Add LONG_TEST flags to backported fix tests using external data

[terriko]

We've had a couple of times where the backported fix code failed due to a network error. That's good information for us in general, but maybe not the most useful thing to have happen in CI. I think we should add LONG_TEST flags to tests that use network data other than NVD in order to improve the reliability of our short tests.

These are the ones that failed earlier, so it's probably those tests that need a LONG_TEST marker:

ERROR test/test_available_fix.py::TestAvailableFixReport::test_debian_available_fix_output
ERROR test/test_available_fix.py::TestAvailableFixReport::test_debian_backport_fix_output
ERROR test/test_available_fix.py::TestAvailableFixReport::test_redhat_available_fix_output

[XDRAGON2002]

Opened PR #1512 that closes this issue.

[BreadGenie]

we might not need to move these tests to LONG_TESTS now that those data are mocked in the test
refer #1513

[XDRAGON2002]

Hm, I looked at #1513 and I suppose the mocks really do make this issue redundant.
So, I guess I'll close #1512 as soon as your PR gets merged.