[DOC] Clear and consise descriptions are needed for how to configure ServiceMesh 2.0 for HA

https://bugzilla.redhat.com/show_bug.cgi?id=1903932

Document URL:
https://docs.openshift.com/container-platform/4.6/service_mesh/v2x/ossm-custom-resources.html#ossm-cr-pilot_ossm-custom-resources-v2x
Section Number and Name:

Deploying to Production  - Istio Pilot configuration - Service Mesh 2.x | Service Mesh | OpenShift Container Platform 4.6

Describe the issue:

  Our customer faced the issue that they cannot configure Anti-Affinity for istio components according to the manual.  Now they are disappointing OSSM 2.0 due to this doc quality, then they asked us to show the example of configuration for HA.
       
  We have the following requests and concerns, would you please find them ?

   Requests:
   
     Would you please provide a clear and concise explanation about how to configure ServiceMeshControlPlance ? And please also clarify the following points in description ?
     
       #1. istiod, istio-egressgateway and istio-ingressgateway have 2 or more replicas for HA respectively (It's more better if we can configure multiple replicas even for prometheus, grafana, kiali, jaeger)
     
       #2. Those pods have a podAntiAffinity rule to avoid running on the same node as other replicas of the same component.
         - By using link [1] we are not able to configure Anti pod affinity rules for Istiod pod.
         [1]
https://docs.openshift.com/container-platform/4.6/service_mesh/v2x/ossm-custom-resources.html#ossm-cr-pilot_ossm-custom-resources-v2x
     
       #3. Those pods should have a toleration rule "operator: Exists" in order to avoid taints marked by user
         (Image Registry or other Operators have this toleration rule, so we want to follow  it in OSSM)

     Please clarify those points in the documents, so that our customers can understand correctly about OSSM2.0.
     (And if OSSM2.0 cannot satisfy the above, our customer would suspend the plan to use OSSM2.0, so we would like to need such a clarification in the product documents.)
     

   Concerns:

     Is Service Mesh 1.0 still supported by Red Hat with OCP4.6?
   

Suggestions for improvement:

  (Please see "Describe the issue" section.)
 
  Also, currently OSSM 2.0's document quality is too low. It's impossible for our customer to use OSSM 2.0 with such quality, so they want to keep using OSSM 1.0 even with OCP4.6. Will Red Hat keep supporting OSSM1.0? If so, until when will Red Hat continue to support it?

Additional information:

    OpenShift ServiceMesh(OSSM) 2.0 was released one month ago, but the quality of its manuals is too low as we already reported several bugzillas.
   
      -
BZ#1898429
: [RHOCP4.6] Service Mesh 2.0's manual still mentions 1.0's way for autoscaling
      -
BZ#1896607
: [RHOCP4.6] Cannot disable the ior feature in Service Mesh 2.0
      -
BZ#1896337
: [RHOCP4.6] Wrong namespace mentioned in smcp template in Service Mesh 2.0 doc
      -
BZ#1897490
: [RHOCP4.6] Wrong template is given for configuring Jaeger with Elastic Search in Service Mesh 2.0
   
    In addition to the aboves, our customer faced the issue that they cannot configure Anti-Affinity for istio components according to the manual.
    Now they are disappointing OSSM 2.0 due to this doc quality, then they asked us to show the example of configuration for HA.
    So, we wrote above, the current doc quality is very low.
    Nobody can use a product with such quality even if the product itself is brilliant.
    We hope Red Hat will improve the doc quality in near future.