You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We have not considered users who are using the AWS Batch runtime but not the Docker runtime, but it's definitely possible!
I think the main point to consider here is how to manage the image used for AWS Batch. Maybe we should consider adding a nextstrain update aws-batch command that can update the docker image version? Maybe add image to the AWS Batch config section? I think either would add more complexity to #167...
The text was updated successfully, but these errors were encountered:
Good to consider this. I think the main decision point is if the AWS Batch runtime image should stay tightly coupled to the Docker runtime image.
If it should, then nextstrain update aws-batch (and nextstrain setup aws-batch) could touch the docker.image config var. But this might be somewhat unintuitive and surprising, as other runtimes are independent of each other.
It would be more consistent to decouple the AWS Batch and Docker runtime images so they can be setup/updated separately. But this would also be a behavioural change which itself may surprise folks who expect them (consciously or not) to be synced.
Steph from WA DOH recently tripped over this. She uses our Conda runtime locally but also our AWS Batch runtime and had previously used our Docker runtime locally so had an old image version in config.
We have not considered users who are using the AWS Batch runtime but not the Docker runtime, but it's definitely possible!
I think the main point to consider here is how to manage the image used for AWS Batch. Maybe we should consider adding a
nextstrain update aws-batch
command that can update the docker image version? Maybe add image to the AWS Batch config section? I think either would add more complexity to #167...The text was updated successfully, but these errors were encountered: