Design level isolation for Applications on GKE

Google Cloud Serverless Applications
, ,
1

Multi-application deployment on GKE cluster

What factors are to be considered to decide the level of isolation among the apps?

e.g.

Scenario 1 -
1 cluster = N Namespace = M services (1 namespace = 2 or more services/apps).

Scenario 1A -
1 cluster = N Namespace = M services (1 namespace = 2 or more services/apps).

Each service having dedicated ingress.

Scenario 1B -
1 cluster = N Namespace = M services (1 namespace = 2 or more services/apps).

Ingress shared across services.

Scenario 2 -

1 cluster = N Namespace = N services (1 namespace = 1 service/app).

2

First, could you please elaborate on your purpose and goals, and which factors are most important to you? For example, which aspects are you primarily focused on - high availability, cost, security, multi-tenancy, or something else?