Quickstart

This document contains basic instructions to get started with your Rockset account. Using the Rockset console, you will learn to:

  1. Onboard a public dataset
  2. Create a collection
  3. Run queries against your collection

If you have already setup an API Key via the console, you can learn to use the rock CLI, or the Python SDK with Rockset. You can also start using Jupyter notebooks to work with your data in Rockset.

If you do not already have an account, please contact us at hello@rockset.com, or request access.

Setup access to Rockset

You will recieve an invite to use Rockset via email. This email will contain a link that allows you to activate your Rockset account. You can setup a username and password, or use your GSuite identity with Rockset. Once you have logged in, you will see your Rockset dashboard that looks like the following:

Console Starting View

You will notice that a collection named "_events" is automatically created for you. This is a system collection that keeps track of audit logs and some metadata. You can learn more about it under Troubleshooting.

Create a collection

In this section, we will create a collection from a public dataset that is hosted on AWS S3. This dataset is all the cities that exist in the world with a population of at least 1000, provided by https://www.geonames.org/. Click on Create Collection in the Overview tab to begin.

Choose an appropriate name, description (optionally), and choose "No Integration" as we'll be using a public AWS S3 bucket as our data source. The full S3 path is s3://rockset-public-datasets/cities. Plug that into the S3 path input and you're ready to go.

Create New Collection

You should now see a new collection in state Created. It can take up to a minute for the collection to become Ready.

Ready Collection

Once the collection is Ready, you can click on it and enter the collection details view. In the collection details view, the Summary tab will show you details about the schema that has been deduced for the collection, as well as some additional information about the occurrences and distribution of your data. You can also inspect some sample documents in your collection using the List view.

Explore Collection

Note that documents are still being ingested in the background and the count of documents will refresh as that happens. Even as this occurs, you can run queries against the collection.

Query the collection

Click on the Query tab from the left hand sidebar. If you wanted to find the 100 most populous cities in the United States, you can run the following.

SELECT cities.fields.name,
       cities.fields.population
FROM cities
WHERE cities.fields.country = 'United States'
ORDER BY cities.fields.population DESC
LIMIT 100

Query Collection

You can also run more complex queries against the collection. For example, to find the top-10 most popular city names across all countries.

SELECT name as city_name,
       count(name) AS nation_count
FROM
  ( SELECT cities.fields.name AS name,
           cities.fields.country AS country
   FROM cities
   GROUP BY cities.fields.name,
            cities.fields.country) a
GROUP BY name
ORDER BY nation_count DESC
LIMIT 10

Look at the supported SQL commands and SQL functions to craft more complex queries against the new Rockset collection that you just created. Note that for your own collections, you may need to first set up an Integration to provide the requisite credentials, as well as configure permissions on your AWS S3 bucket prior to creating a collection.

Invite other users

You can invite other members of your team to your Rockset account through the console. To invite new members, navigate to the Manage > Users section on the top-bar.

Invite Users

Next Steps

Learn more about Rockset's design and benefits here. Also check out our guides for common workflows:

Learn about using Rockset through our APIs and clients:

Please reach out at hello@rockset.com with any questions or feedback, we would love to talk. Rock on!

← ConceptsUse Cases →