On Mon, Mar 30, 2009 at 8:52 PM, Caldarale, Charles R < chuck.caldar...@unisys.com> wrote:
> > From: Gregor Schneider [mailto:rc4...@googlemail.com] > > Subject: Re: configuring tomcat with apache and eclipse. > > > > I'm understanding this scenario in such a way that there's a > > test/dev-environment that being a 1:1-copy of the > > production-environment. Makes sense, IMHO. > > Perhaps, but then why would the OP be concerned about losing session > information when updating with Eclipse, if this were only for > test/development? Just because its a pain. We have an encrypted login and everything else depends on it. Loosing session means needing to do the entire authentication with a remote service again. Anyway yes I want it to be a fairly close mirror of the live enviornment. I want to be able to control what should be secure and insecure just like in production. I wan't to be able to test my mappings so that I know that apache is serving the correct content and tomcat is serving the correct content. I'd like to be able to have a fairly good idea everything is working as it should. Too often in work (this is a private project) I've seen problems where apache won't understand filenames with ;jsessionid attached. Or where the entire functioning depends on a rewrite. Or even worse if the other developers end up writing code like request.getRequestUrl().indexOf("localhost") or where they end up hard coding port 8080 if localhost is in the url for absolue links. Suddenly I need to debug two seperate apps and I can't run either of them on another port because they only work on ports 80 or 443 or 8080. Its for all the above reason I want apache in right in at the beginning. If I can't then I'd pefer to serve all content from tomcat in production also. That opens a can of worms for scalibilty though. On Mon, Mar 30, 2009 at 7:43 PM, Gregor Schneider <rc4...@googlemail.com>wrote: > > > There's an interesting thread here on the list regarding the loss of > sessions during re-deployment which some consider a bug, and there's > even a workaround for that bug: > > http://www.nabble.com/Session-lost-when-app.-is-redeployed-td22660549.html Thanks I'll check it out. > > > My personal oppinion: > > When I re-deploy a webapp (and the webapp is changed!), a session > /must/ be lost, otherwise serialization / de-serialization of > session-data might end up somewhere in Nirvana. > I don't agree if nothing in the session has its serial number changed then it should remain. But thats just my opinion.
