J. Kissel, S. Dwyer

Since we now believe the end station ISC and ALS front-end models and their respective settings are pretty stable, we've turned on the SDF monitoring systems. So far, the only channels I've *removed* from the monitoring system are
${IFO}:ALS-${ARM}_CAM_ITM_PIT_POS
${IFO}:ALS-${ARM}_CAM_ITM_YAW_POS
${IFO}:ALS-${ARM}_CAM_ITM_SUM
${IFO}:ALS-${ARM}_WFS_SWITCH
where the first three are actually the camera's software output being constantly jammed into an EPICs setting (so it's not really a setting, just a monitor); and the the on-off switch for the green WFS, which should be controlled by Guardian. I would post a screen shot of the DIFF screens, but there are no DIFFs, so it's not very exciting.

I've also edited (and committed) the 
/opt/rtcds/userapps/release/sys/common/medm/SDF_OVERVIEW.adl
screen to (a) include the new ALSEX and ALSEY, and (b) change the DCUIDs of ISCEX and ISCEY, which are now one DCUID higher to make room for the new ALS models.

Details:
Steps to get these set up:
(1) Burt restore these systems to 12p PDT last night to ensure that we have good -- or at least something that might be good -- settings to use as our starting point.
(2) Create / update safe.snaps in the userapps repo using makeSafeBackup, e.g. 
makeSafeBackup als h1alsex
For the record, these files live in:
/opt/rtcds/userapps/release/isc/h1/burtfiles/h1iscex_safe.snap
/opt/rtcds/userapps/release/isc/h1/burtfiles/h1iscey_safe.snap
/opt/rtcds/userapps/release/als/h1/burtfiles/h1alsex_safe.snap
/opt/rtcds/userapps/release/als/h1/burtfiles/h1alsey_safe.snap
(3) Ensure the safe.snap in the burt folder of the target area for each model is a softlink to that repo file.
]$ ls -1l --color=auto h1{als,isc}e*/*epics/burt/safe.snap
Feb  3 10:42 h1alsex/h1alsexepics/burt/safe.snap -> /opt/rtcds/userapps/release/als/h1/burtfiles/h1alsex_safe.snap
Feb  3 10:43 h1alsey/h1alseyepics/burt/safe.snap -> /opt/rtcds/userapps/release/als/h1/burtfiles/h1alsey_safe.snap
Mar 10 17:36 h1iscex/h1iscexepics/burt/safe.snap -> /opt/rtcds/userapps/release/isc/h1/burtfiles/h1iscex_safe.snap
Mar 10 17:36 h1iscey/h1isceyepics/burt/safe.snap -> /opt/rtcds/userapps/release/isc/h1/burtfiles/h1iscey_safe.snap
Note, this took some effort, because for some strange reason the ALS auto-generated target folders had different permissions (controls only writeable) than the ISC auto-generated target folders (which had controls and controls group writeable). A little logging in as controls and chmod g+w on the folders did the trick.
(4) Use /ligo/cds/userscripts/sdf_set_monitor.py to add a "1" to the end of each burt setting line, 
]$ sdf_set_monitor 1 /opt/rtcds/userapps/release/als/h1/burtfiles/h1alsex_safe.snap
(5) Load the new safe.snap into the front end as the accepted table of good values but hitting "LOAD TABLE ONLY" on the SDF_RESTORE overview screen for each model.  
(6) Find a few channels that you want to ignore, and remove them from the table,
]$ sdf_set_monitor -c - 0 h1alsex_safe.snap
(channel name, hit return)
(channel name, hit return)
(hit ctrl+d to send'er)
I recall Jonathan / Dave / Betsy working on a different way to use this script, or another script entire that accepts the chanels themselves instead of a file, but we couldn't find a log, and Jamie the python pro was sitting next to me, so he made a few changes to the script and we did it this way. The changes aren't committed, because the script doesn't live in a repo.
(7) Have Sheila test your ability to update the database to accept new changes by having her change a few now-monitored settings, or have Stefan test you by burting an older file and changing vitrually all the settings from what you've stored. Do so, from the SDF OVERVIEW for that model, and at the bottom, for
FILE TYPE SELECTION chose EPICS DB AS BURT
FILE OPTIONS SELECTION chose OVERWRITE
hit SDF SAVE FILE
(wait a second, watch for the "TIME" channel, H1:FEC-${DCUID}_SDF_RELOAD_TIME, to report a modified SDF file)
hit LOAD TABLE ONLY
(8) commit your newly minted SDF monitored safe.snaps to the userapps repo.

Note, I know this will be fixed in the RCG 2.9.1 bug fix / upgrade, but the H1:FEC-${DCUID}_SDF_DIFF_CNT -- the number of differences before the loaded SDF table and what the front end has as settings has a hard limit at 40, even though there may be more than 40 differences. I experienced this when Stefan pushed a burt restore of iscex and iscey. Many channels were different than when Sheila told me was a good time -- many (or at least one) more than 40 channels were different. We didn't know whether we liked midnight last night or an hour ago, so we just went with last night.