Added information about unsupported subgraphs for set affinity (#2872)

docs/IE_DG/supported_plugins/HETERO.md

@@ -32,6 +32,8 @@ Another way to annotate a network is to set affinity manually using <code>ngraph
 
 The fallback policy does not work if even one layer has an initialized affinity. The sequence should be calling of automating affinity settings and then fix manually.
 
+> **NOTE**: If you set affinity manually, be careful at the current moment Inference Engine plugins don't support constant (`Constant`->`Result`) and empty (`Parameter`->`Result`) networks. Please avoid such subgraphs when you set affinity manually.
+
 @snippet openvino/docs/snippets/HETERO1.cpp part1
 
 If you rely on the default affinity distribution, you can avoid calling <code>InferenceEngine::Core::QueryNetwork</code> and just call <code>InferenceEngine::Core::LoadNetwork</code> instead:

docs/nGraph_DG/nGraph_Python_API.md

@@ -188,6 +188,8 @@ for node in function.get_ordered_ops():
 
 Attributes are properties of nodes in the computational graph, which will be stored when the model is serialized to a file. However, there can be additional properties of nodes, which are only important during the execution of a graph. An example of such a property is `affinity`, which determines which operation will be executed on which hardware in a heterogeneous environment. You can get and set runtime information by using the `get_rt_info` method of a node.
 
+> **NOTE**: If you set affinity manually, be careful at the current moment Inference Engine plugins don't support constant (`Constant`->`Result`) and empty (`Parameter`->`Result`) networks. Please avoid such subgraphs when you set affinity manually.
+
 ```python
 rt_info = node.get_rt_info()
 rt_info["affinity"] = "test_affinity"