Thanks. Eelco
On 2/12/07, Weaver, Scott <[EMAIL PROTECTED]> wrote:
Okay, it is up in Jira (WICKET-279) with the corrected patch format. Regards, -scott > -----Original Message----- > From: Upayavira [mailto:[EMAIL PROTECTED] > Sent: Monday, February 12, 2007 11:16 AM > To: wicket-dev@incubator.apache.org > Subject: Re: [PATCH] Wicket-spring Fix for classloading in a clustered > Weblogic 9.x > > Weaver, Scott wrote: > > Hi All, > > > > First off, it has been a loooooong time since I have had to submit a > > patch (with the Jetspeed project I had the luxury of just committing). > > Hopefully the format is correct (it was generated from Subclipse so I am > > guessing it should be). > > > > Okay, that out of the way on to the important part. This patch > > addresses an issue with classloading performed by > > wicket.proxy.LazyInitProxyFactory within the context of a Weblogic > > Server 9.2 clustered application environment (stack trace at the bottom > > of the email). The long and the short of it is that > > Thread.currentThread().getContextClassLoader() will fail to load the > > proxied interface. The fix was quite simple: catch the IAE and attempt > > to load the interface using the current classloader > > (LazyInitProxyFactory.class.getClassLoader()). This works like a charm > > and I have had in production for almost 3 months now with no issues > > whatsoever. > > You should upload this patch to Jira [1]. > > Secondly, it makes patching easier if you create your diff from within > the root of your checkout. That way it won't have D:\ etc in the patch > file. Just makes life easier for whoever applies the patch. > > Regards, Upayavira > > [1] https://issues.apache.org/jira/browse/WICKET