Research if ingestion can be optimized by sorting and partitioning the ingested multivariate data by tag values

[CGodiksen]

This would solve the problem of having to create many small vectors with tag values and calculating the tag hash for each data point that is inserted. A discussion on this was started here: #293 (comment).