Re: 42.0.0.jre7, driver logging takes 30-40% of my server's time

Hi Vladimir,
here are the results from my test of PosgtresSQL JDBC 9.4.1212 vs 42.0.0
(as I mentioned on github).

It is a single run in a VM but it is ease to have several runs.

IMO the results don't show performance degradation at least for the
tested simple testcases - yes, there are some fluctuations but nothing
significant. The computer I'm running the VM is using very modest Intel
Celeron N3150. If someone has access to more powerful hardware it will
be good to run the tests also - the current setup is not so good in
detecting small changes in performance - too much time is spend in CPU
emulation in the VM.

I'm using the code base of TFB and particularly the Java/servlet
implementation. The code is NOT having a logging framework in the
classpath. At least I'm not aware of it and I've checked the maven
dependencies. The servlet container is Resin 4.0.48. If anyone wish to
take a look don't forget to provide the correct maven profile (-P v42 or
-P v9).

The logs of the run are here:
https://github.com/zloster/logs/tree/master/FrameworkBenchmarks/postgreJDBC-9.4.1212-vs-42.0.0
The numbers are here:
https://github.com/zloster/logs/blob/master/FrameworkBenchmarks/postgreJDBC-9.4.1212-vs-42.0.0/results.json
The servlet implementation is here:
https://github.com/zloster/FrameworkBenchmarks/tree/postgrejdbc/frameworks/Java/servlet
The branch is here:
https://github.com/zloster/FrameworkBenchmarks/tree/postgrejdbc
For other details consult the logs or ask me.

About https://github.com/pgjdbc/pgjdbc/issues/793 - it should be fairly
easy to have a servlet with Log4j logger and some servlet with some
other logger also. Even if the test tables are very small there are
quite a lot of resultset processing. If there is performance degradation
maybe it will show up as a difference.

Cheers,
zloster