-
Notifications
You must be signed in to change notification settings - Fork 202
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
limit maxparallel
to 16 by default
#4606
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
Would be nice to motivate this a bit better, for example by benchmarking the speedups we see with a couple of large projects on systems with say 128 cores, for example GCC, OpenFOAM, Qt5, etc. 16 is not an unreasonable value to use as default, but that's based on gut feeling. |
Do we have a template that could be used to auto-set For the examples given, this could also just be done in the easyblocks. |
I do not have a 128 core node available at the moment, but using a 112 core one: Node:
For comparison, I've also tested using
|
I'll start some jobs for the same software that @branfosj is doing, plus Node:
|
@branfosj interesting that your |
There is something weird about building Qt5 in that development environment of mine that causes it to take ages. However, I can also get Qt5 build much more quickly. Our live build time, using 24 cores, for |
wild, that's a huge difference. In any case, 16-32 cores looks like the sweet spot on our system |
We often see it where the large numbers of cores available on modern systems causes either the build to progress slowly or to fail. I propose we add this default to
maxparallel
to counter that.