On 3/31/07, Erik Hatcher <[EMAIL PROTECTED]> wrote:
Are the acts_as_solr folks here? Thoughts on this benchmark?
The test corpus was 414 documents... at that size I would think, one is testing mostly overhead. Probably network overhead is a sizeable portion for both servers (and why they benchmark so closely). The next largest amount of time is probably taken in query parsing, not query execution. Also, if I'm reading it correctly, there was only a single query client (no overlapping queries like one would get in real life). -Yonik
