Per-Entity Metadata Working Group

["Cantor, Scott" <>]

On 7/27/16, 4:46 PM, "Nick Roy" 
<>
 wrote:

> If we design the service to be highly available, I don't think we need to 
> worry about
> on-disk cache, which as Scott has mentioned is at the expense of the 
> ability of other
> non-Shibboleth clients to support the same assumed requirement.

That is an assumption, obviously. ADFS, for example, probably does cache this 
stuff, but that's partly because it doesn't really do on-demand metadata, so 
"supporting MDQ" is really more a scripting exercise than a software feature.

I don't know what Ping might do now or later, but I can imagine I guess that 
the more heavyweight products would store this stuff and be bolting on the 
on-demand option at most anyway.

So perhaps I'm wrong about it, I don't know. Probably wouldn't take much to 
get SSP to consider it.

> That said, if others want to run their own local copy, why don't they just 
> stand up local
> HTTP proxies?

Caching proxies, at least. Obviously they can, but I was reacting to the idea 
that we could somehow rely on that being common as a stopgap for the primary 
being reliable enough. It was never the intent that the Shibboleth 
implementation would demand that backstop.

-- Scott