hi folks, I wanted to check in regarding a bug i am facing and hoping someone could provide pointers: i am running some heavy metaflow tasks and require some additional disk space. I looked into using @batch(use_tmpfs=True) and shared_memory config. However, deploying it using

Step Functions

is resulting in

ValueError: invalid literal for int() with base 10: '127000.0'

errors originating from here and here even though i am passing the parameters as ints. The same code runs successfully when deployed using

AWS Batch

. Please let me know your thoughts and if I can provide more relevant details.

interesting. Sounds like a bug. Investigating..

Was able to verify and identify the cause for the issue. Working on a fix for this today. Issue for bug details and tracking: https://github.com/Netflix/metaflow/issues/1609

wow nice, thanks for the swift action. seems like the fix was merged? can give it a try once new version is released. lmk

I should be able to wrap a release today if all goes accordingly, still at least one thing that should get merged before that (unrelated to this issue) In the mean time, if you want to try out the feature, you can clone the git repository and

pip install .

from it

thank you for the update! please let me know when the release goes out

https://github.com/Netflix/metaflow/releases/tag/2.10.4 is out now and should fix the issue 🙂