Backend availability check

[DmyMi]

Description

Implement a feature to monitor the availability of the backend service and handle cases when it is not responding.

Requirements

Backend availability check

• Make periodic requests to the backend to verify its status.
• If 3 consecutive requests do not receive a positive status response, consider the backend as unavailable.

Error page

• Display a special error page informing the user that something is wrong and to try again later.
• This page should replace the current content when the backend is unavailable.

"Try Again" button with exponential delay

• Include a "Try Again" button on the error page using the /healthz/active endpoint.
• If the backend is still not responding, pressing the button should trigger an exponential backoff delay before it can be pressed again.
  • The delay should start at 1 second and double each time up to a maximum of 1 minute.
  • The delay should persist through page refreshes for at least 10 minutes.

Additional notes

• The implementation should not interfere with other application functionalities.
• The delay persists across page refreshes for at least 10 minutes.
• Once the backend responds positively, the application returns to normal operation.
• Ensure that the backend check does not overload the server with too many requests.
• Handle edge cases where the application starts with the backend already unavailable.