Dynamodb
Chapter 23: Amazon DynamoDB - NoSQL Database
Section titled “Chapter 23: Amazon DynamoDB - NoSQL Database”Serverless NoSQL Database
Section titled “Serverless NoSQL Database”23.1 Overview
Section titled “23.1 Overview”Amazon DynamoDB is a fully managed, serverless, key-value NoSQL database designed for single-digit millisecond performance at any scale.
DynamoDB Overview+------------------------------------------------------------------+| || +------------------------+ || | DynamoDB | || +------------------------+ || | || +---------------------+---------------------+ || | | | || v v v || +----------+ +----------+ +----------+ || | Serverless| | Scalable | | Fast | || | | | | | | || | - No | | - Auto | | - <10ms | || | servers| | scaling| | latency| || | - Managed| | - Any | | - Multi-AZ| || | | | scale | | | || +----------+ +----------+ +----------+ || || Key Features: || - Fully managed, serverless || - Single-digit millisecond latency || - Automatic scaling || - Multi-AZ replication || - Global tables (multi-region) || |+------------------------------------------------------------------+23.2 DynamoDB Data Model
Section titled “23.2 DynamoDB Data Model”Core Concepts
Section titled “Core Concepts” DynamoDB Data Model+------------------------------------------------------------------+| || Table Structure || +----------------------------------------------------------+ || | | || | Table: Users | || | +----------------------------------------------------+ | || | | Primary Key | Attributes | | || | |--------------|--------------------------------------| | || | | PK | SK | Name | Email | Status | | || | |-------|------|---------|--------------|-----------| | || | | USER1 | null | John | john@ex.com | active | | || | | USER2 | null | Jane | jane@ex.com | active | | || | | USER3 | null | Bob | bob@ex.com | inactive | | || | +----------------------------------------------------+ | || | | || +----------------------------------------------------------+ || || Primary Key Types: || +----------------------------------------------------------+ || | | || | 1. Simple Primary Key (Partition Key only) | || | - Single attribute identifies item | || | - Example: UserId | || | | || | 2. Composite Primary Key (Partition + Sort Key) | || | - Two attributes identify item | || | - Partition Key: Hash | || | - Sort Key: Range | || | - Example: UserId + Timestamp | || | | || +----------------------------------------------------------+ || |+------------------------------------------------------------------+Items and Attributes
Section titled “Items and Attributes” DynamoDB Items and Attributes+------------------------------------------------------------------+| || Item (Row equivalent) || +----------------------------------------------------------+ || | | || | { | || | "UserId": "USER1", // Partition Key | || | "Timestamp": 1234567890, // Sort Key | || | "Name": "John Doe", | || | "Email": "john@example.com", | || | "Age": 30, | || | "IsActive": true, | || | "Tags": ["premium", "verified"], | || | "Address": { | || | "Street": "123 Main St", | || | "City": "New York", | || | "Zip": "10001" | || | } | || | } | || | | || +----------------------------------------------------------+ || || Attribute Types: || +----------------------------------------------------------+ || | Scalar Types: | || | - S (String) | || | - N (Number) | || | - B (Binary) | || | - BOOL (Boolean) | || | - NULL (Null) | || | | || | Document Types: | || | - M (Map) - JSON object | || | - L (List) - Array | || | | || | Set Types: | || | - SS (String Set) | || | - NS (Number Set) | || | - BS (Binary Set) | || | | || +----------------------------------------------------------+ || |+------------------------------------------------------------------+23.3 Capacity Modes
Section titled “23.3 Capacity Modes”On-Demand Mode
Section titled “On-Demand Mode” DynamoDB On-Demand Mode+------------------------------------------------------------------+| || Features: || +----------------------------------------------------------+ || | | || | - Pay per request | || | - No capacity planning | || | - Instant scaling | || | - No throttling (within service limits) | || | | || | Pricing: | || | - Read: $1.25 per million request units | || | - Write: $1.25 per million write units | || | - Storage: $0.25 per GB-month | || | | || | Use Cases: | || | - Unpredictable traffic | || | - New applications | || | - Development/testing | || | | || +----------------------------------------------------------+ || |+------------------------------------------------------------------+Provisioned Mode
Section titled “Provisioned Mode” DynamoDB Provisioned Mode+------------------------------------------------------------------+| || Read Capacity Units (RCU) || +----------------------------------------------------------+ || | | || | 1 RCU = 1 strongly consistent read per second | || | for items up to 4 KB | || | | || | 0.5 RCU = 1 eventually consistent read per second | || | | || | Example: | || | - Item size: 8 KB | || | - Strongly consistent: 2 RCU | || | - Eventually consistent: 1 RCU | || | | || +----------------------------------------------------------+ || || Write Capacity Units (WCU) || +----------------------------------------------------------+ || | | || | 1 WCU = 1 write per second | || | for items up to 1 KB | || | | || | Example: | || | - Item size: 5 KB | || | - Required: 5 WCU | || | | || +----------------------------------------------------------+ || || Auto Scaling || +----------------------------------------------------------+ || | | || | Configuration: | || | - Minimum capacity | || | - Maximum capacity | || | - Target utilization (e.g., 70%) | || | | || | Scaling: | || | - Automatic increase/decrease | || | - Based on CloudWatch metrics | || | | || +----------------------------------------------------------+ || |+------------------------------------------------------------------+23.4 DynamoDB Features
Section titled “23.4 DynamoDB Features”Global Secondary Indexes (GSI)
Section titled “Global Secondary Indexes (GSI)” Global Secondary Indexes+------------------------------------------------------------------+| || Purpose: Query on non-key attributes || || Architecture: || +----------------------------------------------------------+ || | | || | Table: Users | || | +----------------------------------------------------+ | || | | PK: UserId | SK: null | Email | Status | | || | |-------------|---------|--------------|-----------| | || | | USER1 | - | john@ex.com | active | | || | | USER2 | - | jane@ex.com | active | | || | +----------------------------------------------------+ | || | | | || | v | || | GSI: EmailIndex | || | +----------------------------------------------------+ | || | | PK: Email | SK: Status | UserId | | | || | |----------------|-------------|-----------|--------| | || | | john@ex.com | active | USER1 | | | || | | jane@ex.com | active | USER2 | | | || | +----------------------------------------------------+ | || | | || +----------------------------------------------------------+ || || Features: || - Different partition key than table || - Can have sort key || - Projected attributes (all, keys-only, include) || - Consumes separate RCU/WCU || - Up to 20 GSIs per table || |+------------------------------------------------------------------+Local Secondary Indexes (LSI)
Section titled “Local Secondary Indexes (LSI)” Local Secondary Indexes+------------------------------------------------------------------+| || Purpose: Alternative sort key for same partition key || || Architecture: || +----------------------------------------------------------+ || | | || | Table: Orders | || | +----------------------------------------------------+ | || | | PK: UserId | SK: OrderDate | Amount | Status | | || | |-------------|-------------|--------|-----------| | || | | USER1 | 2024-01-01 | 100 | shipped | | || | | USER1 | 2024-01-15 | 200 | pending | | || | | USER2 | 2024-01-02 | 150 | shipped | | || | +----------------------------------------------------+ | || | | | || | v | || | LSI: AmountIndex | || | +----------------------------------------------------+ | || | | PK: UserId | SK: Amount | OrderDate | Status | | || | |-------------|-------------|-----------|---------| | || | | USER1 | 100 | 2024-01-01| shipped | | || | | USER1 | 200 | 2024-01-15| pending | | || | | USER2 | 150 | 2024-01-02| shipped | | || | +----------------------------------------------------+ | || | | || +----------------------------------------------------------+ || || Features: || - Same partition key as table || - Different sort key || - Must be created at table creation || - Up to 5 LSIs per table || - Shares RCU/WCU with table || |+------------------------------------------------------------------+DynamoDB Streams
Section titled “DynamoDB Streams” DynamoDB Streams+------------------------------------------------------------------+| || Purpose: Capture table modifications || || Architecture: || +----------------------------------------------------------+ || | | || | DynamoDB Table | || | +------------------+ | || | | Item Operations | | || | | - INSERT | | || | | - MODIFY | | || | | - REMOVE | | || | +------------------+ | || | | | || | v | || | DynamoDB Stream | || | +------------------+ | || | | Stream Records | | || | | (24-hour retention) | || | +------------------+ | || | | | || | +----+----+ | || | | | | || | v v | || | +------+ +----------+ | || | |Lambda| | Kinesis | | || | | | | KCL App | | || | +------+ +----------+ | || | | || +----------------------------------------------------------+ || || Stream View Types: || +----------------------------------------------------------+ || | - KEYS_ONLY: Only key attributes | || | - NEW_IMAGE: Entire new item | || | - OLD_IMAGE: Entire old item | || | - NEW_AND_OLD_IMAGES: Both versions | || +----------------------------------------------------------+ || |+------------------------------------------------------------------+Global Tables
Section titled “Global Tables” DynamoDB Global Tables+------------------------------------------------------------------+| || Purpose: Multi-region, multi-active replication || || Architecture: || +----------------------------------------------------------+ || | | || | US-East-1 EU-West-1 AP-South-1 | || | +----------+ +----------+ +----------+ | || | | Table | | Table | | Table | | || | | Replica |<-------->| Replica |<------>| Replica | | || | +----------+ +----------+ +----------+ | || | ^ ^ ^ | || | | | | | || | Application Application Application | || | (Read/Write) (Read/Write) (Read/Write)| || | | || +----------------------------------------------------------+ || || Features: || - Multi-active replication || - Last-writer-wins conflict resolution || - < 1 second replication || - Requires DynamoDB Streams || - Same table name in all regions || |+------------------------------------------------------------------+23.5 DynamoDB Accelerator (DAX)
Section titled “23.5 DynamoDB Accelerator (DAX)” DynamoDB DAX+------------------------------------------------------------------+| || Purpose: In-memory caching for DynamoDB || || Architecture: || +----------------------------------------------------------+ || | | || | Application | || | +------------------+ | || | | | | || | +------------------+ | || | | | || | v | || | DAX Cluster | || | +------------------+ | || | | Primary Node | | || | | +------------+ | | || | | | Cache | | | || | | +------------+ | | || | | | Replica 1 | | | || | | | Replica 2 | | | || | +------------------+ | || | | | || | v | || | DynamoDB Table | || | +------------------+ | || | | | | || | +------------------+ | || | | || +----------------------------------------------------------+ || || Features: || - Microsecond latency || - Write-through caching || - Eventually consistent reads || - Automatic failover || - Compatible with DynamoDB API || || Performance: || - 10x improvement over DynamoDB || - Up to millions of requests per second || |+------------------------------------------------------------------+23.6 Practical Configuration
Section titled “23.6 Practical Configuration”DynamoDB with Terraform
Section titled “DynamoDB with Terraform”# ============================================================# DynamoDB Table# ============================================================
resource "aws_dynamodb_table" "users" { name = "users" billing_mode = "PAY_PER_REQUEST" # On-demand
# Primary Key hash_key = "UserId" range_key = "CreatedAt"
# Attributes attribute { name = "UserId" type = "S" }
attribute { name = "CreatedAt" type = "N" }
# Global Secondary Index attribute { name = "Email" type = "S" }
global_secondary_index { name = "EmailIndex" hash_key = "Email" range_key = "CreatedAt" projection_type = "ALL" }
# Stream stream_enabled = true stream_view_type = "NEW_AND_OLD_IMAGES"
# TTL ttl { attribute_name = "ExpiresAt" enabled = true }
# Point-in-time Recovery point_in_time_recovery { enabled = true }
# Server-side encryption server_side_encryption { enabled = true kms_key_arn = aws_kms_key.dynamodb.arn }
# Table class table_class = "STANDARD" # or "STANDARD_INFREQUENT_ACCESS"
tags = { Name = "users-table" }}
# ============================================================# DynamoDB Table with Provisioned Capacity# ============================================================
resource "aws_dynamodb_table" "orders" { name = "orders" billing_mode = "PROVISIONED" read_capacity = 20 write_capacity = 10
hash_key = "OrderId" range_key = "OrderDate"
attribute { name = "OrderId" type = "S" }
attribute { name = "OrderDate" type = "S" }
# Local Secondary Index attribute { name = "Amount" type = "N" }
local_secondary_index { name = "AmountIndex" range_key = "Amount" projection_type = "ALL" }
tags = { Name = "orders-table" }}
# ============================================================# Auto Scaling for Provisioned Table# ============================================================
# Read capacity auto scalingresource "aws_appautoscaling_target" "read" { max_capacity = 100 min_capacity = 5 resource_id = "table/${aws_dynamodb_table.orders.name}" scalable_dimension = "dynamodb:table:ReadCapacityUnits" service_namespace = "dynamodb"}
resource "aws_appautoscaling_policy" "read" { name = "read-scaling-policy" policy_type = "TargetTrackingScaling" resource_id = aws_appautoscaling_target.read.resource_id scalable_dimension = aws_appautoscaling_target.read.scalable_dimension service_namespace = aws_appautoscaling_target.read.service_namespace
target_tracking_scaling_policy_configuration { target_value = 70
predefined_metric_specification { predefined_metric_type = "DynamoDBReadCapacityUtilization" } }}
# Write capacity auto scalingresource "aws_appautoscaling_target" "write" { max_capacity = 100 min_capacity = 5 resource_id = "table/${aws_dynamodb_table.orders.name}" scalable_dimension = "dynamodb:table:WriteCapacityUnits" service_namespace = "dynamodb"}
resource "aws_appautoscaling_policy" "write" { name = "write-scaling-policy" policy_type = "TargetTrackingScaling" resource_id = aws_appautoscaling_target.write.resource_id scalable_dimension = aws_appautoscaling_target.write.scalable_dimension service_namespace = aws_appautoscaling_target.write.service_namespace
target_tracking_scaling_policy_configuration { target_value = 70
predefined_metric_specification { predefined_metric_type = "DynamoDBWriteCapacityUtilization" } }}
# ============================================================# Global Table# ============================================================
# Primary region tableresource "aws_dynamodb_table" "global" { provider = aws.primary
name = "global-table" billing_mode = "PAY_PER_REQUEST"
hash_key = "Id"
attribute { name = "Id" type = "S" }
stream_enabled = true stream_view_type = "NEW_AND_OLD_IMAGES"
# Enable global table replica { region_name = "us-east-1" }
replica { region_name = "eu-west-1" }}
# ============================================================# DAX Cluster# ============================================================
resource "aws_dax_cluster" "main" { cluster_name = "main-dax" node_type = "dax.t3.small" replication_factor = 3
# IAM role iam_role_arn = aws_iam_role.dax.arn
# VPC configuration subnet_group_name = aws_dax_subnet_group.main.name security_group_ids = [aws_security_group.dax.id]
# Parameter group parameter_group_name = "default.dax-1.0"
tags = { Name = "main-dax" }}
resource "aws_dax_subnet_group" "main" { name = "main-dax-subnet" subnet_ids = var.private_subnet_ids}
# ============================================================# DynamoDB Stream to Lambda# ============================================================
resource "aws_lambda_function" "process_stream" { function_name = "process-dynamodb-stream" role = aws_iam_role.lambda.arn runtime = "python3.11" handler = "index.handler"
filename = "function.zip"
source_code_hash = filebase64sha256("function.zip")}
resource "aws_lambda_event_source_mapping" "stream" { event_source_arn = aws_dynamodb_table.users.stream_arn function_name = aws_lambda_function.process_stream.arn starting_position = "LATEST"
# Batch size batch_size = 100
# Filter criteria (optional) filter_criteria { filter { pattern = jsonencode({ dynamodb = { Keys = { UserId = { S = ["USER1"] } } } }) } }}23.7 DynamoDB API Operations
Section titled “23.7 DynamoDB API Operations” DynamoDB Operations+------------------------------------------------------------------+| || Item Operations || +----------------------------------------------------------+ || | | || | PutItem: Create or replace item | || | GetItem: Retrieve item by key | || | UpdateItem: Update item attributes | || | DeleteItem: Delete item by key | || | BatchGetItem: Get multiple items | || | BatchWriteItem: Put/delete multiple items | || | | || +----------------------------------------------------------+ || || Query Operations || +----------------------------------------------------------+ || | | || | Query: Query by partition key + optional sort key | || | Scan: Read all items in table/index | || | | || | Query Parameters: | || | - KeyConditionExpression | || | - FilterExpression | || | - ProjectionExpression | || | - Limit | || | - ExclusiveStartKey (pagination) | || | | || +----------------------------------------------------------+ || || Conditional Writes || +----------------------------------------------------------+ || | | || | ConditionExpression: | || | - attribute_exists(attr) | || | - attribute_not_exists(attr) | || | - attr = value | || | - attr < value | || | | || | Use Cases: | || | - Prevent overwrites | || | - Optimistic locking | || | | || +----------------------------------------------------------+ || |+------------------------------------------------------------------+23.8 Exam Tips
Section titled “23.8 Exam Tips”
- Capacity Modes: On-demand (pay per request), Provisioned (RCU/WCU)
- RCU: 1 RCU = 1 strongly consistent read (4 KB)
- WCU: 1 WCU = 1 write (1 KB)
- GSI: Different partition key, separate capacity
- LSI: Same partition key, different sort key, created at table creation
- Streams: 24-hour retention, triggers Lambda
- Global Tables: Multi-region, multi-active, requires streams
- DAX: In-memory cache, microsecond latency
- TTL: Automatic item expiration
- Conditional Writes: Optimistic locking, prevent overwrites
Next Chapter
Section titled “Next Chapter”Chapter 24: Amazon ElastiCache - In-Memory Caching
Last Updated: February 2026