Hi is this a problem with perm memory? if so I think it is the same problem Matt Railble describes on his blog: http://raibledesigns.com/page/rd/20060419
the fix is to collect the memory consumtion data and express a solution on that problem by setting the perm mem size accordingly -XX:MaxPermSize=256m to MAVEN_OPTS. Matt says it to add to JAVA_OPTS but that doesn't really work for maven. On 4/20/06, Burkhard Graves <[EMAIL PROTECTED]> wrote: > > Hi, just a small remark: > > I experienced similar out-of-memory problems with my > spring/hibernate-app after around 10 or 12 reloads - but under tomcat > 5.5.12! I'm switching to jetty right now - as soon as everything is > running (actually some other problems hinder me ;-) I'll tell you if I > still have out-of-memory problems... > > Regards > Burkhard > > Jan Bartel schrieb: > > Hi dub, > > > > The jetty maven plugin is up to release beta14 > > so I would give that a go and see if it helps > > with your memory issue. > > We don't currently have any reported issues with out-of-memory problems > > for the plugin. The webapp > > classloader is ditched and then re-created on > > each restart so stuff loaded from the webapp's > > dependencies and classes should not be leaking. > > > > Are the spring jars explicitly on the plugin's classpath or are they as > > dependencies of the > > project? > > > > regards > > Jan > > > > > > gdub wrote: > >> I use the jetty6:run (6.0 beta 9) target > >> to launch Jetty with my web app under > >> integration. > >> > >> It has a not-too-extensive Spring/Hibernate > >> configuration. Jetty detects code changes > >> just fine but after maybe 10 reloads, it > >> starts reporting out-of-memory problems > >> and refuses to reload. The machine isn't > >> out of memory so it's the JVM itself that > >> hits a wall. > >> > >> Is this a known Jetty plug-in problem? Or > >> should I be looking for memory leaks in > >> Spring and Hibernate (or, e gads, my own > >> code). Is there something I need to > >> configure to make sure that Jetty releases > >> all app objects before reloading? > >> > >> BTW, I also ran into out of memory > >> problems under surefire when running > >> integration tests but was able to solve > >> it by using a singleton Spring application > >> context as a class member. But it leads > >> me to think that my Spring context isn't > >> releasing everything when it stops being > >> referenced (closing the context and > >> explicitly setting all references to it > >> null helped me get about 10 more tests > >> in a run). > >> > >> This is really only an issue during this > >> final integration phase so it's not too > >> big a deal but it does stop the thought > >> flow when it happens. It's also an appli- > >> cation confidence issue but I will do > >> some memory profiling later. > >> > >> > >> TIA, > >> > >> -dub > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > >
