The last improvement of memory usage introduced a performance
regression. The ingestion performance dropped by 50%-80%, because
for every inserted entry the Tags were created inefficient.
Before we could only group by a single field. But it is acutally
very useful to group by multiple fields. For example to see the
graph for a small set of methods grouped by host and project.
Store values in sequences of variable length. Instead of using 8 bytes
per entry we are now using between 2 and 20 bytes. But we are also able
to store every non-negative long value.
