Amazon Neptune
Cloud service · AWS · Updated

Amazon Neptune

In short

Amazon Neptune is a fast, reliable, and fully managed graph database service that makes it easy to build and run applications that work with highly connected datasets.

Cloud Services Hub →

7 key features of Amazon Neptune

Amazon Neptune is a managed graph database service with support for multiple graph models, designed for high scalability and availability, fast performance, robust security features, and ease of use - all while being tightly integrated with the AWS ecosystem.

01
Fully Managed Graph Database Service

Amazon Neptune is a fully managed graph database service that simplifies the setup, operation, and scaling of graph databases, allowing users to build and run applications that work with highly connected datasets without worrying about database management tasks.

02
Support for Popular Graph Models

Neptune supports popular graph models like Property Graph and the RDF (Resource Description Framework), along with their respective query languages such as Apache TinkerPop Gremlin and SPARQL, enabling flexibility in building graph-based applications.

03
Highly Scalable and Available

The service is designed to offer high performance at scale with read replica, point-in-time recovery, and continuous backup to Amazon S3, ensuring high availability and durability for graph databases.

04
Fast and Predictable Performance

Neptune is optimized for low-latency, high-throughput performance, and can handle complex query patterns that involve multiple hops and large data sets with ease.

05
Enterprise-Ready with Security Features

It includes security features such as encryption at rest and in transit, IAM authentication, and support for VPC, making it suitable for sensitive and regulatory-compliant workloads.

06
Fully Integrated with the AWS Ecosystem

Neptune easily integrates with other AWS services offering seamless connectivity for data import/export, analytics, and AI/ML capabilities.

07
Easy to Use

Amazon Neptune offers a user-friendly console and APIs, making it simple to manage clusters, instances, and other resources, significantly reducing the complexity typically associated with graph database management.

Where Amazon Neptune is used

Amazon Neptune is used for analyzing social networks, detecting fraud in financial transactions, building knowledge graphs for improved insights, monitoring network and IT operations, personalizing recommendations in e-commerce, and managing complex data in life sciences.

  1. 01
    Social Network Analysis

    Amazon Neptune can be used to manage and navigate social networks by handling complex relationships and connections between data points. It supports the implementation of feeds, recommendations, and friend-finding algorithms by efficiently querying connected data.

  2. 02
    Fraud Detection

    By analyzing data relationships, Neptune helps detect potentially fraudulent patterns and anomalies in financial systems. Its graph capabilities enable real-time processing of banking transactions to identify and prevent fraud.

  3. 03
    Knowledge Graphs

    Organizations use Neptune to build knowledge graphs that store and link large volumes of data across various sources in an interconnected way, supporting more effective data discovery, improving business insights, and enhancing customer experience.

  4. 04
    Network and IT Operations

    Neptune can be employed for monitoring network infrastructure, managing IT operations, and ensuring security compliance. Its ability to quickly traverse network datasets aids in root cause analysis and real-time network optimization.

  5. 05
    Recommendation Systems

    By using graph databases, Amazon Neptune can power recommendation engines that personalize user experiences on e-commerce sites or content platforms through fast and accurate relationship queries.

  6. 06
    Life Sciences Data Management

    In the life sciences field, Neptune aids in managing complex biological, chemical, and medical data. It helps in drug discovery processes by linking vast datasets and uncovering insights into relationships between genetic and protein functions.

How Amazon Neptune is priced

Amazon Neptune offers on-demand and reserved instances with hourly pricing, backup storage charges beyond a free tier, standard data transfer rates, and fees for additional I/O requests.

On-Demand Instances

Amazon Neptune charges on an hourly basis depending on the instance class and region. You pay only for the compute time that you consume without any upfront commitments.

Backup Storage

Backup storage pricing applies to both Neptune Database and Neptune Analytics. Backup storage up to 100% of each cluster's total Neptune storage has no additional charge; excess backup storage and snapshots retained after cluster deletion are charged.

Data Transfer

Transfers between Neptune and EC2 instances in the same Availability Zone are free, as is cross-AZ replication for Multi-AZ deployments. For transfers between EC2 and Neptune instances inside a VPC in different Availability Zones of the same Region, EC2 Regional data transfer charges apply on both sides.

I/O Rate

Usage-based I/O charges are billed per million requests consumed, without requiring advance provisioning.

Database Savings Plans

Database Savings Plans offer savings in exchange for a commitment to a specific amount of usage, measured in dollars per hour, over a one-year term.

Neptune Serverless

Neptune Serverless capacity is billed per second in Neptune Capacity Units (NCUs). Each NCU provides approximately two GB of memory with corresponding CPU and networking capacity.

Neptune Analytics

Neptune Analytics is billed using memory-optimized Neptune Capacity Units (m-NCUs), each corresponding to one GB of memory and associated compute and networking capacity for one hour. A paused graph costs 10% of its normal compute price while preserving its data and settings.

Database Storage

Neptune Database storage is billed in GB-month increments for storage consumed, without requiring advance provisioning.

Neptune I/O-Optimized

Neptune I/O-Optimized charges for database instances and storage usage, with no charges for read or write I/O operations.

Neptune Global Database

Neptune Global Database charges for replicated write I/Os between the primary Region and each secondary Region, in addition to standard instance, storage, and backup storage charges.

Neptune Workbench

Neptune workbench notebook instances are charged per instance-hour while in the Ready state, independently of the selected Neptune engine.

T4g and T3 CPU Credits

T4g and T3 instances run in unlimited mode and incur CPU credit charges when average CPU usage over a rolling 24-hour period exceeds the instance baseline. CPU credits cost $0.15 per vCPU-hour across all AWS Regions.

Neptune Database Free Trial

Customers new to Neptune can receive 750 hours of db.t3.medium or db.t4g.medium instance usage, 10 million I/O requests, one GB of storage, and one GB of backup storage free for 30 days starting with their first Neptune Database cluster. Usage beyond the allowance or trial period is charged at standard on-demand rates.

Services Amazon Neptune works with

Amazon CloudWatchAmazon CloudWatch

Delivers monitoring capabilities via logs and metrics to gain insight into the performance, health, and usage of Neptune resources.

AWS LambdaAWS Lambda

Enables you to run code in response to triggers from Amazon Neptune, facilitating serverless data processing or ETL tasks.

Amazon API GatewayAmazon API Gateway

The Neptune export service exposes an API through API Gateway for starting and monitoring graph-data exports. This lets applications initiate exports without running the Neptune export utility themselves.

Amazon AthenaAmazon Athena

Neptune provides an Athena federated query connector so you can query graph data using SQL. You can join results from Neptune with data in Amazon S3 or other Athena data sources.

Amazon Bedrock

Amazon Bedrock Knowledge Bases uses Neptune Analytics to store and query graphs for GraphRAG. This allows retrieval to follow relationships between entities and provide richer context for generated answers.

Amazon DynamoDBAmazon DynamoDB

Neptune ingestion reference implementations can consume DynamoDB Streams and translate item changes into graph updates. This supports maintaining a graph representation of relationships in DynamoDB application data.

Amazon EMRAmazon EMR

Neptune's Spark integration supports graph-data processing and loading from Spark jobs running on Amazon EMR. Jobs can transform large datasets into vertices and edges before writing them to Neptune.

Amazon EventBridgeAmazon EventBridge

Neptune management API activity recorded through CloudTrail can be matched by EventBridge rules. For example, a rule can trigger an automated workflow when a Neptune cluster is deleted or its configuration changes.

Amazon Kinesis Data StreamsAmazon Kinesis Data Streams

Neptune provides reference ingestion implementations that consume Kinesis Data Streams through Lambda and write records as graph updates. This supports continuously loading high-volume event data into a graph.

What is Amazon Neptune

4 min read

Amazon Neptune is a fast, reliable, and fully managed graph database service that is part of the Amazon Web Services (AWS) portfolio. It is designed to store and navigate relationships between data points in a highly efficient manner, offering both high availability and durability for various applications. With its capacity to handle complex queries over large datasets, Neptune has become a potent tool for developers working on applications that involve highly connected data, such as social networks, fraud detection, recommendation engines, and knowledge graphs. 

One of the key features of Amazon Neptune is its support for multiple popular graph models and query languages. It supports property graph queries using Apache TinkerPop Gremlin, and for RDF (Resource Description Framework) graph queries, it's compatible with the SPARQL query language. This dual model approach provides developers with the flexibility to utilize the graph model that best fits their domain and query requirements. Data can be represented as vertices and edges in property graphs or as triples or quads in RDF graphs, with each model offering its own set of capabilities and syntaxes. 

Neptune also supports openCypher for property-graph queries.

Furthermore, Neptune is optimized for storing billions of relationships and querying the graph with milliseconds latency. The database is designed to offer high throughput for both read and write operations, which is essential for applications that require real-time insights from their connected data. 

Each Neptune DB cluster has one primary DB instance that performs all writes and modifications to the cluster volume.

Neptune Analytics is a separate analytics database engine for graph analytics, graph algorithms, and vector search. It can load graph data from an existing Neptune Database or Amazon S3 and analyze tens of billions of relationships in seconds.

Amazon has engineered Neptune to be highly available and durable. It automatically replicates data across multiple availability zones in an AWS region, and it also provides continuous backups to Amazon S3, thus safeguarding against data loss and providing the ability to restore data if necessary. 

Neptune Global Database is an optional multi-Region configuration providing low-latency global reads and disaster recovery for Region-wide outages.

Amazon Neptune is fully managed, which means that AWS handles much of the heavy lifting involved in database management, such as hardware provisioning, software patching, setup, configuration, or backups. Developers can thus focus on building their applications rather than worrying about the underlying infrastructure. 

Neptune management operations can be asynchronous, requiring polling or callbacks to determine when a command has been applied. Depending on the operation, changes can apply immediately, on the next instance reboot, or during the maintenance window.

Neptune and Amazon RDS share certain per-account, per-Region resource limits, including limits involving DB instances and DB clusters. Once a resource limit is reached, additional creation calls fail with an exception.

Neptune Database also offers Neptune Serverless, which automatically scales database capacity up or down to meet application workload needs.

A Neptune Serverless instance supports a maximum capacity of 128 NCUs, corresponding to 256 GB of memory.

The Neptune lookup cache is not compatible with Serverless DB instances.

Changing a Serverless cluster's capacity range changes some configuration defaults; some changes require a reboot, indicated by a pending-reboot status.

Neptune Database clusters support Standard and I/O-Optimized configurations. Standard suits typical access patterns with low-to-moderate I/O usage, while I/O-Optimized is designed to improve price performance for I/O-intensive applications.

Additionally, Neptune integrates with other AWS services, offering additional capabilities like identity and access management through AWS IAM, encryption at rest and in transit, and monitoring via Amazon CloudWatch. Given its cloud-native nature, Amazon Neptune is built to scale. It provides the ability to handle an increasing number of requests without degrading performance. It can scale read operations by adding read replicas, and storage automatically grows as you add more data, ensuring that your applications have the resources they need as they grow. 

With its emphasis on security, performance, and ease of use, Amazon Neptune is tailored for modern, data-driven application development. Whether it's building a complex knowledge graph that powers intelligent applications, maintaining a social network with millions of connections, or identifying patterns for fraud detection, Neptune offers a robust and scalable solution for managing and querying relational data. Its managed service approach frees up developers from operational tasks, allowing them to focus on innovation and creating value through their applications.

Concepts related to Amazon Neptune

Amazon Neptune exam questions

Real practice exam questions on Amazon Neptune, with the correct answer and its explanation. Answer them with every option in the exam simulator.

SAA-C03

A manufacturer is building a component-recall application. The application must trace relationships among suppliers, parts, assemblies, and finished products to identify products affected by a defective part. Queries follow a variable number of relationships across millions of entities. Relationships change throughout the day, and queries must reflect committed changes within seconds.

A prototype that stores every possible dependency path requires excessive storage and expensive updates whenever relationships change. The company wants a fully managed database that can query these relationships directly, without maintaining precomputed dependency paths or implementing multi-step traversal logic in application code.

Which solution BEST meets these requirements while minimizing data-maintenance and application-development costs?

Show answerCORRECT ANSWER

Use Amazon Neptune to store entities as vertices and relationships as edges. Execute graph traversal queries against the current relationships.

Neptune is a fully managed graph database designed for highly connected datasets and multi-hop relationship queries. Graph queries can follow the required dependencies directly without storing every possible path or implementing traversal through application-managed database requests. This reduces the maintenance and development costs described in the scenario.

CLF-C02

A manufacturing company needs to track how components, suppliers, and finished products are connected. When a component is recalled, analysts must follow several levels of relationships to identify all affected products. The company wants a fully managed NoSQL database optimized for querying these connections. Which AWS service is the best choice?

Show answerCORRECT ANSWER

Amazon Neptune

Neptune is a fully managed graph database designed for highly connected datasets. It can model components, suppliers, and products as entities and efficiently follow their relationships to identify products affected by a recall.

Practice Amazon Neptune in a full exam

Questions on Amazon Neptune in each practice exam, with every answer option and an explanation for each:

Start a free practice exam

Learn Amazon Neptune in lessons

5 short lessons on Amazon Neptune in 3 certification courses.

Open the lessons

AWS architecture icons are the property of Amazon Web Services. BlowStack is not affiliated with or endorsed by AWS.