Quickstart

This quickstart walks you through:

  1. Setting up a TigerGraph schema and loading data to a TigerGraph instance.

  2. Running an instance of TigerGraph GraphQL service and connecting to the graph in the TigerGraph instance.

  3. Running GraphQL queries against the service.

  4. (Optional) Running a web application built on top of the service

You can view the walk-through in video format:

1. Download example solution and data

  1. Download the example TigerGraph solution.

  2. Download the example data set and unzip it.

2. Load data

  1. Start up a TigerGraph instance, either locally or on TigerGraph Cloud.

  2. Navigate to the GraphStudio homepage and click Import an Existing Solution. Select the .tar.gz file you downloaded in the previous step/.

  3. In the upper left corner of the GraphStudio Home page, click Global View  DemoGraph to switch to the demo graph.

  4. On the left-side navigation, click Map Data To Graph, which should no longer be grayed out after switching out of global view.

  5. Click the Add Data File button and upload all the .csv files from the decompressed data folder.

    Add data file
    Figure 2. Add data file

  6. Navigate to the Load Data page by clicking Load Data on the left-side navigation menu. Click the ▶ button in the toolbar above the canvas to start data loading.

After data loading completes, your graph is ready for use.

3. Start TigerGraph GraphQL Service

  1. Download the service binary:

    OS Download link

    Linux X86_64

    https://tigergraph-release-download.s3.us-west-1.amazonaws.com/tigergraphql/tigergraph-graphql-0.7.1-linux-x86_64.tar.gz

    MacOS x64

    https://tigergraph-release-download.s3.us-west-1.amazonaws.com/tigergraphql/tigergraph-graphql-0.7.1-macos-x64.tar.gz

    Windows x64

    https://tigergraph-release-download.s3.us-west-1.amazonaws.com/tigergraphql/tigergraph-graphql-0.7.1-windows-x64.zip

  2. Start the service

    • Linux/MacOS

    • Windows

    1. Decompress the tarball, and there will be a binary named tigergraphql.

    2. Run the following command from the terminal to start the service and connect to the TigerGraph instance:

      $ ./tigergraphql -s http://<url_to_tigergraph:port> -p 4000 -u tigergraph -w tigergraph -g DemoGraph (1) (2) (3) (4) (5)
      1 -s: The TigerGraph server URL
      2 -p: The port TigerGraph GraphQL Service is served on
      3 -u: GSQL username
      4 -w: GSQL password
      5 -g: The name of the graph to connect to

    1. Unzip the file, and there will be a file named tigergraphql.exe.

    2. Run the executable with the following command to start the service and connect to the TigerGraph instance:

      tigergraphql.exe -s http://<url_to_tigergraph:port> -p 4000 -u tigergraph -w tigergraph -g DemoGraph (1) (2) (3) (4) (5)
      1 -s: The TigerGraph server URL and port for GraphStudio
      2 -p: The port TigerGraph GraphQL Service is served on
      3 -u: GSQL username
      4 -w: GSQL password
      5 -g: The name of the graph to connect to

    The URL and port is the URL and port to access GraphStudio in the browser. If you are accessing a TigerGraph Cloud instance, you will need to use https instead of http, and you do not need to specify the port.

If the connection is successful, you will see in the terminal that the schema is loaded and that the service is being served. Now you can access the GraphQL API from the browser at the URL and port you specified:

graphql browser
Figure 3. GraphQL API in browser

4. Run a query

Having set up the GraphQL service, now let’s try running a simple query against the API. You’ll learn about running GraphQL queries on TigerGraph in more depth in the Queries section. For now, let’s run a simple query to get all person vertices in the graph and where they are born:

  • Query

  • Results

query {
  DemoGraph {
    person {
      name
      born_in {
        birthday
        to {
          name
        }
      }
    }
  }
}
{
  "data": {
    "DemoGraph": {
      "person": [
        {
          "born_in": [
            {
              "birthday": "1982-07-27 00:00:00",
              "to": {
                "name": "redwood city"
              }
            }
          ],
          "name": "John"
        },
        {
          "born_in": [
            {
              "birthday": "1991-12-21 00:00:00",
              "to": {
                "name": "san jose"
              }
            }
          ],
          "name": "Kevin"
        },
        {
          "born_in": [
            {
              "birthday": "1995-01-01 00:00:00",
              "to": {
                "name": "palo alto"
              }
            }
          ],
          "name": "Jenny"
        },
        {
          "born_in": [
            {
              "birthday": "1973-10-05 00:00:00",
              "to": {
                "name": "mountain view"
              }
            }
          ],
          "name": "Smith"
        },
        {
          "born_in": [
            {
              "birthday": "1990-09-12 00:00:00",
              "to": {
                "name": "redwood city"
              }
            }
          ],
          "name": "Tom"
        },
        {
          "born_in": [
            {
              "birthday": "1992-05-23 00:00:00",
              "to": {
                "name": "san jose"
              }
            }
          ],
          "name": "Emily"
        }
      ]
    }
  },
  "errors": null
}

5. (Optional) Run a GraphQL-powered web application

You can run our example React app built on top our demo data and explore how a front-end application interacts with a GraphQL API.

  1. Download the example application, unzip the file and change directory into the tigergraphql_example_react folder

  2. If TigerGraph GraphQL Service is running on another machine, change config.json with the IP and port to that machine. If the service is running on localhost, skip this step.

    {
    "GRAPHQL_API": "http://localhost:4000/graphql" (1)
}
    1 Change the root url to the IP and port that the service is running on

  3. Run the following command to serve the app:

    $ python server.py
serving at port 3000