Agreed, this should be considered a bug. On Thu, May 22, 2008 at 1:15 PM, Gert Vanthienen <[EMAIL PROTECTED]> wrote: > Ivan, > > I can't imagine that being the intended behavior so feel free to raise a > JIRA for this > > Gert > > > > Ivan Stojic wrote: >> >> Hi, >> >> I'm testing the hot (re)deployment feature of the ServiceMix 4 Kernel. I >> created an OSGi bundle with a simple BundleActivator which only prints out >> when start/stop methods are called. When I copy the bundle to the deploy >> dir, the start message is printed and everything works fine. >> >> But, when I copy the bundle again (overwriting the old one) or when I >> simply >> 'touch' the old one in the deploy dir, I noticed that it is being stopped >> and started twice in a row. >> >> I traced the problem to the >> FileMonitor.refreshPackagesAndStartOrUpdateBundles() method, where my >> bundle >> is first being refreshed via packageAdmin.refreshPackages(bundles) >> (probably >> causing the first stop/start of the bundle) and, right after that, the >> same >> bundle is being updated via bundle.update() (causing the second >> stop/start). >> >> Is this the intended behavior or some kind of a bug? >> >> Thanks, >> Ivan >> > >
-- Cheers, Guillaume Nodet ------------------------ Blog: http://gnodet.blogspot.com/
