Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

Hello, I'm going to deploy my workflow on the local container. But the workflow fails with this message:
```[1/1] currentAttempt done. Last Error: USER::containers with unready status: [fba452891990446a7a49-n0-0]|Back-off pulling image "cosine:2ce90d3ad559b4e5b1981af8726f4d19eeedc835"```
Is there any way I can use to debug what's happened?

It seems you did not push the container to a registry

So one option when working with demo / sandbox is to build the container into demo 

This is done by mounting the source inside - handled automatically by flytectl

<https://docs.flyte.org/projects/cookbook/en/latest/auto/larger_apps/larger_apps_deploy.html|https://docs.flyte.org/projects/cookbook/en/latest/auto/larger_apps/larger_apps_deploy.html>

But I think in a doc restructure we missed this

Custom images. Cc <@U03BGLXSWMT> <@U02MEAKMMU1> 

I've built the image using `docker_build_and_tag.sh`, and packed the workflow using `pyflyte --pkgs flyte.workflows package --image &lt;registry/repo:version&gt;`. But I don't know how to run my workflows using this image, executing `flytectl create execution --project flytesnacks --domain development --execFile exec_spec.yaml`  results in the same container with unready status error

I finally managed to deploy my workflow to local demo container. I noticed that when the task passes a type that flyte doesn't support, the object is converted using `PythonPickle` , and attached with a s3 url. Does it mean the dumped file is uploaded to aws?

In local case to the bundled minio - which has an s3 like interface 

But you are right in prod it will be one of the blob stores

Yeah, we'll use our private data storage service. I found this doc: <https://docs.flyte.org/projects/flytekit/en/latest/data.extend.html> but it's somewhat.. simple. I still don't know exactly how to extend flyte to use my custom data storage

<@U03GZAV819N> let me dm you I can definitely help with this. Can we do a quick VC call and we can discuss 

It seems that we have jet lag here.. I'm going to have some meetings in next few hours. Let me have a try before we talk :D