Benchmarks - overhead of instrumented code without a tracer

[jpkrohling]

See opentracing/opentracing-java#295

Similar to the linked issue, this task is about measuring the overhead of traced JDBC calls when compared to their non-traced counter-parts.

Scenarios to test:

• [ ] Using an in-memory database, execute as many simple queries as possible (like select 1). Run it by time (X statements per minute) and by a fixed number of statements (task took N seconds to complete). Run twice, one with a traced JDBC driver and one without.

[jpkrohling]

@malafeev could you add me as a contributor to this repo? I can't seem to be able to add labels to issues in this repo. Alternatively, please create/add the labels outreachy and performance to this issue, like there is for opentracing/opentracing-java#295

[malafeev]

@jpkrohling done

[jpkrohling]

Thanks!