def reindex(self, entities):
        # Note - for some period of time, this doubles the memory usage of your ES cluster
        # If this is a problem, the ES reindex function can take source/destination cluster so you can
        # throw away the entire old cluster after everything is migrated.
        # This assumes you use an alias (shown as `self.index` in this example) for your API calls

        # 1. Create a new index with a unique name
        #    Set all of the index settings on the new index
        #    Set all of the doctype mappings on it
        # 2. Use the Elasticsearch reindex helper to copy all documents over to new index
        # 3. Swap the alias so API calls will use the new index. Drop the old index.
        # 4. Reload all documents onto the new index to get any newly mapped fields that may need to be indexed

        old_index = list(self.es.indices.get_alias(self.index).keys())[0]

        new_index = self._generate_actual_index()  # Unique name of the underlying index
        self.create_index(new_index)  # create the index with any index and analyzer settings

        # initialize the DocType mappings on the new index
        for entity in entities:
            # All of the entities in ES implement `doc_type()`
            doc_type = entity.doc_type()
            doc_type.init(new_index, using=self.es)

        helpers.reindex(self.es, old_index, new_index)
        self.es.indices.put_alias(new_index, self.index)
        self.es.indices.update_aliases({
            "actions": [
                {"remove": {"index": old_index, "alias": self.index}},
                {"add": {"index": new_index, "alias": self.index}}
            ]
        })
        self.es.indices.delete(old_index)

        for entity in entities:
            # Some function to pull batches out of the database (limit/offset)
            # Then use the `streaming_bulk` ES helper to push them
            pass