IWF-350: Sticky state execution: use previous worker instance

In SDK, worker service will optionally take in a url of the current instance. Then state options will allow setting a flag that the state execution will try to use the same worker instance as the previous state execution. When enabled, SDK will also send over the worker url for executing the state.

This feature should be used together with the failure recovery feature— it’s possible that the instance got killed and the processing may need to be started over.

No files or tests are named. Start by locating the SDK worker service and state-options entry points, then read the existing failure-recovery feature because this work is intended to be used with it. Done means state execution can optionally reuse the previous worker instance and pass its worker URL, while recovery handles an instance that was killed.

Written by the indexing model from the issue text.