Programatically implement auto failover

We are trying to implement AutoFailover from Server A to Server B in case of any issues in servers or to implement Client side Load Balancing . The active cluster details is exposed through an API . In Java we have NameResolver , which updates the connection config and in Go we have Resolver for the same . I am looking for some suggestions to implement the same in Python as well since there is no interface exposed for Python.

I do not believe we have this currently. In Java and Go we use those language’s gRPC mechanism but not with core-based SDKs where we use Rust-based SDK. Let me confer with the team on the best solution here.

To confirm, are you looking for a target host resolution callback, or could you work with something similar like a DNS resolver or being able to provide multiple target hosts?

We are looking for target host resolution callback

We have opened [Feature Request] Dynamic client endpoints · Issue #477 · temporalio/sdk-core · GitHub to support this feature. When implemented, you will be able push changes to your “service set” dynamically. So it’s not a callback necessarily but just as effective for most use cases. I am afraid I cannot give a time table on when it will be implemented.