I have a few questions about Ferret’s performance when loading and
querying indexes. How do load time and query time scale with the
number of indexes being loaded or searched at once? Does the size of
these indexes matter?
From what I have seen here, the time it takes to load the indexes
seems to scale more or less linerarly (i.e., O(n)) with the number of
indexes to load, but not necessarily with their size. Is that correct?
What about query times?
On a semi-related note, is there a way to figure out how many
documents are there in an index without running a query on it?