improve timeout handling

[alee]

• add a setInterval when issuing a query that informs the user that things are taking too long, allow them to reset / extend the timeout until it reaches a maximum
• clear timeout error messages when the query succeeds

[alee]

handling timeouts progressively requires that we navigate multiple layers of the technology stack:

1. haproxy timeouts as they broker the connection between the client browser and the backend API service
2. FASTAPI uvicorn timeouts: https://www.uvicorn.org/server-behavior/#timeouts
3. client side timeouts that can be extended up to a limit (defined by the server side)

To this end, what is the upper bound for max time we can handle?

[alee]

This is turning out to be a bigger issue to resolve than simply increasing timeouts. We could have a very long server side timeout and a lower client side timeout that can be progressively increased but that can choke the server / API with long running requests.

A more robust solution would require additional funding for development / engineering and would entail designing the application server and skope-api to use an asynchronous task queue like Celery

https://celery-node.js.org/