Skip to content
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

Change default for max.load #102

Closed
krlmlr opened this issue Aug 31, 2015 · 1 comment
Closed

Change default for max.load #102

krlmlr opened this issue Aug 31, 2015 · 1 comment

Comments

@krlmlr
Copy link

krlmlr commented Aug 31, 2015

In makeClusterFunctionsMulticore, the argument max.load defaults to ncpus - 1. I think this should rather be detectCores() - 1 or detectCores() (not even related to the mc.cores option). The load should be limited by the available physical resources, not by what we allow the current process to use.

Example: Assume four cores, two of which are busy. I want to limit a computation to the two remaining cores and therefore say makeClusterFunctionsMulticore(2), to the effect that no workers are allowed to start because the load (2) is clearly larger than ncpus - 1.

On that note, what is the practical difference between ncpus and max.jobs?
#62 is slightly related.

@mllg
Copy link
Member

mllg commented Sep 30, 2015

Yep, that is a much better default. Note that we are well aware that the our choices are pretty defensive (as we want to avoid automatic over-utilization). We kind of expect the user to choose sane and efficient defaults.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants