I see, for example, https://pagure.io/atomic-images/issue/55 and https://pagure.io/atomic-images/issue/56, which are failures from Fedora-Rawhide-20160707.n.0 and Fedora-Rawhide-20160708.n.0, and it really appears to be the same ongoing failure. Creating a new issue adds noise, which I think makes it easier to ignore the whole thing. Instead, the first open bug with the same issue should be updated.

(If the earlier bug is closed, it should stay that way, I guess, but it'd be handy for the new one to note that this failure is a regression and link to the earlier bug.)

(This behavior was briefly described at https://fedoraproject.org/wiki/Changes/Two_Week_Atomic#Testing, BTW)