Seems odd to publish a benchmark against Lucene where your results are "Lucene crashes at 4 concurrent users". Since there are thousands of people using Lucene and products built on Lucene every day, I think the issue may be in your benchmarking code.