Redis support for StorageService in opensaml-java?

[Cantor, Scott]

> Actually I am implementing SP, and the replay checking is mostly for assertion
> replays. Also with an LB on top of the hosts, per node cache doesn't really
> work.

A per node cache simply means that the cache is not global and is better than nothing, though not ideal. Virtually all Shibboleth SPs don't have a clustered replay cache, it's just not worth it.

> Also, what is your reason for not implementing SP? I understand there's a SP
> package implemented by shib, but is there any other reason for that?

Most people who implement SAML themselves do a very bad job of it at best, and implement it incorrectly at worst. Using an existing implementation is almost always a better choice.

If you're doing something internal and very limited in scope, it might be defensible.

-- Scott