Add node selection strategy documentation

Author: steveburnett

presto-docs/src/main/sphinx/connector/iceberg.rst

@@ -541,7 +541,8 @@ Property Name                                         Description
                                                       assign a split to. Splits which read data from the same file within
                                                       the same chunk will hash to the same node. A smaller chunk size will
                                                       result in a higher probability splits being distributed evenly across
-                                                      the cluster, but reduce locality.
+                                                      the cluster, but reduce locality. 
+                                                      See :ref:`develop/connectors:Node Selection Strategy`.
 ``iceberg.parquet_dereference_pushdown_enabled``      Overrides the behavior of the connector property                        Yes                 No
                                                       ``iceberg.enable-parquet-dereference-pushdown`` in the current session.
 ===================================================== ======================================================================= =================== =============================================

presto-docs/src/main/sphinx/develop/connectors.rst

@@ -8,7 +8,8 @@ you adapt your data source to the API expected by Presto, you can write
 queries against this data.
 
 ConnectorSplit
-----------------
+--------------
+
 Instances of your connector splits.
 
 The ``getNodeSelectionStrategy`` method indicates the node affinity
@@ -81,3 +82,14 @@ Given a split and a list of columns, the record set provider is
 responsible for delivering data to the Presto execution engine.
 It creates a ``RecordSet``, which in turn creates a ``RecordCursor``
 that is used by Presto to read the column values for each row.
+
+Node Selection Strategy
+-----------------------
+
+The node selection strategy is specified by a connector on each split.  The possible values are:
+
+* HARD_AFFINITY - The Presto runtime must schedule this split on the nodes specified on ``ConnectorSplit#getPreferredNodes``.
+* SOFT_AFFINITY - The Presto runtime should prefer ``ConnectorSplit#getPreferredNodes`` nodes, but doesn't have to. Use this value primarily for caching.
+* NO_PREFERENCE - No preference. 
+
+Use the ``node_selection_strategy`` session property in Hive and Iceberg to override this.