DynamoDB, ItemCollection#query makes additional get_item request. about aws-sdk-ruby HOT 2 CLOSED

No, it is not necessary. The issue is the SDK does not know what or how many attributes you require. In the abscence of the list of attributes to select, it will request only the table schema attributes. In your example, this is "id" and "date". The table schema attributes are required so the yielded Item object will know how to operate on itsel f (for example, what if you only wanted to delete each yielded item, you would need to know its schema attributes).

If you pass in a list of attributes to select, the block will instead yield an ItemData object (instead of an Item object).

table.items.query(:select => [ATTRIBUTES_TO_SELECT], :hash_value => "1", :range_value => "22") do |item_data|
    puts item_data.attributes #=> a hash of selected attributes
    item_data.item #=> the AWS::DynamoDB::Item object for this attribute bag
end

from aws-sdk-ruby.

Thanks for clarify. As reference in Doc: Item objects do not have any attribute data populated. The Attributes#to_hash is the one that makes get_item request.

from aws-sdk-ruby.