Per-Entity Metadata Working Group

["Cantor, Scott" <>]

> Scott C, do you envision for "caching"
> 
> a) writing to disk the metadata for an entity each time it has
> been fetched successfully to be used to accomplish a SAML flow

Yes. Currently we cache only in RAM. It may be possible to tweak the Java 
code to enable the HttpClient cache to persist, but our initial suspicion is 
that doing that would be risky and confusing and potentially not work anyway, 
because it's really like a browser, it doesn't necessarily just use the 
cached copy if the remote one is unavailable, at least not predictably.

> b) periodically downloading an aggregate that contains all of
> the entities (since we expect aggregates to stay around for
> some time) and writing it to the side on disk but not using it
> except when detecting that MDQ queries are failing (hence not
> having to take up memory when not necessary).

No, apart from discovery-motivated necessities, but not for basic use, no.

-- Scott