What's the correct way to handle state in a SignalR Hub?
I am going to implement a SignalR hub which will be connected to by a client application. The purpose of this being SignalR and not just HTTP requests is so that I can send messages from the server back to the client at any point without the need for something like port forwarding, just ad-hoc, and I will also use the connection to request general data too, so I don't need more than one connection to communicate data either direction.
I would like to authenticate the client with an API key when it connects, or allow it to connect unauthenticated so I can pop it up on my control panel and assign it a new API key and identity remotely (remote provisioning, essentially)
What would the correct way to do this be? I understand that sessions are not available as a result of "duplex messaging" which makes perfect sense, so my other idea was to do this with a scoped service to hold some state data for the connection, but as far as I can tell, the DI scope is not guaranteed especially if the hub is connected via HTTP polling instead of WS.
I would like to authenticate the client with an API key when it connects, or allow it to connect unauthenticated so I can pop it up on my control panel and assign it a new API key and identity remotely (remote provisioning, essentially)
What would the correct way to do this be? I understand that sessions are not available as a result of "duplex messaging" which makes perfect sense, so my other idea was to do this with a scoped service to hold some state data for the connection, but as far as I can tell, the DI scope is not guaranteed especially if the hub is connected via HTTP polling instead of WS.
