AWS Developer Blog

DynamoDB APIs

by Pavel Safronov | on | in .NET | Permalink | Comments |  Share

Amazon DynamoDB is a fast NoSQL database service offered by AWS. DynamoDB can be invoked from .NET applications by using the AWS SDK for .NET. The SDK provides three different models for communicating with DynamoDB. This blog post is the first of a series that describes the various APIs, their respective tradeoffs, best practices, and little-known features.

The Models

The SDK provides three ways of communicating with DynamoDB. Each one offers a different tradeoff between control and ease of use.

  • Low-level : Amazon.DynamoDBv2 namespace—This is a thin wrapper over the DynamoDB service calls. It matches all the service features. You can reference the service documentation to learn more about each individual operation.
  • Document Model : Amazon.DynamoDBv2.DocumentModel namespace—This is a model that provides a simpler interface for dealing with data. DynamoDB tables are represented by Table objects, while individual rows of data are represented by Document objects. Conversion of .NET objects to DynamoDB data is automatic for basic types.
  • Object Persistence Model : Amazon.DynamoDBv2.DataModel namespace—This set of APIs allow you to store and load .NET objects in DynamoDB. Objects must be marked up to configure the target table and the hash/range keys. DynamoDBContext acts on marked up objects. It is used to store and load DynamoDB data, or to retrieve .NET objects from a query or scan operation. Basic data types are automatically converted to DynamoDB data and converters allow arbitrary types to be stored in DynamoDB.

The three models provide different approaches to working with the service. While the low-level approach requires more client-side code—the user must convert .NET types such as numbers and dates to DynamoDB-supported strings—it provides access to all service features. By comparison, the Object Persistence Model approach makes it easier to use the service—since the user is for the most part working with familiar .NET objects—but does not provide all the functionality. For example, it is not possible to make conditional Put calls with the Object Persistence Model.

Sample code

The best way to gain an understanding of the different models is with a code sample. Below are three examples of storing and retrieving data from DynamoDB, each using a different model.


var client = new AmazonDynamoDBClient();

// Store item
client.PutItem(new PutItemRequest
    TableName = "Books",
    Item = new Dictionary<string, AttributeValue>
        { "Title", new AttributeValue { S = "Cryptonomicon" } },
        { "Id", new AttributeValue { N = "42" } },
        { "Authors", new AttributeValue {
            SS = new List<string> { "Neal Stephenson" } } },
        { "Price", new AttributeValue { N = "12.95" } }

// Get item
Dictionary<string, AttributeValue> book = client.GetItem(new GetItemRequest
    TableName = "Books",
    Key = new Dictionary<string, AttributeValue>
        { "Id", new AttributeValue { N = "42" } }

Console.WriteLine("Id = {0}", book["Id"].S);
Console.WriteLine("Title = {0}", book["Title"].S);
Console.WriteLine("Authors = {0}",
    string.Join(", ", book["Authors"].SS));

Document Model

var client = new AmazonDynamoDBClient();
Table booksTable = Table.LoadTable(client, "Books");

// Store item
Document book = new Document();
book["Title"] = "Cryptonomicon";
book["Id"] = 42;
book["Authors"] = new List<string> { "Neal Stephenson" };
book["Price"] = 12.95;

// Get item
book = booksTable.GetItem(42);
Console.WriteLine("Id = {0}", book["Id"]);
Console.WriteLine("Title = {0}", book["Title"]);
Console.WriteLine("Authors = {0}",
    string.Join(", ", book["Authors"].AsListOfString()));

Object Persistence Model

This example consists of two parts: first, we must define our Book type; second, we use it with DynamoDBContext.

class Book
    public int Id { get; set; }
    public string Title { get; set; }
    public List<string> Authors { get; set; }
    public double Price { get; set; }
var client = new AmazonDynamoDBClient();
DynamoDBContext context = new DynamoDBContext(client);

// Store item
Book book = new Book
    Title = "Cryptonomicon",
    Id = 42,
    Authors = new List<string> { "Neal Stephenson" },
    Price = 12.95

// Get item
book = context.Load<Book>(42);
Console.WriteLine("Id = {0}", book.Id);
Console.WriteLine("Title = {0}", book.Title);
Console.WriteLine("Authors = {0}", string.Join(", ", book.Authors));


As you can see, the three models differ considerably. The low-level approach is quite verbose, but it does expose service capabilities that are not present in the other models. The tradeoffs, specific features unique to each model, and how to use all models together will be the focus of this series on the .NET SDK and DynamoDB.