Updates to Helm Chart Package Structure Table

[nghtm]

A few comments on HyperPod EKS Helm_Chart Github package structure...

1/ EFA K8's Device Plugin is not included in the Helm Chart Package Structure Table, making it seem like it is not required for customers to install it. Is this intentional? Is the EFA K8's Device Plugin pre-installed somehow on HyperPod nodes?

2/ Requesting to add a "Required" column to specify if each chart is required in order to successfully create a HyperPod Cluster. Customers are asking for the bare minimum requirements needed to helm install in-order for HyperPod Resiliency features to work which should be listed in the Helm Chart Package Structure Table

[pintaoz-aws]

1/ aws-efa-k8s-device-plugin is required and enabled by default here. I have added it to the helm chart package structure table: #96.
2/ The only required helm chart for HyperPod Resiliency is health-monitoring-agent. As there are multiple features supported by the CLI, it will cause some confusion if we add one column for a specific feature. We may add a new section for the required charts of different features in the future.

[nghtm]

Thank you for the response @pintaoz-aws.

on 2/ Requesting to add a "Required" column for Helm Chart Package Structure Table. It is still important that we signal to customers helm charts are required for HyperPod CLI, Resiliency and Task Governance features. Customers frequently want to install only the required charts for their specific workload requirements. Installing additional charts runs the risk of breaking existing installations on customers clusters.

Can we add an additional column in the readme titled "Required for", and add the specific hyperpod features for which this helm chart must be added (ie CLI, Resiliency, Task Governance, etc).

[nghtm]

We should be following the design paradigm "batteries included, but swappable"

[nghtm]

@pintaoz-aws I just got off a call where it was stated that there are 4 helm charts required for HyperPod basic functionalities (resiliency):

• Health Moniotring Agent
• Deep Health Check
• Job Auto Restart
• HyperPod Patching