Nirmal Fernando <[email protected]> writes:

> Hi Rick,
>
> On Tue, Jun 5, 2012 at 12:52 AM, Rick Hillegas
> <[email protected]> wrote:
>
>     If you are planning to tackle tasks on the 10.9.1 checklist, could
>     you take a moment to mark those tasks with your name:
>     http://wiki.apache.org/db-derby/TenNineOneChecklist If we can
>     develop a sense of where our gaps are, then we may be able to fill
>     those gaps.
>     
>
> I would love to help testing
> "org.apache.derbyTesting.perf.clients.Runner", but I've not looked
> into that Class yet. Is it a good pick for a novice (of this area)?
>
> Is there any guide line to follow?

I don't think there is one. Previously, the testers have typically just
picked a few combinations that they ran in their environment with both
the latest released version and the current release candidate, and
compared the results from the two versions to verify that the
performance hasn't gone down.

I'm going to run some of these tests on Solaris, but it would be useful
to see results from other platforms too. And two testers will probably
configure the JVM/db/test/OS differently, which will increase the
likelihood of discovering regressions that only manifest themselves in
certain configurations.

-- 
Knut Anders

Reply via email to