I ran your test with the alternative harness restraint and it passed. https://beaker.engineering.redhat.com/jobs/903203 All I did was add the following to the job xml.. In ks_meta I added harness='restraint-rhts beakerlib' and I added the following repo to include the harness repo: <repos> <repo name="restraint" url="http://file.bos.redhat.com/~bpeck/restraint/el7/"/> </repos> Hope that helps.
I'm guessing your tests are fiddling with the system time? This is an issue we have seen before (bug 849587, bug 639255) where Twisted generally does not behave if the system time is changed. At this point, using restraint as Bill suggests in comment 1 is probably the best choice. The likelihood of this being fixed in Twisted and/or Beah is low. We are planning to make restraint available in the default Beaker harness repos soon.
Actually bug 1063647 is the still-open bug where some unknown piece of Beah code passes a bad timeout value, possibly because the system clock has gone backwards. *** This bug has been marked as a duplicate of bug 1063647 ***