Plans and Placements

GrapheneDB offers a wide spectrum of plans for different data set sizes, workloads, and development stages. Choosing the right plan will depend on your dataset and data model, complexity of queries, and throughput expectations. Other factors to take into account are development stage, product availability, and support expectations from your team.

You can get started with GrapheneDB and Neo4j with our free Sandbox plan in the Hobby tier and scale as you need.

For detailed information, features and prices of our different our tiers and plans, please refer to our pricing page. You will also find a feature comparison matrix for our plan tiers on our plans page.

Plans and tiers

Hobby Tier

Hobby databases are designed for learning Neo4j, developing, and testing your code.

Hobby databases are not meant to run a production application and have the following limitations:

• Enforced storage limits by node and relationship count:
  • 1,000 nodes & 10,000 relationships in Sandbox (free)
  • 100,000 nides & 1,000,000 relationships in Developer
• Asleep when there is no inbound traffic during a 30 minute time window
• No scheduled or on-demand hot backups
• Unannounced maintenance windows
• Inability to configure memory distribution and limited instance memory
• Insights (metrics and slow queries) with 1 hour max. retention

Please refer to the Storage Limit Enforcement section for further information on how limits are enforced.

🚧

Hobby databases are unsuitable for production.

Standard Tier

Standard databases are suitable for budget-conscious projects with small datasets and light server loads.

All Standard databases include:

• Unlimited nodes and relationships (storage limited by disk size only)
• Announced maintenance windows
• Scheduled and on-demand hot backups
• Ability to configure memory distribution and limited instance memory.
• Insights (metrics and slow queries) with 12 hour max. retention
• 24x7monitoring
• 24x7 urgent tickets with 1-hour response time

📘

Unlike Hobby databases, Standard databases are never put to sleep.

Performance Tier

Performance databases are best suited for production-level server workloads.

All Performance databases include:

• Unlimited nodes and relationships (storage limited by disk size only)
• Encryption at rest
• Private networks: IP whitelisting & VPC peering
• Self-service / scheduled maintenance windows
• Scheduled and on-demand hot backups
• Ability to configure memory distribution and limited instance memory.
• Insights (metrics and slow queries) with 7 days max. retention
• 24x7 monitoring
• 24x7 urgent tickets with 1-hour response time
• Priority service restoration on disruptions

📘

Unlike Hobby databases, Performance databases are never put to sleep

Enterprise Tier

Enterprise databases are the best choice for mission critical projects with highly concurrent traffic and enterprise-level resiliency. Enterprise database deployments run on Neo4j Enterprise Edition, with 3 or more server nodes which are usually deployed across multiple availability zones for higher resiliency.

Enterprise deployments require a commercial Neo4j Enterprise license, which can be provided by the client (bring your own) or by GrapheneDB.

All Enterprise databases include:

• Resilient, highly available multi-AZ deployments with 3 or more servers
• Neo4j Enterprise features such as Enterprise Lock Manager, Schema constraints, High Availability
• Unlimited nodes and relationships (storage limited by disk size only)
• Encryption at rest
• Private networks: IP whitelisting & VPC peering
• Self-service / scheduled maintenance windows
• Scheduled and on-demand hot backups
• Ability to configure memory distribution and limited instance memory
• Insights (metrics and slow queries) with 7 days max. retention
• 24x7 monitoring
• 24x7 urgent tickets with 1-hour response time
• Priority service restoration on disruptions
• Technical account manager

📘

Unlike Hobby databases, Enterprise databases are never put to sleep

Storage Limit Enforcement

Database storage is limited by disk size in the Standard, Performance, and Enterprise tiers. It is uncommon for a user to hit the size limit on these plans.

Storage in Hobby databases is limited by the node and relationship count stored in the database, as well as by size. It is possible for users to hit these count limits, especially in scenarios where data is bulk loaded from a script or CSV file.

GrapheneDB sends the following automated notifications to inform the user when the database is close to its storage limit:

1. A warning email is sent to the account owner when a database reaches 70% of the storage limit (count or disk size)
2. A final notification is sent to the account owner when a database reaches 100% of its storage limit. At this point, the database is set to read-only mode

Information on how to deal with databases that have exceeded their limits and have been switched to read-only can be found here.

Placements

To ensure maximum performance, databases should be hosted as close as possible to the application. GrapheneDB offers hosting in most AWS regions.

The Hobby tier is only available in certain regions. The following table includes the tiers available by AWS region:

Heroku

GrapheneDB is available on Heroku as add-on on the Elements marketplace.

The add-on is available in the Heroku regions Heroku US and Heroku EU, with databases being co-located in the same regions where Heroku applications are hosted (N.Virginia and Ireland).

For further information on the Heroku add-on, please visit the GrapheneDB page in the Heroku Dev Center.

Other PaaS and IaaS Providers

GrapheneDB databases managed by GrapheneDB can also be used from Heroku and other PaaS providers, as well as servers hosted with other IaaS and hosting providers.

Use the table below to pick the closest location.

For other providers and regions, we recommend to measure latencies to identify the AWS regions with the lowest network latency to your application servers.