You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Current device is scheduler pod to allocate,then write annotation;in device-plugin Allocate method Inject libvgpu.so, there are several problems with doing this:
every node have a node lock, so the concurrency of node allocation device is 1.
The current implementation method will cause confusion in pod resource allocation, and the set resources will be inconsistent with the actual ones.
requires frequent interaction with kube-apiserver
We can implement an NRI plugin in the device-plugin component and use NRI to inject devices, which can solve the three problems mentioned above.
What type of PR is this?
/kind feature
What this PR does / why we need it:
Which issue(s) this PR fixes:
Fixes #
Special notes for your reviewer:
Does this PR introduce a user-facing change?:
The text was updated successfully, but these errors were encountered:
What would you like to be added:
Current device is scheduler pod to allocate,then write annotation;in device-plugin
Allocate
method Injectlibvgpu.so
, there are several problems with doing this:kube-apiserver
We can implement an NRI plugin in the device-plugin component and use NRI to inject devices, which can solve the three problems mentioned above.
What type of PR is this?
/kind feature
What this PR does / why we need it:
Which issue(s) this PR fixes:
Fixes #
Special notes for your reviewer:
Does this PR introduce a user-facing change?:
The text was updated successfully, but these errors were encountered: