Hi There has been some sporadic errors in camel-netty4-http after a commit change a while back. I suspect there is a race condition in some close channel code that can cause it to fail some tests.
Though if the re-run was applied it may hide this problem. The CI tests is pretty good as-is, if there is suddenly a "port number in use" issue then we often see that as 30+ jetty tests failed etc. The re-run should imho be more intelligent and detect it was a "port number in use" problem and re-run the test a bit later etc. On Thu, Jan 14, 2016 at 5:56 PM, Andrea Cosentino <ancosen1...@yahoo.com.invalid> wrote: > Hi Claus, > > If we mantain the re-run number on a default value of 2, we should be able to > notice potential issues. > > In my opinion the PR is ok. > > Thanks, > Andrea > > -- > Andrea Cosentino > ---------------------------------- > Apache Camel PMC Member > Apache Karaf Committer > Email: ancosen1...@yahoo.com > Twitter: @oscerd2 > Github: oscerd > > > > On Thursday, January 14, 2016 5:44 PM, Claus Ibsen <claus.ib...@gmail.com> > wrote: > Hi > > We got a PR pending at > https://github.com/veithen/camel/commit/6bcdb7f30f4984508fab6f683c3e4775bfc84a6e > > I haven't seen that option before on the surefire plugin. > > It seems at first glance as a nice way to work around any oddiness. > But wonder on the other hand if it would cause these "oddiness" to not > surface and we aren't noticing that there is a potential issue in a > Camel component. > > Any thoughts on this PR ? > > -- > Claus Ibsen > ----------------- > http://davsclaus.com @davsclaus > Camel in Action 2: https://www.manning.com/ibsen2 -- Claus Ibsen ----------------- http://davsclaus.com @davsclaus Camel in Action 2: https://www.manning.com/ibsen2
