Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

Hi, I’m following the instructions in <https://towardsdatascience.com/how-to-cut-your-aws-ecs-costs-with-fargate-spot-and-prefect-1a1ba5d2e2df|this> great article on using ECS. I’m now setting up a custom docker image to run dbt (I’m assuming I can’t use `DbtShellTask` because dbt wouldn’t have been installed). It’s coming back with an error where I need to specify a `execution_role_arn` in the `ECSRun` which I’ve now done, but can I use the previously defined role, or do I need to define something else? Thanks!

You can just add the dbt library to the image. If you don’t want to make your own image, you can just use the base prefect image and add the env var `EXTRA_PIP_PACKAGES` which takes in packages to install. You can find that <https://docs.prefect.io/orchestration/flow_config/docker.html#installing-extra-dependencies-at-runtime|here> , only works if you use the base prefect image.

But could you show me the ECSRun code? Just remove sensitive info

<https://www.prefect.io/blog/orchestrating-elt-on-kubernetes-with-prefect-dbt-and-snowflake-part-2/|this post> seems to answer your question

the only difference is that this post is using EKS rather than ECS Fargate. If you use EKS on Fargate, you can follow the above article 1:1 with the only difference that the cluster could be created using the Fargate profile with `--fargate` flag:
```eksctl create cluster --fargate```

Fantastic! Thanks both. I actually managed to get it working using the initial post. I’d just skipped a step somehow. I also didn’t know about the `EXTRA_PIP_PACKAGES` env var, so that’s super helpful to know. Thanks