Retry availability until the receiver of the request is dropped (#2763)

* guide updates

* keep interactions alive until receivers drop

* retry indefinitely

* cancel approval tasks on finality

* use swap_remove instead of remove

polkadot/node/network/availability-recovery/src/error.rs

@@ -16,7 +16,7 @@
 
 //! The `Error` and `Result` types used by the subsystem.
 
-use futures::channel::{mpsc, oneshot};
+use futures::channel::oneshot;
 use thiserror::Error;
 
 /// Error type used by the Availability Recovery subsystem.
@@ -34,9 +34,6 @@ pub enum Error {
 	#[error("failed to send response")]
 	CanceledResponseSender,
 
-	#[error("to_state channel is closed")]
-	ClosedToState(#[source] mpsc::SendError),
-
 	#[error(transparent)]
 	Runtime(#[from] polkadot_subsystem::errors::RuntimeApiError),