Menu
▾
▴
Re: [Resteasy-developers] Fixed? Re: Failed unit tests
|
From: Jozef H. <joz...@gm...> - 2010-01-29 08:51:37
|
Tests run: 350, Failures: 0, Errors: 0, Skipped: 0 Awesome! Well done. On 01/29/2010 02:55 AM, Bill Burke wrote: > Ok, I think I solved this issue. You'll have to get a clean checkout > though to see the changes. > > I invested TJWS a bit more and it seems that there is a race condition > in both startup and shutdown. I've forked TJWS into a resteasy project > to fix these problems. This seems to fix the problems. > > Let me know if you still get these issues. > > Solomon Duskis wrote: >> We could add in an auto increment or port round robin in the >> TestPortProvider class. It's a quick change, but would may pose a >> problem in some cases. A lot of the RESTEasy tests can take advantage >> of the InMemoryClientExecutor, but that would take a lot of work. >> >> -Solomon >> >> On Wed, Jan 27, 2010 at 9:20 AM, Bill Burke<bb...@re... >> <mailto:bb...@re...>> wrote: >> >> This happens. I'm not sure how to fix it. I'm guessing a test or >> two (or more) is not shutting down TJWS or...sockets aren't being >> cleaned up fast enough by the OS. >> >> >> Stef Epardaud wrote: >> >> On Wed, Jan 27, 2010 at 05:55:39AM -0500, Solomon Duskis wrote: >> >> It works fine for me. Could it be that you have a web >> server running when >> you ran all of your tests? What were the exceptions you >> were getting? >> >> >> It's hard to tell with Maven and their horrible logging policy. >> First >> there's no global way of finding out which tests failed. Second, I >> notice some errors show up in stderr and not in the surefire >> reports. I >> do get "address already in use" for the gzip interceptor for >> example, >> even though my 8080 and 8081 ports are free. Lingering sockets >> perhaps? >> >> >> -- >> Bill Burke >> JBoss, a division of Red Hat >> http://bill.burkecentral.com >> >> > -- Jozef Hartinger joz...@gm... |