fig.1)

They are reference transfer function at line frequencies for tracking calibration lines. They had been accepted once as tentative values for some test. In this time, they updated again with finalized calibration parameters.
 

K1:CAL-CS_TDEP_PCAL_LINE?_CORRECTION_{REAL,IMAG}, K1:CAL-CS_TDEP_PCAL_LINE?_PCAL_DEMOD_SIG (Fig.2)

These values are the pcal correction factors for line tracking. For simple management of line tracking parameters, I merged gain in filterbank was merged to EPICS records.

Following updates are acceppted on observation.snap.
 

k1 STEPPER

K1:STEPPER-ITMY_GAS_4_ACTUAL_POSITION

K1:STEPPER-ITMY_GAS_4_TARGET_POSITION

These values are the stepper motor positions changed to offload the BF GAS in IY. I accepted the change before returning to Observing.

K1:CAL-CS_TDEP_PCAL_LINE{1,2,3}_*, K1:CAL-CS_TDEP_SUS_LINE3_* (Fig.1)