Using Amazon DynamoDb with the AWS.NET API Part 3: table operations

Introduction

In the previous post we started coding our DynamoDb demo application. In particular we saw how to extract information about tables in DynamoDb.

In this post we’ll look at various table operations, more specifically how to create, delete and update tables in DynamoDb.

Open the demo application we started working on previously and let’s get to it!

Creating tables

We can set the properties of our table through the CreateTableRequest object. This is where we set the table name, the provisioned read and write throughput, the key schema and their attribute definitions, and the secondary indexes. The following code sets all of these except the secondary keys.

Insert the following method into DynamoDbDemoService.cs:

public void CreateNewTableDemo(string tableName)
{
	try
	{
		using (IAmazonDynamoDB client = GetDynamoDbClient())
		{				
			CreateTableRequest createTableRequest = new CreateTableRequest();
			createTableRequest.TableName = tableName;
			createTableRequest.ProvisionedThroughput = new ProvisionedThroughput() { ReadCapacityUnits = 1, WriteCapacityUnits = 1 };
			createTableRequest.KeySchema = new List<KeySchemaElement>()
			{
				new KeySchemaElement()
				{
					AttributeName = "Name"
					, KeyType = KeyType.HASH
				},
				new KeySchemaElement()
				{
					AttributeName = "Birthdate"
					, KeyType = KeyType.RANGE
				}
			};
			createTableRequest.AttributeDefinitions = new List<AttributeDefinition>()
			{
				new AttributeDefinition(){AttributeName = "Name", AttributeType = ScalarAttributeType.S}
				, new AttributeDefinition(){AttributeName = "Birthdate", AttributeType = ScalarAttributeType.S}
			};
			CreateTableResponse createTableResponse = client.CreateTable(createTableRequest);

			TableDescription tableDescription = createTableResponse.TableDescription;
			Debug.WriteLine(string.Format("Table {0} creation command sent to Amazon. Current table status: {1}", tableName, tableDescription.TableStatus));

			String tableStatus = tableDescription.TableStatus.Value.ToLower();
			while (tableStatus != "active")
			{
				Debug.WriteLine(string.Format("Table {0} not yet active, waiting...", tableName));
				Thread.Sleep(2000);
				DescribeTableRequest describeTableRequest = new DescribeTableRequest(tableName);
				DescribeTableResponse describeTableResponse = client.DescribeTable(describeTableRequest);
				tableDescription = describeTableResponse.Table;
				tableStatus = tableDescription.TableStatus.Value.ToLower();
				Debug.WriteLine(string.Format("Latest status of table {0}: {1}", tableName, tableStatus));
			}

			Debug.WriteLine(string.Format("Table creation loop exited for table {0}, final status: {1}", tableName, tableStatus));
		}
	}
	catch (AmazonDynamoDBException exception)
	{
		Debug.WriteLine(string.Concat("Exception while creating new DynamoDb table: {0}", exception.Message));
		Debug.WriteLine(String.Concat("Error code: {0}, error type: {1}", exception.ErrorCode, exception.ErrorType));
	}
}

You’ll recognise the Hash and Range key types from the first post of this series. The KeySchemaElement object lets you define the name and type of the key. The AttributeDefinition will further let you define the schema of the keys.

We set up a compound key that consists of “Name” and “Birthdate”. We’ll insert some people-related objects in the next post where each record will have a compound key of Name and Birthdate, i.e. the name and birthday properties will uniquely identify each record. You’ll note the attribute type of S in the AttributeDefinition object. S stands for String. We saw earlier that DynamoDb supports 3 data types: strings, numbers and binary data. Numbers are represented by ScalarAttributeType.N and binary types by ScalarAttributeType.B. As dates are not supported by DynamoDb we’ll go with strings instead.

After constructing the CreateTableRequest object we call the CreateTable method of the DynamoDb client. The CreateTableResponse includes a TableDescription object which helps us read the initial status of the table. It should be “creating” at first. We then enter a loop where we continuously check the status of the new table until it reaches status “active”.

Let’s call the method from Main as follows:

DynamoDbDemoService service = new DynamoDbDemoService();
service.CreateNewTableDemo("a-demo-table");

If all goes well then you should see output similar to the following:

Table a-demo-table creation command sent to Amazon. Current table status: CREATING
Table a-demo-table not yet active, waiting…
Latest status of table a-demo-table: creating
Table a-demo-table not yet active, waiting…
Latest status of table a-demo-table: creating
Table a-demo-table not yet active, waiting…
Latest status of table a-demo-table: active
Table creation loop exited for table a-demo-table, final status: active

The table is up and running in the DynamoDB GUI as well:

It’s obviously not allowed to have two tables with the same name. Try running the same code again and you should get an exception message:

Exception while creating new DynamoDb table: {0}Table already exists: a-demo-table
Error code: {0}, error type: {1}ResourceInUseExceptionUnknown

Updating a table

You can update a table through the UpdateTableRequest object. You can update the read and write throughput values and the global secondary indexes, nothing else. The following method demonstrates the usage of the object:

public void UpdateTableDemo(string tableName)
{
	try
	{
		using (IAmazonDynamoDB client = GetDynamoDbClient())
		{
			UpdateTableRequest updateTableRequest = new UpdateTableRequest();
			updateTableRequest.TableName = tableName;
			updateTableRequest.ProvisionedThroughput = new ProvisionedThroughput() { ReadCapacityUnits = 2, WriteCapacityUnits = 2 };					
			UpdateTableResponse updateTableResponse = client.UpdateTable(updateTableRequest);
			TableDescription tableDescription = updateTableResponse.TableDescription;
			Debug.WriteLine(string.Format("Update table command sent to Amazon for table {0}, status after update: {1}", tableName
				, tableDescription.TableStatus));
		}			
	}
	catch (AmazonDynamoDBException exception)
	{
		Debug.WriteLine(string.Concat("Exception while updating DynamoDb table: {0}", exception.Message));
		Debug.WriteLine(String.Concat("Error code: {0}, error type: {1}", exception.ErrorCode, exception.ErrorType));
	}
}

Run the method like this:

DynamoDbDemoService service = new DynamoDbDemoService();
//service.UpdateTableDemo("a-demo-table");

Here’s the Debug output:

Update table command sent to Amazon for table a-demo-table, status after update: UPDATING

…and the table has been updated with the new throughput values:

Deleting a table

Deleting a table is as trivial as providing the name of the table to the DeleteTableRequest object and sending it to AWS by way of the DynamoDb client:

public void DeleteTableDemo(string tableName)
{
	try
	{
		using (IAmazonDynamoDB client = GetDynamoDbClient())
		{
			DeleteTableRequest deleteTableRequest = new DeleteTableRequest(tableName);
			DeleteTableResponse deleteTableResponse = client.DeleteTable(deleteTableRequest);
			TableDescription tableDescription = deleteTableResponse.TableDescription;
			TableStatus tableStatus = tableDescription.TableStatus;
			Debug.WriteLine(string.Format("Delete table command sent to Amazon for table {0}, status after deletion: {1}", tableName
				, tableDescription.TableStatus));
		}
	}
	catch (AmazonDynamoDBException exception)
	{
		Debug.WriteLine(string.Concat("Exception while deleting DynamoDb table: {0}", exception.Message));
		Debug.WriteLine(String.Concat("Error code: {0}, error type: {1}", exception.ErrorCode, exception.ErrorType));
	}
}

Run the method as follows:

DynamoDbDemoService service = new DynamoDbDemoService();
service.DeleteTableDemo("a-demo-table");

The table won’t be deleted immediately but have the status “deleting” at first before the table is finally deleted:

Delete table command sent to Amazon for table a-demo-table, status after deletion: DELETING

Here’s what it looks like in the DynamoDb GUI:

We’ll explore how to insert records into a DynamoDb table in the next post.

View all posts related to Amazon Web Services and Big Data here.