[Docs] Missing env setup in remote example

[soyking]

Description

In Running your workflow, commands for remote running missed env input, which cause remote error like:

File "/root/hello_world.py", line 19, in <module>
        registry=os.environ['FLYTE_IMAGE_REGISTRY']

This would work at my local machine:

FLYTE_IMAGE_REGISTRY=localhost:30000 pyflyte run --env FLYTE_IMAGE_REGISTRY=localhost:30000 --remote --project my-project --domain development hello_world.py hello_world_wf

Are you sure this issue hasn't been raised already?

• Yes

Have you read the Code of Conduct?

• Yes

[welcome]

Thank you for opening your first issue here! 🛠

[runllm]

The Flyte documentation does not explicitly show passing environment variables (like FLYTE_IMAGE_REGISTRY) when running workflows remotely with pyflyte, which can lead to errors if your code expects such variables. To ensure environment variables are available in remote executions, you should use the --env flag, e.g., pyflyte run --env FLYTE_IMAGE_REGISTRY=localhost:30000 --remote ... as you described. This is confirmed as the correct approach in recent Flyte issues and discussions, and omitting it can cause KeyError exceptions when accessing os.environ in your workflow code.

If you need to pass environment variables to tasks, use the --env flag with pyflyte run for both local and remote executions. Note that there was a bug in local executions with --env/--envvars, but it is fixed in recent versions; the --env flag works for remote runs as expected. See the linked issues for more details and workarounds if you encounter problems in older versions. Would you like a step-by-step example or more detail on environment variable handling in Flyte?

Sources:

• GitHub Issue #4503

---

This is an AI-generated response and your feedback is appreciated! Please leave a 👍 if this is helpful and 👎 if it is not.

Mention @RunLLM to ask follow-up questions. See runllm.com for more!

[0yukali0]

#take