chore: log Ray Pool run batches

[hfutatzhanghb]

Summary

• Add info-level logs before and after each index Ray Pool run_batch business batch.
• Include batch size, a bounded fragment ID preview, and result status in the log messages.
• Add a unit test that exercises the shared index Pool helper and verifies the run-batch log messages.

Why

The index helper submits fragment_batches to ray.util.multiprocessing.Pool.map_async(..., chunksize=1). In Ray Pool, each submitted chunk is executed by PoolActor.run_batch. With chunksize=1, each run_batch contains one balanced fragment batch.

For large vector index builds, these batches can be memory-heavy or long-running. Bracketing each business batch with info-level logs makes it easier to identify which fragment batch is starting, which one completed, and what status it returned, without logging an unbounded list of fragment IDs.

Validation

• python -m py_compile lance_ray/index.py tests/test_vector_index_options.py
• python -m pytest tests/test_vector_index_options.py
• uv run --no-sync ruff check lance_ray/index.py tests/test_vector_index_options.py
• git diff --check