DynamoDB supported features:
- Batched Operations - the current DynamoDB API only supports batched fetch operations. Additionally DynamoDB makes it painful to fetch more than a few items at a time. Pynamo provides a useful batched operation facade.
- Set Operations - DynamoDB has native support for sets of strings or numbers, which Pynamo supports with a much simplified API.
- UpdateItem - Pynamo automatically uses UpdateItem when it makes sense.
Pynamo builds on the great boto package, whose underlying connection api automatically uses a connection pool and supports keep-alive and timeouts.
Additionally, Pynamo comes with some features of it's own:
- Declaritive Schema - Pynamo's schema goes beyond defining just a hash key and range key. Create a safe and consistent way of representing your data in DynamoDB
- Compound Keys - keys that are composed of multiple keys.
- Synthesized Types - attributes can be container types like list or dict
- Automated, Sortable Unique ID Generation - Pynamo provides the LexicalUUIDField which may be used to generate hash and range keys.
- Validated Attributes - attributes are validated against the type chosen when writing persisted classes
- No Overhead The objects are backed by the same dictionaries that Boto uses so there is virtually no overhead to using Pynamo unlike with an most ORMs.
First, you will need an Amazon AWS account. If you don't have one of those, why are you running these tests?
Create a file: ~/.pynamo.cfg which provides the following data (filled in):
[aws] access_key_id = secret_access_key = [dynamodb] table_prefix =
I use nose which pretty much rocks. From the main repository directy simply run nosetests which will pick up all the Pynamo tests and run them.
Running the full test suite takes a very long time. It has to create and destroy many DynamoDB tables, and at the time of writing that seems to take quite a while.
If some tests fail, you may have to delete the tables from the AWS Console or another command line client by hand.