Same ShibSP session across two vhosts?

[Jon Gorrono]

On Sun, Oct 23, 2011 at 11:14 AM, Cantor, Scott <cantor.2 at osu.edu> wrote:
> On Oct 22, 2011, at 9:56 PM, "Jon Gorrono" <jpgorrono at ucdavis.edu> wrote:
>
>> Yeah, the apache mod is written in c++... but I said 'something like
>> terracotta'...are you saying that that the SP process/agent can't be
>> clustered at all?
>
> No, but Terracotta replicates objects and state in a way that's about as foreign to C++ as you can get, so no, nothing like it is going to work.

That is a narrow definition of 'like' ... I wouldn't expect that a
clustering solution that works well for C++ would be engineered quite
differently than terracotta.... by 'like' I was referring to OSS,
fast, binary, works....

>
> The formal clustering support is via coding to an explicit persistence API. Most of the time applications that can be clustered at all have their own session subsystem anyway and the SP doesn't need to be clustered directly.


This is a good point... born out by the discussion in NativeSPClustering

Thanks


>
> -- Scott
> --
> To unsubscribe from this list send an email to users-unsubscribe at shibboleth.net
>



-- 
Jon Gorrono
PGP Key: 0x5434509D -
http{pgp.mit.edu:11371/pks/lookup?search=0x5434509D&op=index}
GSWoT Introducer - {GSWoT:US75 5434509D Jon P. Gorrono <jpgorrono - gswot.org>}
http{middleware.ucdavis.edu}