Not all of the data we might want to load into Treasure Data is time series data. When we bulk load a data set that does not have a time column, the 'time' values are all pretty much the same. It would be nice to be able to select another column, perhaps a categorical one, to do the partitioning on. It is not clear that there would be significant query or load performance improvements (there might be!), but it is a data architecture/elegance issue that really bugs me.2 votes
- Don't see your idea?