SLO session lookup failure

[Cantor, Scott]

> To recap, we get a repeatable SessionNotFound error during SLO when it
> happens after 30min of inactivity.

That happens to be the default Java sesssion timeout, so I would speculate that the problem is related to getting the client-side session data loaded back into the IdP. If the container session is still alive, the data's cached on the server in the session and only reloaded from the client when it has to be.

The relevant logging here is going to be client side, you need traces of what it's doing and I would be curious whether the IdP is in control of the UI frame. If it's not it's probably not going to work.

The other low level logging would be from the storage classes, down to TRACE I would imagine. org.opensaml.storage

-- Scott