---
title: 'Introducing Cosmos 1.6: The Best Way to Run dbt-core in Airflow'
description: >-
We're happy to announce that the newest version of Astronomer’s dbt-core
integration, Cosmos 1.6.0, is now available, featuring a list of enhancements
and some great new additions to serve the community.
date: 2024-09-27T00:00:00.000Z
authors:
- author: src/content/people/pankaj-singh.md
tags: []
categories: []
canonical_url: 'https://www.astronomer.io/blog/introducing-cosmos-1-6/'
---
Cosmos 1.6.0, is now available, featuring a range of enhancements and new additions to serve the community. This blog post will walk you through the features introduced in Cosmos 1.6, released on August 20th.
**Details:**
* 📦 PyPI:
[https://pypi.org/project/astronomer-cosmos/1.6.0/](https://pypi.org/project/astronomer-cosmos/1.6.0/)
* 📚 Docs:
[https://astronomer.github.io/astronomer-cosmos/getting\_started/index.html](https://astronomer.github.io/astronomer-cosmos/getting_started/index.html)
* 🛠 Release Notes:
[https://github.com/astronomer/astronomer-cosmos/blob/main/CHANGELOG.rst#160-2024-08-20](https://github.com/astronomer/astronomer-cosmos/blob/main/CHANGELOG.rst#160-2024-08-20)
## When to use Cosmos
The open-source provider package Cosmos allows you to [integrate dbt-core
jobs](https://www.astronomer.io/solutions/dbt/)
into Airflow by automatically creating Airflow tasks from dbt models,
giving you the ability to turn your dbt-core projects into an Airflow task
group and DAGs with just a few lines of code. Since its release in early
2023, Cosmos has become the most popular way to run dbt-core with Airflow, recently passing 1 million monthly downloads.
This blog post dives into the newest Cosmos features released in version 1.6.0. If you are new to Cosmos, check out our [tutorial](https://www.astronomer.io/docs/learn/airflow-dbt/?utm_source=cosmos-1-6-blog&utm_medium=blog&utm_campaign=dbt) and the [documentation](https://astronomer.github.io/astronomer-cosmos/getting_started/index.html) for more information.
## Airflow Task Representation for dbt Source Nodes
A new way of rendering [source nodes as Airflow
tasks](https://astronomer.github.io/astronomer-cosmos/configuration/source-nodes-rendering.html)
is now available through the RenderConfig API. The default behavior is set
to not render source nodes. You can control this rendering behavior using
the source\_rendering\_behavior parameter in RenderConfig.
With this addition, Cosmos now supports running freshness checks on source
nodes, executing source tests, and rendering source nodes as
EmptyOperators in the Airflow UI for nodes without tests or freshness
checks.
This feature was heavily requested across various channels by community
members and developed by [Alejandro Rojas
Benítez](https://github.com/arojasb3), a data engineer at
[GlossGenius](https://glossgenius.com/). We thank all our users for their
continued suggestions and inspiration on how we can further improve
Cosmos!
### Example of how to render dbt Source Nodes
In the Airflow graph view above, you can see that source nodes are
rendered in sky blue on the left side.
For more information on rendering source nodes with Cosmos check out
the
[documentation](https://astronomer.github.io/astronomer-cosmos/configuration/source-nodes-rendering.html).
## Persisting the virtual env directory for LoadMode.VIRTUALENV
Additionally, Cosmos 1.6. added a new option in the ExecutionConfig API to
persist the virtual environment for the virtualenv load mode on a single
worker. This allows the worker to maintain and reuse the virtual
environment, enhancing DAG run time efficiency. Before this change, Cosmos
would create a new Python virtualenv every time, which could significantly
impact the task execution time.
## Support using remote manifest files (from GCS/S3/Azure Blob Storage)
Cosmos 1.6 better supports users who want to keep their dbt project(s) and
Airflow in different repositories and in different Docker container
images. Hosting manifest files in remote object storage in either GCS, S3
or Azure blob storage is compatible with implementations that use the
Cosmos manifest load method.
For more information, See [Load Cosmos dbt manifest file from
S3/GCS/ABS](https://astronomer.github.io/astronomer-cosmos/configuration/parsing-methods.html#dbt-manifest)
in the documentation.
### Example implementation of using a remote manifest file
## Support remote caching
In addition to being able to store manifests in remote object storage, we
have enhanced `dbt ls` output caching to enable storage in remote
systems such as S3, GCS, and Azure blob storage as well. This improvement
increases the scalability of dbt ls output caching and reduces the load on
the Airflow metadata database. This is an alternative to the Cosmos 1.5
Airflow Variable-based cache.
This feature utilizes the Airflow ObjectStorage feature, which is
dependent on Airflow version 2.8.0+.
You can enable remote caching for dbt ls by setting the following
environment variables in your Airflow environment. See the [Cosmos
documentation](https://astronomer.github.io/astronomer-cosmos/configuration/caching.html#caching-the-dbt-ls-output)
for more information.
## Caching of the dbt package
Cosmos 1.6 also adds support for [caching the dbt package-lock.yml
file](https://astronomer.github.io/astronomer-cosmos/configuration/cosmos-conf.html#enable-cache-package-lockfile)
to ensure predictable installation of dbt project dependencies and to
enhance the performance of [dbt
deps](https://docs.getdbt.com/reference/commands/deps) in Cosmos for LOCAL
and VIRTUALENV execution modes.
You can enable or disable package caching using the following environment
variable: `AIRFLOW__COSMOS__ENABLE_CACHE_PACKAGE_LOCKFILE=True`
## Teradata profile mapping
Lastly, this release added support for profile mapping to enable an
integration between the [Teradata Airflow
provider](https://airflow.apache.org/docs/apache-airflow-providers-teradata/stable/index.html)
and the [dbt-teradata adapter](https://github.com/Teradata/dbt-teradata).
This enhancement expands Cosmos' capabilities, allowing seamless
connections with Teradata Vantage and improved workflow management.
You can learn more about Teradata profile mapping in the [Cosmos
Docs](https://astronomer.github.io/astronomer-cosmos/profiles/TeradataUserPassword.html).
## Conclusion
The 1.6 release of Cosmos added many highly anticipated features. We’d
like to thank the Cosmos contributors making this release possible, in
particular [Alejandro Rojas Benítez](https://github.com/arojasb3), [Daniel
Reeves](https://github.com/dwreeves), [Giovanni Corsetti
Silva](https://github.com/CorsettiS),
[Jaegwon.seo](https://github.com/wornjs), [Micah
Victoria](https://github.com/mvictoria), [Pankaj
Koti](https://github.com/pankajkoti), [Pankaj
Singh](https://github.com/pankajastro), [Satish
Chinthanippu](https://github.com/sc250072) and [Tatiana
Al-Chueyr](https://github.com/tatiana).
By combining Cosmos and dbt Deploys in one platform, dbt on Astro enables
teams to observe and deploy their dbt and Airflow code from a single
place. Experience the latest firsthand with our [14-day Free
Trial](https://www.astronomer.io/lp/signup/)
(new signups receive $300 in credits to get started). Sign up today to get
started!