IdP Clustering

[Jarno Huuskonen]

Hi,

On Thu, Jun 11, NPTabunakawai wrote:
> I have a few questions about clustering in IdP 3. I'd be grateful for any
> feedback as it will help to decide on which approach to take in clustering
> or whether to stick to a single IdP.

Did you already read:
https://wiki.shibboleth.net/confluence/display/IDP30/Clustering

> - how many shib user sessions can an IdP instance handle?
> - regarding the dns round robin approach, what IdP metadata will be used by
> the SP? Will it use chaining or will metadata from each node be combined
> into one metadata file.
> - how will the certs and signing keys from each node be used ?

Why would you use different metadata/certs for each idp node ? Why not
just use same metadata/certs for each idp node ?

-Jarno

-- 
Jarno Huuskonen