docs: clarify chunking caveats

Signed-off-by: memleakd <121398829+memleakd@users.noreply.github.com>

Author: memleakd

user_guide_src/source/models/model.rst

@@ -980,6 +980,8 @@ Sometimes, you need to process large amounts of data and would run the risk of r
 This is best used during cronjobs, data exports, or other large tasks. To make this simpler, you can
 process the data in smaller, manageable pieces using the methods below.
 
+.. note:: The ``chunk()``, ``chunkRows()``, ``chunkById()``, and ``chunkRowsById()`` methods do not trigger Model callbacks.
+
 chunk()
 -------
 
@@ -1020,7 +1022,8 @@ If you want the entire ID-based chunk to be passed to the Closure at once, use `
 
 .. warning:: ID-based chunking needs the model to define a primary key, and that primary key must be selected.
     If you customize the selected columns, make sure the primary key is included.
-    When joining tables, make sure the model table primary key remains available under its normal field name.
+    When using joins, the model primary key must remain available under its normal field name,
+    and each primary key value must appear only once in the result set. Otherwise, rows may be skipped.
     These methods manage their own primary key ordering and cannot be used with
     ``orderBy()``, ``groupBy()``, ``limit()``, ``offset()`` or ``union()``.
     You should also avoid changing primary key values while processing chunks.