On Tue, Sep 15, 2009 at 3:03 PM, Matej Knopp <matej.kn...@gmail.com> wrote:
> Well,
>
> there are two kinds of overhead with SecondLevelCacheSessionStore.
> Object serialization and writing to disk. I can imagine antivirus
> interfering with the writing.
>
>
I did mention in comment #2 under the blog post the point about Windows /
AntiVirus possibly having something to do with it. IMO the difference in
performance is negligible and yes, may not even exist on *nix (needs
confirmation) - and as Matej said, the benefits are reduced page expiration
situations etc.
When posting the blog entry, my intention was not to recommend using the
HttpSession store. Just that the performance test does not simulate the
back-button at all and Tapestry does not have anything similar to the
DiskStore etc.
I do want to report a couple of settings that seemed to make a difference:
getMarkupSettings().setCompressWhitespace(true);
This almost cut page sizes in half for large HTML pages (having lots of
repeater indenting etc), and improved performance.
Also:
@Override
protected IRequestCycleProcessor newRequestCycleProcessor() {
return new UrlCompressingWebRequestProcessor();
}
This setting can help reduce the size of HTML in cases where you have a
large-sized repeater containing hyperlinks.
- Peter.
P.S: DZone link ;)
http://www.dzone.com/links/rss/wicket_beats_grails_jsfseam_and_tapestry_in_shoot.html
> -Matej
>
> On Tue, Sep 15, 2009 at 11:25 AM, Johan Compagner <jcompag...@gmail.com>
> wrote:
> > And that back and forward buttons in the browser dont really work well in
> > the HttpSessionStore
> >
> > But i cant believe that the overhead he is reporting is really the
> > overhead..
> >
> > It could be an Antivirus checker or other stuff like that.
> > Besides that if that would be a problem just get an SSD and place that
> into
> > your server where you point the DiskStore to
> > Then you will really not see any difference
> >
> > johan
> >
> >
> > On Tue, Sep 15, 2009 at 11:07, Matej Knopp <matej.kn...@gmail.com>
> wrote:
> >
> >> There really is no good reason to use http session store instead of
> >> secondlevelcachesessionstore unless you are experiencing performance
> >> problems. HttpSessionStore uses more memory and keeps much shorter
> >> history so your users will be experiencing page expiration much more
> >> often.
> >>
> >> -Matej
> >>
> >> On Tue, Sep 15, 2009 at 10:54 AM, Eyal Golan <egola...@gmail.com>
> wrote:
> >> > Hi,
> >> > I read this post
> >> >
> >>
> http://ptrthomas.wordpress.com/2009/09/14/perfbench-update-tapestry-5-and-grails/
> >> > and he suggested to use HttpSessionStore instead of the second
> cache...
> >> >
> >> > I looked into the javadoc and the code itself and understood the
> >> > differences.
> >> >
> >> > However, can anyone explain in a bit more detail the implications /
> >> > differences of using the Http instead of the SecondLayer?
> >> > The reason I'm asking this, is because, if it's better use, then I
> want
> >> to
> >> > convince my team that we change the implementation of our application
> to
> >> use
> >> > Http store.
> >> >
> >> >
> >> > Thanks very much.
> >> >
> >> > Eyal Golan
> >> > egola...@gmail.com
> >> >
> >> > Visit: http://jvdrums.sourceforge.net/
> >> > LinkedIn: http://www.linkedin.com/in/egolan74
> >> >
> >> > P Save a tree. Please don't print this e-mail unless it's really
> >> necessary
> >> >
> >>
> >> ---------------------------------------------------------------------
> >> To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
> >> For additional commands, e-mail: users-h...@wicket.apache.org
> >>
> >>
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
> For additional commands, e-mail: users-h...@wicket.apache.org
>
>
