Skip to content

Commit

Permalink
Update scheduling.md
Browse files Browse the repository at this point in the history
Co-authored-by: Jason Deal <[email protected]>
  • Loading branch information
bryantbiggs and jmdeal authored Oct 31, 2024
1 parent 309c27d commit 30e265d
Showing 1 changed file with 1 addition and 1 deletion.
2 changes: 1 addition & 1 deletion website/content/en/preview/concepts/scheduling.md
Original file line number Diff line number Diff line change
Expand Up @@ -104,7 +104,7 @@ The [Neuron scheduler extension](https://awsdocs-neuron.readthedocs-hosted.com/e

> The Neuron scheduler extension finds sets of directly connected devices with minimal communication latency when scheduling containers. On Inf1 and Inf2 instance types where Neuron devices are connected through a ring topology, the scheduler finds sets of contiguous devices. For example, for a container requesting 3 Neuron devices the scheduler might assign Neuron devices 0,1,2 to the container if they are available but never devices 0,2,4 because those devices are not directly connected. On Trn1.32xlarge and Trn1n.32xlarge instance types where devices are connected through a 2D torus topology, the Neuron scheduler enforces additional constraints that containers request 1, 4, 8, or all 16 devices. If your container requires a different number of devices, such as 2 or 5, we recommend that you use an Inf2 instance instead of Trn1 to benefit from more advanced topology.

However, Karpenter is not aware of the decisions made by the Neuron scheduler extension which precludes it from making any optimizations to consolidate and bin pack pods requiring Neuron resources.
However, Karpenter is not aware of the decisions made by the Neuron scheduler extension which precludes it from making any optimizations to consolidate and bin pack pods requiring Neuron resources. To ensure Karpenter's bin-packing is consistent with the decisions made by the scheduler extension, containers must have like-sized, power of 2 requests (e.g. 1, 2, 4, etc). Failing to do so may result in permanently pending pods.

### Pod ENI Resources (Security Groups for Pods)
[Pod ENI](https://github.com/aws/amazon-vpc-cni-k8s#enable_pod_eni-v170) is a feature of the AWS VPC CNI Plugin which allows an Elastic Network Interface (ENI) to be allocated directly to a Pod. When enabled, the `vpc.amazonaws.com/pod-eni` extended resource is added to supported nodes. The Pod ENI feature can be used independently, but is most often used in conjunction with Security Groups for Pods. Follow the below instructions to enable support for Pod ENI and/or Security Groups for Pods in Karpenter.
Expand Down

0 comments on commit 30e265d

Please sign in to comment.