The pipeline runs much faster when a short time window is used when considering samples to try adding back into the tree, as it doesn't waste effort retrying samples which failed the previous n days. However this causes problems when only the sample_date is considered for inclusion in the tree as samples may not be uploaded before the sample_date is outside this time window
The pipeline runs much faster when a short time window is used when considering samples to try adding back into the tree, as it doesn't waste effort retrying samples which failed the previous n days. However this causes problems when only the sample_date is considered for inclusion in the tree as samples may not be uploaded before the sample_date is outside this time window