[iota-data-ingestion-core]: Handle graceful shutdown

[sergiupopescu199]

Description

The current implementation of IndexerExecutor attempts to provide graceful shutdown functionality, but it falls short in its execution. While the exit_receiver triggers loop termination, there's missed potential for a more controlled shutdown.

Unused Graceful Shutdown Mechanisms:

• CheckpointReader offers a dedicated channel for graceful shutdown, but it remains unutilized.
• Each worker in the WorkerPool possesses a term_receiver intended for graceful shutdown, but its functionality is not yet implemented. Dropping the workers vector doesn't initiate graceful shutdown. The checkpoint synchronization process is particularly sensitive. If a task is interrupted during checkpoint processing, it might not be able to update the progress store, leading to potential issues upon restart. This could result in duplicate checkpoint processing or conflicts in the data, depending on the specific implementation of the worker.