The first- and last- buckets in date histograms are often incomplete. Rather than showing artificially-low values near the edges (which look like steep declines in whatever metric is aggregated), it'd be nice to optionally "trim" N buckets off the ends. The ES data source allows this in its Date Histogram with the "Trim Edges" option.
For my current use-case, I can get away with AND date(field) < date(NOW() - INTERVAL '1' day), but the option would be extra snazzy.
The first- and last- buckets in date histograms are often incomplete. Rather than showing artificially-low values near the edges (which look like steep declines in whatever metric is aggregated), it'd be nice to optionally "trim" N buckets off the ends. The ES data source allows this in its Date Histogram with the "Trim Edges" option.
For my current use-case, I can get away with
AND date(field) < date(NOW() - INTERVAL '1' day)
, but the option would be extra snazzy.