FAQ

Can I query/join multiple tables at one time?

Not directly no. A Superset SQLAlchemy datasource can only be a single table or a view.

When working with tables, the solution would be to materialize a table that contains all the fields needed for your analysis, most likely through some scheduled batch process.

A view is a simple logical layer that abstract an arbitrary SQL queries as a virtual table. This can allow you to join and union multiple tables, and to apply some transformation using arbitrary SQL expressions. The limitation there is your database performance as Superset effectively will run a query on top of your query (view). A good practice may be to limit yourself to joining your main large table to one or many small tables only, and avoid using GROUP BY where possible as Superset will do its own GROUP BY and doing the work twice might slow down performance.

Whether you use a table or a view, the important factor is whether your database is fast enough to serve it in an interactive fashion to provide a good user experience in Superset.

How BIG can my data source be?

It can be gigantic! As mentioned above, the main criteria is whether your database can execute queries and return results in a time frame that is acceptable to your users. Many distributed databases out there can execute queries that scan through terabytes in an interactive fashion.

How do I create my own visualization?

We are planning on making it easier to add new visualizations to the framework, in the meantime, we’ve tagged a few pull requests as example to give people examples of how to contribute new visualizations.

https://github.com/airbnb/superset/issues?q=label%3Aexample+is%3Aclosed

Why are my queries timing out?

There are many reasons may cause long query timing out.

  • For running long query from Sql Lab, by default Superset allows it run as long as 6 hours before it being killed by celery. If you want to increase the time for running query, you can specify the timeout in configuration. For example:

    SQLLAB_ASYNC_TIME_LIMIT_SEC = 60 * 60 * 6

  • Superset is running on gunicorn web server, which may time out web requests. If you want to increase the default (50), you can specify the timeout when starting the web server with the -t flag, which is expressed in seconds.

    superset runserver -t 300

  • If you are seeing timeouts (504 Gateway Time-out) when loading dashboard or explore slice, you are probably behind gateway or proxy server (such as Nginx). If it did not receive a timely response from Superset server (which is processing long queries), these web servers will send 504 status code to clients directly. Superset has a client-side timeout limit to address this issue. If query didn’t come back within clint-side timeout (45 seconds by default), Superset will display warning message to avoid gateway timeout message. If you have a longer gateway timeout limit, you can change client-side timeout limit settings from /superset/superset/assets/javascripts/constants.js file and rebuild js package:

    export const QUERY_TIMEOUT_THRESHOLD = 45000;

Why is the map not visible in the mapbox visualization?

You need to register to mapbox.com, get an API key and configure it as MAPBOX_API_KEY in superset_config.py.

How to add dynamic filters to a dashboard?

It’s easy: use the Filter Box widget, build a slice, and add it to your dashboard.

The Filter Box widget allows you to define a query to populate dropdowns that can be use for filtering. To build the list of distinct values, we run a query, and sort the result by the metric you provide, sorting descending.

The widget also has a checkbox Date Filter, which enables time filtering capabilities to your dashboard. After checking the box and refreshing, you’ll see a from and a to dropdown show up.

By default, the filtering will be applied to all the slices that are built on top of a datasource that shares the column name that the filter is based on. It’s also a requirement for that column to be checked as “filterable” in the column tab of the table editor.

But what about if you don’t want certain widgets to get filtered on your dashboard? You can do that by editing your dashboard, and in the form, edit the JSON Metadata field, more specifically the filter_immune_slices key, that receives an array of sliceIds that should never be affected by any dashboard level filtering.

..code:

{
    "filter_immune_slices": [324, 65, 92],
    "expanded_slices": {},
    "filter_immune_slice_fields": {
        "177": ["country_name", "__from", "__to"],
        "32": ["__from", "__to"]
    },
    "timed_refresh_immune_slices": [324]
}

In the json blob above, slices 324, 65 and 92 won’t be affected by any dashboard level filtering.

Now note the filter_immune_slice_fields key. This one allows you to be more specific and define for a specific slice_id, which filter fields should be disregarded.

Note the use of the __from and __to keywords, those are reserved for dealing with the time boundary filtering mentioned above.

But what happens with filtering when dealing with slices coming from different tables or databases? If the column name is shared, the filter will be applied, it’s as simple as that.

How to limit the timed refresh on a dashboard?

By default, the dashboard timed refresh feature allows you to automatically requery every slice on a dashboard according to a set schedule. Sometimes, however, you won’t want all of the slices to be refreshed - especially if some data is slow moving, or run heavy queries. To exclude specific slices from the timed refresh process, add the timed_refresh_immune_slices key to the dashboard JSON Metadata field:

..code:

{
   "filter_immune_slices": [],
    "expanded_slices": {},
    "filter_immune_slice_fields": {},
    "timed_refresh_immune_slices": [324]
}

In the example above, if a timed refresh is set for the dashboard, then every slice except 324 will be automatically requeried on schedule.

Why does fabmanager or superset freezed/hung/not responding when started (my home directory is NFS mounted)?

By default, superset creates and uses an sqlite database at ~/.superset/superset.db. Sqlite is known to don’t work well if used on NFS due to broken file locking implementation on NFS.

You can override this path using the SUPERSET_HOME environment variable.

Another work around is to change where superset stores the sqlite database by adding SQLALCHEMY_DATABASE_URI = 'sqlite:////new/location/superset.db' in superset_config.py (create the file if needed), then adding the directory where superset_config.py lives to PYTHONPATH environment variable (e.g. export PYTHONPATH=/opt/logs/sandbox/airbnb/).

How do I add new columns to an existing table

Table schemas evolve, and Superset needs to reflect that. It’s pretty common in the life cycle of a dashboard to want to add a new dimension or metric. To get Superset to discover your new columns, all you have to do is to go to Menu -> Sources -> Tables, click the edit icon next to the table who’s schema has changed, and hit Save from the Detail tab. Behind the scene, the new columns will get merged it. Following this, you may want to re-edit the table afterwards to configure the Column tab, check the appropriate boxes and save again.

How do I go about developing a new visualization type?

Here’s an example as a Github PR with comments that describe what the different sections of the code do: https://github.com/airbnb/superset/pull/3013

What database engine can I use as a backend for Superset?

To clarify, the database backend is an OLTP database used by Superset to store its internal information like your list of users, slices and dashboard definitions.

Superset is tested using Mysql, Postgresql and Sqlite for its backend. It’s recommended you install Superset on one of these database server for production.

Using a column-store, non-OLTP databases like Vertica, Redshift or Presto as a database backend simply won’t work as these databases are not designed for this type of workload. Installation on Oracle, Microsoft SQL Server, or other OLTP databases may work but isn’t tested.

Please note that pretty much any databases that have a SqlAlchemy integration should work perfectly fine as a datasource for Superset, just not as the OLTP backend.