Introducing Developer Deployments on ArangoDB ArangoGraph

Estimated reading time: 4 minutes

Today we’re announcing the introduction of Developer deployments as a beta feature on the ArangoGraph platform.

In this blog post, we’ll tell you what Developer deployments are, what you can do with them, what you should not do with them, and how to get started.

What are Developer deployments?

Since we launched ArangoGraph, a deployment on ArangoGraph has always been a highly available ArangoDB cluster. That is great for high availability, and it allows you to scale your deployment to incredibly large sets of data.

Many customers have told us that they would like something a bit smaller that can be used by an individual developer, or better yet, let every developer have their own deployment.

That request is exactly what we are answering with the introduction of Developer deployments. 

A Developer deployment consists of only a single server on a single node.

With that configuration, there is obviously no high availability, and scaling is limited to vertical scaling of that single server.

For those reasons, Developer deployments are not suitable for any kind of production environment, and because of that, Developer deployments are excluded from audit logs.

Support is given for Developer deployments on a best effort basis. It is not possible to buy an additional support plan with a Developer deployment.

All other features, like backups, full encryption, and Foxx, are fully available.

What can you do with a Developer deployment?

Developer deployments are ideal when you want to experiment with ArangoDB, or are just learning its features. 

They are also ideal for (small scale) analytics experiments. You can quickly load your data into them, configure your graphs and perform your analysis.

The lack of high availability is usually not a problem for such experiments.

Of course, you can also give a single Developer deployment to all of your developers to develop an application against. When you do so, you have to keep in mind that there are small differences between a single instance of ArangoDB and a cluster.

Within ArangoGraph, we have reduced these changes by requiring the use of the WITH statement in exactly the same way that a cluster deployment requires that keyword (from ArangoDB version 3.7.12 and higher).

How are Developer deployments priced?

A Developer deployment launched in AWS Ohio is available for as little as $ 0.058 per hour or $42.2 per month. The same deployment with General CPU is available for $ 0.068 per hour.

Similar to our OneShard and Sharded deployments, network traffic and backup storage is charged separately.

For more details on pricing, please log in to the ArangoGraph dashboard and visit the Pricing page.

What not do with Developer deployments

As mentioned before, Developer deployments are not highly available and can be restarted at any time. For that reason, you should not use them for any kind of production environment.

Since we strongly believe that a staging environment should reflect the production environment as much as possible, you should also not use a Developer deployment for a staging environment.

How to get started with Developer deployments

To create a Developer deployment on ArangoGraph, log in to your ArangoDB ArangoGraph account on cloud.arangodb.com.
Then go to your project and click on New Deployment.

Enter all the normal fields such as the name of your deployment and select a cloud provider & region.
Then click on the Developer (beta) button to choose the Developer model.

Select the size of your deployment and click on Create.

Your Developer deployment will now bootstrap.

Once that has finished, you’ll receive an email and can start to use your deployment.

Hear More from the Author

OCB: Challenges in Building Multi-Cloud-Provider Platform With Kubernetes

Getting Started with ArangoDB ArangoGraph

Continue Reading

ArangoML Part 1: Where Graphs and Machine Learning Meet

A story of a memory leak in GO: How to properly use time.After()

ArangoDB 3.7 – A Big Step Forward for Multi-Model