Amazon Neptune

Amazon Neptune is a managed graph database, or a database-as-a-service, product offered by Amazon Web Services (AWS). The product was announced on November 29, 2017 and was launched to general availability on May 30, 2018. Amazon Neptune achieved HIPAA eligibility on September 12, 2018 and Amazon Web Services announced on December 12, 2018 that the product was in-scope for Payment Card Industry Data Security Standard (PCI DSS) certification and ISO certification. Amazon Neptune is also covered under AWS's ISO 9001, 27001, 27017, and 27018 certifications, which are for global security standards, information security management, and the protection of personally identifiable information.

Amazon Neptune is built to work with connected datasets, the core of which is built to be a graph database engine optimized for storing billions of relationships and querying the graph with milliseconds of latency. The product supports popular graph models Property Graph and W3C's RDF, and their respective query languages Apache TinkerPop Gremlin and SPARQL. Neptune is intended to help build queries for product customers, including recommendation engines, fraud detection, knowledge graphs, drug discovery, and network security.

Through AWS's management console, Amazon Neptune's compute and memory resources that power the production cluster will change, with increased computing and memory resources automatically allocated to the graph database as necessary. As well, Amazon Neptune will automatically grow the size of the product users databased volume as storage needs grow, increasing in increments of 10 GB up to a maximum of 64 TB. As well, in order to reduce overall latency, Amazon Neptune creates up to fifteen database read replicas, which share the underlying storage as the source instance that lowers overall costs and avoids the need to perform writes at the replica nodes.

Amazon Neptune offers continuous monitoring of instances. If the instance powering a database fails, the database and associated processes are automatically restarted. These restart times are typically thirty seconds or less. During an instance failure, Amazon Neptune automates failover of up to 15 Neptune replicas created. In the case where no replicas exist, Neptune attempts to create a new database instance automatically. Each 10GB chunk of a database is replicated six ways. Neptune's storage is also self-healing, with data blocks and disks continuously scanned for errors and replaced.

Neptune supports the property graph model using the open source Apache TinkerPop Gremlin traversal language and provides a Gremlin Websockets server that supports TinkerPop version 3.3. And existing Gremlin applications can use Neptune by changing the Gremlin service configuration point. As well, Neptune supports the W3C's Semantic Web standards of RDF 1.1 and SPARQL 1.1 and provides and HTTP REST endpoint.

Neptune runs in Amazon VPC (Virtual Private Cloud), which allows a user to isolate their database on their own virtual network and connect to an on-premises IT infrastructure using encrypted IPsec VPNs. The VPC also allows for firewall configuration and network access control. Network access control also comes with identity and access management integrated through the AWS Identity and Access Management (IAM) product. As well, users can encrypt their databases using keys created through AWS Key Management Service, which encrypts the instance running, data stored at rest, and the automated backups, snapshots, and replicas. Database events can also be logged, and these logs can later be analyzed for database management, security, governance, and regulatory compliance.

Upon launching a Neptune database instance through the AWS management console, the database is pre-configured based on the parameters and settings chosen by a user for the instance class selected. It can be launched in relatively little time, without additional configuration. The parameter groups also offer control and fine-tuning of the database. Other AWS products also allow a user to automatically manage and monitor a database. This can include viewing key operational metrics for a database instance (including computing, memory, storage, query throughput, and active connections) or notifications through email or SMS of important database events.

Amazon Neptune supports parallel bulk loading for Property Graph data that is stored in S3, which can in turn be given a specified location through a REST interface. As well, Neptune uses CSV delimited format to load data into the Nodes and Edges. The N-Triples (NT), N-Quads (NQ), RDF/XML, and Turtle RDF 1.1 serializations are supported by Neptune through RDF bulk loading.

Neptune powers graph use cases such as recommendation engines, fraud detection, knowledge graphs, drug discovery, and network security.

Amazon Neptune is priced through on-demand pricing, which allows users to pay for a database by the hour with no long-term commitments. This is intended to allow users to pay for database use as they need, rather than based on a planned need. Instance pricing applies to primary instances and replicas, such that the cost of replicas is the cost of the primary instance plus the cost of a replica. The cost of a standard db.t3.medium instance is $0.098 per hour.

Storage consumed by an Amazon Neptune database is billed in per GB-month increments and IOs consumed are billed in per million request increments. This results in a storage rate of $0.10 per GB a month and the IO rate at $0.20 per one million requests.

Backup storage with Amazon Neptune is the storage associated with automated database backups and customer-initiated database cluster snapshots. The backup storage is charged at $0.021 per GB a month.

Data transfer is charged based on the data transferred "in" and "out" of Amazon Neptune. For data transferred out, the base price is $0.00 per GB up to one GB per month. This changes dependent on region, ranging from $0.01 to $0.02 per GB per month.

This describes the pricing for different instances running in Ready state for Amazon Neptune and the virtualized computer in use.