SQL Clone
SQLServerCentral is supported by Redgate
Log in  ::  Register  ::  Not logged in

Azure Cosmos DB Free Trial – walk through of Gremlin API for Graph

After the 2018 Microsoft Ignite event, Microsoft announced a free trial of Azure Cosmos DB. For those who are eager to check out Cosmos DB, this is a great opportunity to familiarize yourself with the hottest No SQL database in the market. Cosmos DB is currently sitting pretty at Rank 29 in the DB Engines Ranking page. Quite an achievement considering that the product is “only 1 year old”.

MS recently went public with the support for Cassandra API. A few years from now, Cosmos DB will be the most popular database offering from Microsoft. If you want to get a gentle intro to Cosmos DB, check out my previous posts – Introduction to Azure Cosmos DB and Azure Cosmos DB – Partition and Throughput.

This article is a walk through of using the Free Trial to get started with Cosmos DB.

Limited Time Free Trialhttps://azure.microsoft.com/en-au/try/cosmosdb/

There is no need for a credit card or subscription to avail this free trial. Please do note though that MS is likely to withdraw trial in a few months.

Step 1: Choosing an API/data model
Once you click the above URL, you will be asked to pick an API and data model. Go ahead and pick SQL to check out document (JSON) data model if you are unsure. I want to check out Gremlin API for Graph, so I am choosing Graph in this example.
You will be asked to login with your Microsoft account. If you don’t have one, create a new account.

CosmosDB Trial

Step 2: Choose default options and click on Create container.
Note that container will be created free of cost at a low throughput of 400 RU/s. If you are not familiar with throughput, read my article about Azure Cosmos DB – Partition and Throughput.
You now have a brand new container and the familiar Azure portal access. My trial container has read/write location of Central US.

Azure Portal Database

Closer look: There is a lot going on here, so let us take a closer look.
My container is called Persons and it is in graphdb database. Cosmos DB uses Apache Tinkerpop Gremlin API for graph traversal. Currently, my container Persons is empty, so let us connect a sample application. If you are new to Graph DB, I recommend reading this free ebook from another popular Graph DB product – Neo4j.

Graph DB - Persons

Step 3: Sample project
Click on the Quick start blade to download a sample project and explore graph data.
The sample project already has the connection string set to the trial database, so we can execute the project right away. How awesome is that!


In Step 1, I chose Graph API and model, so my sample project has Gremlin API queries. Read more about Gremlin API here – https://aka.ms/gremlin.

Gremlin API Queries

Step 4: Execute sample project
Now it’s time to execute the project!

Sample Program Execution

Step 5: Data Explorer
Looks like the program has added a number of graph documents to the container. Let us head over to Azure Portal and explore using Data Explorer blade. Click on the Execute Gremlin Query button to take a look at the data added by sample program. Now let us take a closer look at the output.

Graph DB Data Explorer.jpg

Closer look: A quick look shows that the query g.V() returned 4 nodes. Result can be viewed in either JSON or Graph mode.

Output - closer look 1

If we click on the second node – Ben – and zoom in, we get a nice graphical view of ben’s relationships. Click and explore to view related nodes.

Output - closer look 3

Switch to JSON mode to view result in JSON format. Pretty self-explanatory there ??

Output - closer look 2

Step 6: Monitor activity
Head over to Activity Log blade to take a look at activity in the Cosmos DB account. This could be downloaded as a .csv or exported to Event Hub for further analysis.

Activity Log


We took a look at the Cosmos DB Trial offered by Microsoft and got started with Graph API. Graph DB has a number of attractive use cases such as Fraud Detection and Recommendation Engine. Another popular Graph DB is Neo4j. I hope this article helped you understand how powerful Cosmos DB’s multi model support is. If you have any questions, drop me an email or add a comment. I will be happy to help. Add me on LinkedIn to stay connected – https://www.linkedin.com/in/arjunsivadasan/

Further Reading

  1. Introduction to Azure Cosmos DB and Azure Cosmos DB – Partition and Throughput
  2. Intro to Cassandra API in Cosmos DB – https://docs.microsoft.com/en-us/azure/cosmos-db/cassandra-introduction
  3. Intro to Graph DB – https://neo4j.com/graph-databases-book/
  4. Explore Gremlin API – https://aka.ms/gremlin
  5. Cosmos DB playground – https://www.documentdb.com/sql/demo
  6. Microsoft Ignite 2018 Updates – https://news.microsoft.com/uploads/prod/sites/507/2018/09/IGNITEBOOKOFNEWS-5ba95469d658b.pdf

SQL Roadie

Hi, I'm Arjun. I'm a Database programmer with over 10 years of experience designing and optimizing OLTP / transaction database systems. Of late, I have been focusing on Data Analytics. I presently work in the Data Analytics team at Oaks Hotels, Brisbane. Most of my work has been on SQL Server, but I have also worked on NoSQL databases such as CosmosDB and MongoDB, and Oracle. I am quite excited about the new Azure suite of products. My goal is to help organizations design scalable data frameworks and take decisions based on actionable data insights.


Leave a comment on the original post [sqlroadie.wordpress.com, opens in a new window]

Loading comments...