The current code assumes every openqa test will have an entry
in TESTSUITES and breaks if it doesn't. All tests *should* have
entries, but it's susceptible to going wrong (as the tests
are in a different git repo so we have to sync the commits
carefully) and it is at least possible we might add tests to
openQA that don't match anything in the matrices, at least right
away. So instead of blowing up the report process with a
KeyError if the testsuite isn't there, just log a warning and
continue.
NOTE: this actually goes with D679: I want to add the N-1 upgrade
tests now, but they aren't in the matrices yet and won't be
until at least after we discuss it at Monday's meeting. So it
seemed smarter to add this check than to add an empty entry
to TESTSUITES, especially since I've seen this blow up before.