Explore option to make webhook fail open rather than fail close

This can be made configurable or kept as fail close.

**What would you like to be enhanced**:

**Why is the change needed and what use case will it solve**:
In the event of controller being restarted, webhook will be unavailable, If pod is deployed during that time. Webhook will fail open causing pod to schedule with generic ENI rather than branch ENI and security group.

Start by locating the webhook implementation and the controller restart or pod admission path. Determine how webhook unavailability is handled during a controller restart and clarify whether the desired behavior is configurable or should remain fail-closed. Done means the chosen behavior and its security implications are documented and covered by appropriate tests.

Written by the indexing model from the issue text.