[platform-releng-dev] Re-running tests on CentOS Machine for I20160512-1

["David M Williams" <david_williams@xxxxxxxxxx>]

Earlier today, while several of us were
trying to figure out a "DNF" from last night's build, I noticed
that the heap size on both of our Linux test boxes was overly constrained
(probably for years). For details, see https://bugs.eclipse.org/bugs/show_bug.cgi?id=493485#c9

While the fix was not made in time for
today's respin and tests, there was a DNF on the CentOS machine today (in
the org.eclipse.e4.ui.tests.css.swt suite) and that made me look at the
logs closer, and that made me remember the CentOS machine needed the memory
fix made also (I was mistakenly thinking they shared a configuration file).

So out of curiosity more than anything,
I would like to re-run the tests on the CentOS machine. While one run won't
prove "no more DNFs" it should be interesting to begin to see
the effects of allowing Linux to have more memory when it runs. 

This will "replace" the JUnit
test results from I20160512-1000 that are on display now (for CentOS) ...
in about 6 or 8 hours. I hope they do not get worse. :)