Daniel
Daniel
Just need to get postgres working!
@cbaker6 @mtrezza would either of you know if this feature is compatible with Postgres? Or should the old auth lookup be used?
It should be a minor performance increase, but mostly intended as a functionality change as at the moment errors in auth config are only called once the auth is used....
Cache is required here otherwise each adapter is constructed each time an auth method is called. An auth adapter should be constructed once and then used across the rest of...
@Moumouls could you add your thoughts?
The cache is just to reduce the method from being called constantly when the same session token is used. If there are multiple servers, the session will just be continued...
> Ok, I understand that a shared cache in this case is a separate feature that doesn't exist currently, correct? Correct > Is the session expiration date also in the...
It's an arbitrary number that is currently used - it just represents the window in which using the same token _won't_ be extended (it will at the start of the...
No, I don't think so, the mechanism is a `throttle` so it should throttle it to one request every 500ms. Previously with the mechanism at a debouncer, this could occur....
The LRU cache here is a fixed, un-configurable time of 500ms. I think the 5s was just presented as a potential scenario, which wouldn't make a difference anyway as again,...