If nameko rpc is called from standalone flask app using ClusterRpcProxy and if the message is lost.
does the nameko/rabbit retrys the message?
It’s not really possible for an RPC request message to be “lost”.
The ClusterRpcProxy will error if the message can’t be delivered to an RPC queue on the broker. The queue will exist as long as the target service is running or was previously running. Once the message is queued, it’ll be consumed by the next available service worker.
Does that answer your question?