On Tue, Jan 07, 2020 at 02:39:01AM +0100, Fernando Fernandez Mancera wrote: > Hi all, > > I have noted that something is wrong with the codecov reports in the > pull request. I have opened a PR [1] which just change a few lines of > code in the same file and the codecov is reporting a total coverage > decrease of 26.21%. > > I think this is wrong. I will investigate more about it. Any > suggestions? Haven't look into detail yet. In travis, multiple test run will upload coverage data, if any of them failed or not uploaded within their(codecov) allowed range, codecov will think the partial data as complete coverage results, hence generate incorrect data. To fix this: * codecov: centos8-nmstate-dev --test-type integ * coveralls: nmstate/fedora-nmstate-dev --test-type unit_py36 Till mentioned he will look into it, please coordinates with him.