Hi, we are deploying another instance of metaflow/argo on a different EKS cluster. How do you work with the metaflow

config.json

? Do we have to manually set it as some env var so that it can use the correct cluster?

Hi Shantanu, if I’m understanding the use case correctly - you have two clusters you’re using, and you want to be able to pick which cluster your flow should be submitted to? In that case: 1.

config.json

can point to your primary/default cluster. 2. Have another config corresponding to your second cluster - let’s call it

config_secondary.json

3. Then you can use

METAFLOW_PROFILE=secondary python flow.py run

and this will use the cluster pointed to by

config_secondary.json