Summary
Once we have the ability to query SPICE using specific timestamps, not just days, we will need to update batch_starter to query for the entire MAG range (which is the full day +-30 minutes) to prevent failures on runs that do not have complete SPICE data.
Current issue, is there are processing jobs running against incomplete SPICE data and failing. This doesn't have a big impact, but it does make it hard to distinguish legitimate failures.
Requirements
The more specific the better.
Related tickets
Blocked by #1059