Getting Started with RedisGraph

RedisGraph is a Redis module that enables enterprises to process any kind of connected data much faster than with traditional relational or existing graph databases. RedisGraph implements a unique data storage and processing solution (with sparse-adjacency matrices and GraphBLAS) to deliver the fastest and most efficient way to store, manage, and process connected data in graphs. With RedisGraph, you can process complex transactions 10 - 600 times faster than with traditional graph solutions while using 50 - 60% less memory resources than other graph databases!

Step 1. Register and subscribe#

Follow this link to register and subscribe to Redis Enterprise Cloud

RedisGraph RedisGraph

Step 2. Create a database with RedisGraph Module#

RedisGraph

Step 3. Connect to Redis database#

Follow this link to learn about RedisInsight - a free Redis GUI to connect to Redis database.

Step 4. Getting Started with RedisGraph#

In the following steps, we will use some basic RediGraph commands to insert data into a graph and then query the graph. You can run them from the Redis command-line interface (redis-cli) or use the CLI available in RedisInsight. (See part 2 of this tutorial to learn more about using the RedisInsight CLI.)

RedisGraph

Step 5: Insert data into a graph#

Insert actors#

To interact with RedisGraph you will typically use the GRAPH.QUERY command and execute Cypher queries. Let’s start to insert some actors into the graph:movies graph name, which is automatically created using this command:

>> GRAPH.QUERY graph:movies "CREATE (:Actor {name:'Mark Hamill', actor_id:1}), (:Actor {name:'Harrison Ford', actor_id:2}), (:Actor {name:'Carrie Fisher', actor_id:3})"
1) 1) "Labels added: 1"
2) "Nodes created: 3"
3) "Properties set: 6"
4) "Query internal execution time: 0.675400 milliseconds"

This single query creates three actors, along with their names and unique IDs.

Insert a movie#

> GRAPH.QUERY graph:movies "CREATE (:Movie {title:'Star Wars: Episode V - The Empire Strikes Back', release_year: 1980 , movie_id:1})"
1) 1) "Labels added: 1"
2) "Nodes created: 1"
3) "Properties set: 3"
4) "Query internal execution time: 0.392300 milliseconds"

This single query creates a movie with a title, the release year, and an ID.

Associate actors and movies#

The core of a graph is the relationships between the nodes, allowing the applications to navigate and query them. Let’s create a relationship between the actors and the movies:

> GRAPH.QUERY graph:movies "MATCH (a:Actor),(m:Movie) WHERE a.actor_id = 1 AND m.movie_id = 1 CREATE (a)-[r:Acted_in {role:'Luke Skywalker'}]->(m) RETURN r"
1) 1) "r"
2) 1) 1) 1) 1) "id"
2) (integer) 1
2) 1) "type"
2) "Acted_in"
3) 1) "src_node"
2) (integer) 0
4) 1) "dest_node"
2) (integer) 3
5) 1) "properties"
2) 1) 1) "role"
2) "Luke Skywalker"
3) 1) "Properties set: 1"
2) "Relationships created: 1"
3) "Query internal execution time: 0.664800 milliseconds"

This command created a new relation indicating that the actor Mark Hamill acted in Star Wars: Episode V as Luke Skywalker.

Let’s repeat this process for the other actors:

> GRAPH.QUERY graph:movies "MATCH (a:Actor), (m:Movie) WHERE a.actor_id = 2 AND m.movie_id = 1 CREATE (a)-[r:Acted_in {role:'Han Solo'}]->(m) RETURN r"
> GRAPH.QUERY graph:movies "MATCH (a:Actor), (m:Movie) WHERE a.actor_id = 3 AND m.movie_id = 1 CREATE (a)-[r:Acted_in {role:'Princess Leila'}]->(m) RETURN r"

You can also do all of this in a single query, for example:

> GRAPH.QUERY graph:movies "CREATE (:Actor {name:'Marlo Brando', actor_id:4})-[:Acted_in {role:'Don Vito Corleone'}]->(:Movie {title:'The Godfather', release_year: 1972 , movie_id:2})"
1) 1) "Nodes created: 2"
2) "Properties set: 6"
3) "Relationships created: 1"
4) "Query internal execution time: 0.848500 milliseconds"

Querying the graph#

Now that you have data in your graph, you’re ready to ask some questions, such as:

“What are the titles of all the movies?”#

> GRAPH.QUERY graph:movies "MATCH (m:Movie) RETURN m.title"
1) 1) "m.title"
2) 1) 1) "Star Wars: Episode V - The Empire Strikes Back"
2) 1) "The Godfather"
3) 1) "Query internal execution time: 0.349400 milliseconds"

“What is the information for the movie with the ID of 1?”#

> GRAPH.QUERY graph:movies "MATCH (m:Movie) WHERE m.movie_id = 1 RETURN m"
1) 1) "m"
2) 1) 1) 1) 1) "id"
2) (integer) 3
2) 1) "labels"
2) 1) "Movie"
3) 1) "properties"
2) 1) 1) "title"
2) "Star Wars: Episode V - The Empire Strikes Back"
2) 1) "release_year"
2) (integer) 1980
3) 1) "movie_id"
2) (integer) 1
3) 1) "Query internal execution time: 0.365800 milliseconds"

“Who are the actors in the movie 'Star Wars: Episode V - The Empire Strikes Back' and what roles did they play?”#

> GRAPH.QUERY graph:movies "MATCH (a:Actor)-[r:Acted_in]-(m:Movie) WHERE m.movie_id = 1 RETURN a.name,m.title,r.role"
1) 1) "a.name"
2) "m.title"
3) "r.role"
2) 1) 1) "Mark Hamill"
2) "Star Wars: Episode V - The Empire Strikes Back"
3) "Luke Skywalker"
2) 1) "Harrison Ford"
2) "Star Wars: Episode V - The Empire Strikes Back"
3) "Han Solo"
3) 1) "Carrie Fisher"
2) "Star Wars: Episode V - The Empire Strikes Back"
3) "Princess Leila"
3) 1) "Query internal execution time: 0.641200 milliseconds"

Visualizing graph databases using RedisInsight#

If you are using RedisInsight, you can visualize and navigate into the nodes and relationships graphically. Click on the RedisGraph menu entry on the left and enter the query:

MATCH (m:Actor) return m

Click on the Execute button, and double click on the actors to follow the relationships You should see a graph like this one:

RedisGraph

Resources#

Next Step#