-
-
Notifications
You must be signed in to change notification settings - Fork 144
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
LSF scheduler seems to have redundant memory keyword #205
Comments
Also it appears that memory should be of form '24 GB', but mem need to be passed as an integer number of bytes. |
Historically, the memory keyword was for the We discussed leaving or not the duplicated ncpus or mem kwarg in #82 or more recently in #172 (comment), and maybe in #181 too. I'm still undecided on this, especially for the memory kwarg. It seems unlikely that a user wants to have a different reservation that what it uses on dask_side. And in this case, maybe he could implement it with |
Okay, I don't have a strong preference either way except that I got a little confused. Also I assume we would want the input form for the two to match (Using unit in the name as a string, vs passing an int for memory). To update this I assume we'd want to test the type for an int or str on mem keyword and do the conversion that way so as to not break previously existing scripts? Or would you rather it is just updated for both to take a str? |
Hm, I think it would be a minor breaking change and not impact a lot of people. I'd be more in favor of just update |
I think we can do two things to clarify this:
|
It looks like specifying the lsf scheduler requires two memory keywords (mem, and memory). Should mem on the LSF scheduler be turned into memory so that it lines up with other schedulers and the core object?
The text was updated successfully, but these errors were encountered: