Errors during observations (eg a rotator error) cause observations to get skipped. This is because observations are marked as completed once they are loaded. Add a new field that gets written to the schedules for attempted which increments a counter for the number of times it is attempted.
[ ] Simplest fix: the easiest approach here is to let things get attempted up to N times (2?)
[ ] Fancier fix: record the number of dithers actually executed so that on subsequent loads it will only do the remaining dithers.
I think we need both of these, because we need to still ensure that no observation gets stuck in an infinite loop where it is continually loaded and failed.
The preferred logic is:
load observation, increment attempted from 0 --> 1
increment completed dithers after each completed dither
only once all dithers are completed set completed from 0 --> 1
Errors during observations (eg a rotator error) cause observations to get skipped. This is because observations are marked as
completedonce they are loaded. Add a new field that gets written to the schedules forattemptedwhich increments a counter for the number of times it is attempted.I think we need both of these, because we need to still ensure that no observation gets stuck in an infinite loop where it is continually loaded and failed.
The preferred logic is: