Skip to content
Branch: master
Find file History
natekupp Add config mapping context
Summary: This adds `execution_epoch_time` as a tag on the `RunConfig` object, and threads that through context creation to be added as a property on a `ConfigMappingContext` object which is provided to config mapping functions.

Test Plan: unit

Reviewers: schrockn, alangenfeld

Reviewed By: schrockn

Differential Revision:
Latest commit 7a968ed Jun 26, 2019



This library provides an integration with Dask / Dask.Distributed, to support distributed execution of Dagster workloads. It is still early, and has some limitations which are discussed below.

Presently, it provides a single API, execute_on_dask, which can execute a Dagster pipeline on either local Dask or a remote Dask cluster.


To use dagster-dask, you'll need to install Dask / Dask.Distributed.

If you want to use a cluster for distributed execution, you'll need to set up a Dask cluster. Note that you'll need to ensure that Dagster can access the host/port on which the Dask scheduler is running.

Getting Started

There is a simple example of how to use this library in the tests folder. This example showcases how to set up a "hello, world" with local Dask execution.

For distributed execution on a Dask cluster, you'll just need to provide a DaskConfig object with the address/port of the Dask scheduler:

    env_config={'storage': {'s3': {}}},


  • Presently, dagster-dask does not support launching Dask workloads from Dagit.
  • For distributed execution, you must use S3 for intermediates and run storage, as shown above.
  • Dagster logs are not yet retrieved from Dask workers; this will be addressed in follow-up work.

While this library is still nascent, we're working to improve it, and we are happy to accept contributions!

You can’t perform that action at this time.