cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Geoff Howard" <coc...@leverageweb.com>
Subject RE: [RT] improving the session concept
Date Thu, 10 Apr 2003 01:57:50 GMT
Ok, this is what I had assumed.  I thought there was some discussion about 
eliminating access to the session from the flow but sounds like I either 
misunderstood or that didn't happen.  I have yet to go through the new 
idl docs carefully.

[OT]How's gmane working out?

Thanks,
Geoff

> -----Original Message-----
> From: news [mailto:news@main.gmane.org]On Behalf Of Stefano Mazzocchi
> Sent: Wednesday, April 09, 2003 4:23 PM
> To: cocoon-dev@xml.apache.org
> Subject: Re: [RT] improving the session concept
> 
> 
> on 4/8/03 5:14 PM Geoff Howard wrote:
> 
> >>>Stefano Mazzocchi wrote:
> > 
> > 
> > <snip/>
> > 
> >>it will be nice if the flow could wrap around the session objects and
> >>provide a way to add hooks to the sessione events and this simply with
> >>using the servlet API.
> > 
> > 
> > I think this thread may be answering a question I was getting
> > ready to pose about continuations and load balancing.
> > 
> > In a scenario with multiple servlet containers sharing session
> > information, is it possible now to share continuation information
> > across those containers?  
> 
> currently is not possible (AFAIK).
> 
> > If not, what paths to make that possible exist?
> 
> maybe store the continuation into the session? I don't know, just
> thought about it but since the continuation is a real native object of
> the javascript engine, it should be fully serializable. this means it
> can be transported from a clustered machine as much as any other data.
> 
> of course, the concept of having a jsession-id-aware load balancer up
> front holds true for this as well, but if you need to replicate those
> sessions around your machines (for whatever reasons) the virtualization
> provided by having a javascript engine allows us to ship the
> continuation (which is, in short, the rest of the program to execute) to
> the other machine.
> 
> but don't ask me how big (in data size) a continuation is because I have
> no idea.
> 
> But I'm sure Chris will be able to answer the question with much more
> details.
> 
> -- 
> Stefano.
> 
> 
> 
> 

Mime
View raw message