-
Notifications
You must be signed in to change notification settings - Fork 16
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
MIG GPU topic #313
base: main
Are you sure you want to change the base?
MIG GPU topic #313
Conversation
First-draft PR, for when we are ready to move forward. cc: @nwneisen |
1. Obtain the default MKE 4 configuration file: | ||
|
||
``` | ||
mkectl init |
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.
Redirect the output to actually save it to a file mkectl init > mke4.yaml
mkectl apply -f <mke-configuration-file> | ||
``` | ||
|
||
4. Optional. You can apply a profile to a node using the `nvidia.com/mig.config: |
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.
This is the part that is a little up in the air for this. I don't think the profiles are optional and when a node doesn't have one, MIG will just stay disabled on that node.
I'm also trying to think if there is a way for us to do this for them. For example, if they are in single mode and provide a profile in the mke config, we apply it to all their GPU nodes. But then we will likely run in to customers who don't want all their GPUs to be MIG. And then there's handling mixed mode..
No changes ATM but this is the part of MIG that is still in flux
No description provided.