Collected OBSERVE.snap files for use in full observing configuration monitoring.
With the SDF happy (green) and Guardian Nominal (Robust Isolated) and green, I made an OBSERVE.snap file with SDF_SAVE screen
choosing EPICS DB TO FILE & SAVE AS == OBSERVE
This saves the current values of all switches and maintains the monitor/not monitor bit into the target area, e.g. /opt/rtcds/lho/h1/target/h1hpietmy/h1hpietmyepics/burt as OBSERVE.snap
This file is then copied to the svn: /opt/rtcds/userapps/release/hpi/h1/burtfiles as h1hpietmy_OBSERVE.snap and added/commited as needed.
The OBSERVE.snap in the target area is now deleted and a soft link is created for OBSERVE.snap to point to the svn copy:
ln -s /opt/rtcds/userapps/release/hpi/h1/burtfiles/h1hpietmy_OBSERVE.snap OBSERVE.snap
Finally, the SDF_RESTORE screen is used to select the OBSERVE.snap softlink and loaded with the LOAD TABLE button.
Now, for the HEPIs for example, the not monitored channels dealt with by the guardian will be a different value from the safe.snap but, the not monitored channels are still not monitored so the SDF remains green and happy. And if the HEPI platform trips, it will still be happy and green because, the not monitored channels are still not monitored.
What's the use of all this you say? Okay, I say, go to the SDF_TABLE screen and switch the MONITOR SELECT choice to ALL (vs MASK.) Now, the not monitored channel bit flag is ignored and all records are monitored and differences (ISO filters when the platform is tripped for example) will show in the DIFF list until guardian has the platform back to nominal.
Notice too that the SDF_OVERVIEW has the pink light indicating monitor ALL is set. This should stay this way unless Guardian is having trouble reisolating the platform and then the operator may want to reenable the bit mask to make more evident any switches that guardian isn't touching more apparent.
But rather than rely on selecting ALL in the SDF_MON_ALL selection, I would suggest you actual set the monitor bit to True for all channels in the OBSERVE.snap. That way we don't have to do a two-step select process to activate it, and we can indicate if there are special channels that we don't monitor, for whatever reason.
Yes Jameson. That is why I selectied the ALL button allowing all channels to be monitored.
Hugh, I think the OBSERVE snaps should have the montor bit set for all channels. In some sense that's the whole point of having separate OBSERVE files to be used in this way, that we use them to define setpoints against which every channel should be moniotred.