I don't know if you are interested by my apache cellar fork has all the changes necessary to update the tests to use the apache karaf 3.0.0 commands which are different. I also added a group manager property in a couple blueprint files to prevent NPEs during startup. Unfortunately, even though those problems s are fixed, cellars 3.0.0 deadlocks during startup, one of the blocked threads seemed to be waiting on the "Karaf User Console" which didn't seem to make sense. When it doesn't deadlock, it will error out after a few seconds D's with concurrent modification exceptions due to the combined class loader being constantly requested to load a RemoteRepositoryEvent by simultaneous threads. When I hooked in a debugger, there were alternating RemoteRepositoryEvents saying the cellar feature repository was remove and added. They just kept rapidly firing and the console would show a concurrent modification except stack trace every few minutes.
Unfortunately this looks .like it's going to be a pain in the but of a problem to fix.... have you seen that happen with Cellar 3.x yet? Ryan On Jan 16, 2013 10:54 AM, "Jean-Baptiste Onofré" <[email protected]> wrote: > Hi Ryan, > > yes, I started to work on both Cellar 2.3.x and trunk yesterday. > It should be better by the end of this week. > > Regards > JB > > On 01/16/2013 05:12 AM, Ryan Moquin wrote: > >> I know I was told that this version isn't stable yet, but I still wanted >> to see if the build should be able to complete successfully or not. I >> want to determine if some issues I'm having are something on my end or >> not. Currently the itests always seem to fail with cellar trunk, is >> that the current known state of it? Or is it failing on my side only? >> >> Thanks, >> Ryan >> >> > -- > Jean-Baptiste Onofré > [email protected] > http://blog.nanthrax.net > Talend - http://www.talend.com >
