Amazon DynamoDB is a NoSQL database used in serverless applications. It supports key-value and document data models. If you need to create a DynamoDB table that will be used by your application service, here are the basic steps to do that.
1. The bare-minimum requirements to create a DynamoDB table are the table name, a partition key and a sort key column. For example, the table below:
Tablename: dbaworks-courses
PartitionKey: teacher_id
SortKey: course_id
The name of the table is dbaworks-courses. Table names are case-sensitive, and contains only characters a-z or A-Z, the numbers 0-9, or the symbols dash (-), underscore (_) or dot (.)
The partition key and sort key are the types of primary keys for DynamoDB tables. The partition key values are used in an internal hash function that determines the physical storage partition where the data item is stored. In the example, the partition key is teacher_id. The data items in the table dbaworks-courses can be accessed directly using the teacher-id value.
The use of sort key together with the partition key makes up the composite primary key. In this case, more than one data items can have the same value for the partition key teacher-id and are stored together in their respective partitions. In each partition, the data are further sorted based on the sort key, in this case the course_id.
2. You may optionally need to add tags to your new table for better categorisation and reporting.
Examples of tags and their values:
creator = dbaworks
environment = prod
department = computer-science
jiraticket = CS-2233
requestor = cmary
application = courseapps
3. Create the DynamoDB table in the AWS Console or using the AWS CLI command below:
aws dynamodb create-table \
--table-name dbaworks-courses \
--attribute-definitions \
AttributeName=teacher_id,AttributeType=S \
AttributeName=course_id,AttributeType=S \
--key-schema AttributeName=teacher_id,KeyType=HASH \
AttributeName=course_id,KeyType=RANGE
You need to supply the required parameters such as the table-name,the attribute-definitions, and the key-schema. Our table name is dbaworks-courses. The attribute-definitions parameter section is divided further into the attribute name and type. This is the part where you define all the columns of your table and their data types. In our example, these are 2 columns or attributes, and they are the teacher_id and the course_id.
The key-schema is where you define the primary or composite key for your table. Again, you can declare specific attributes like key name and key type under this parameter. In our case, the teacher_id is the the partition key, and it has the HASH value for its key type to indicate that this is a hash attribute and will be used in the internal hash function to define the storage partition. The other key attribute is course_id and it has the RANGE value for its key type to indicate that this is the sort key.
The output of the create-table command is below:
{
"TableDescription": {
"AttributeDefinitions": [
{
"AttributeName": "course_id",
"AttributeType": "S"
},
{
"AttributeName": "teacher_id",
"AttributeType": "S"
}
],
"TableName": "dbaworks-courses",
"KeySchema": [
{
"AttributeName": "teacher_id",
"KeyType": "HASH"
},
{
"AttributeName": "course_id",
"KeyType": "RANGE"
}
],
"TableStatus": "CREATING",
"CreationDateTime": 1595058082.896,
"ProvisionedThroughput": {
"NumberOfDecreasesToday": 0,
"ReadCapacityUnits": 0,
"WriteCapacityUnits": 0
},
"TableSizeBytes": 0,
"ItemCount": 0,
"TableArn": "arn:aws:dynamodb:us-west-1:123456789123:table/dbaworks-courses",
"TableId": "123456f3-1234-5f33-65cc-4cfd456e3nn7",
"BillingModeSummary": {
"BillingMode": "PAY_PER_REQUEST"
}
}
}
Take note of the resource ARN or the TableArn in the output of the create-table command, because you will need this for further modifications or updates in your table. In this example, the resource ARN or TableArn is "arn:aws:dynamodb:us-west-1:123456789123:table/dbaworks-courses"
4. You can monitor the table creation using the command below:
aws dynamodb describe-table --table-name dbaworks-courses | grep TableStatus
"TableStatus": "ACTIVE",
5. Once the table is created or in ACTIVE status, you can optionally add the tags to the table, using the tags in Step 2 as values. Here is how to add tags to the new DynamoDB table. The resource-arn is the TableArn value that resulted from the create-table command from Step 3.
aws dynamodb tag-resource \
--resource-arn arn:aws:dynamodb:us-west-1:123456789123:table/dbaworks-courses \
--tags \
Key=creator,Value=dbaworks \
Key=environment,Value=prod \
Key=department,Value=computer-science \
Key=jiraticket,Value=CS-2233 \
Key=requestor,Value=cmary \
Key=application,Value=courseapps
You can list the tags added for this table by running the command below:
aws dynamodb list-tags-of-resource \
--resource-arn arn:aws:dynamodb:us-west-1:123456789123:table/dbaworks-courses
6. You can also add secondary indexes to the table. These indexes can either be a global secondary index (GSI) or a local secondary index (LSI). Secondary indexes are used to query data using different keys other than the primary keys. GSI is an index with its own partition key and sort key different from the table's primary key, and LSI uses the partition key of the table but has a different sort key. These secondary indexes give more flexibility in querying the data. Secondary index declarations are also similar in form as creating the table and its attributes, and it requires such parameter values as IndexName, KeySchema, and Projection.
In the example below, we used the update-table command to add a global secondary index to the table:
aws dynamodb update-table \
--table-name dbaworks-courses \
--attribute-definitions \
AttributeName=student_uid,AttributeType=S \
AttributeName=created,AttributeType=N \
--global-secondary-index-updates \
'[{"Create":{"IndexName": "student-index","KeySchema":[{"AttributeName":"student_uid","KeyType":"HASH"},{"AttributeName":"created","KeyType":"RANGE"}],"Projection":{"ProjectionType":"ALL"}}}]'
The Projection declaration specifies that these column attributes will be copied or projected from the main table into the index. The projection type ALL means all the attributes will be projected to the index. You can also chose to project the index and primary keys only (KEYS_ONLY) or a specified list of table attributes only (INCLUDE).
7. You can learn more of the comprehensive options and functionalities of the create-table command by using the help option such as below:
aws dynamodb create-table help
