"global.proxy.protocolDetectionTimeout" "meshConfig.protocolDetectionTimeout"
    "global.proxy.holdApplicationUntilProxyStarts" "meshConfig.defaultConfig.holdApplicationUntilProxyStarts"
    "pilot.ingress" "meshConfig.ingressService, meshConfig.ingressControllerMode, and meshConfig.ingressClass"
    "global.mtls.enabled" "the PeerAuthentication resource"
    "global.mtls.auto" "meshConfig.enableAutoMtls"

and allows a bit more flexibility.
- the dedicated pilot-per-ingress is required if the gateway needs to support k8s-style ingress.

# Port and basic host configuration

In order to configure the Service object, the install/upgrade needs to provide a list of all ports.
In the past, this was done when installing/upgrading full istio, and involved some duplication - ports configured

-- simplified configuration
-- multiple versions of the ingress can be used, to minize upgrade risks

- the new chart uses the default namespace service account, and doesn't require
additional RBAC permissions.

- simplified label structure. Label change is not supported on upgrade.

- for 'internal load balancer' you should deploy a separate gateway, in a different
namespace.

  $ helm get all {{ .Release.Name }} -n {{ .Release.Namespace }}

Next steps:
  * Deploy an HTTP Gateway: https://istio.io/latest/docs/tasks/traffic-management/ingress/ingress-control/