duplicate endpoint stats

[zirain]

Same endpoint exists in multiple priority levels (e.g., priority 0 and priority 1), it will generate multiple metrics with the same labels.

reproduce yaml:

admin:
  address:
    socket_address:
      address: 0.0.0.0
      port_value: 9901

static_resources:
  listeners:
  - name: listener_0
    address:
      socket_address:
        address: 0.0.0.0
        port_value: 10000
    filter_chains:
    - filters:
      - name: envoy.filters.network.http_connection_manager
        typed_config:
          "@type": type.googleapis.com/envoy.extensions.filters.network.http_connection_manager.v3.HttpConnectionManager
          stat_prefix: ingress_http
          route_config:
            name: local_route
            virtual_hosts:
            - name: backend
              domains: ["*"]
              routes:
              - match:
                  prefix: "/"
                route:
                  cluster: test_cluster
          http_filters:
          - name: envoy.filters.http.router
            typed_config:
              "@type": type.googleapis.com/envoy.extensions.filters.http.router.v3.Router

  clusters:
  - name: test_cluster
    type: STATIC
    lb_policy: ROUND_ROBIN
    
    # Enable per-endpoint stats - this is required to see endpoint_healthy metrics
    track_cluster_stats:
      per_endpoint_stats: true
    
    # Configure multiple priorities with the SAME endpoint
    # This reproduces the duplicate metrics issue
    load_assignment:
      cluster_name: test_cluster
      endpoints:
      # Priority 0 - endpoint is healthy
      - priority: 0
        lb_endpoints:
        - endpoint:
            address:
              socket_address:
                address: 127.0.0.1
                port_value: 8443
            hostname: "test-endpoint.example.local"
          health_status: HEALTHY  # This will show healthy=1
      
      # Priority 1 - SAME endpoint but unhealthy
      - priority: 1
        lb_endpoints:
        - endpoint:
            address:
              socket_address:
                address: 127.0.0.1
                port_value: 8443
            hostname: "test-endpoint.example.local"
          health_status: UNHEALTHY  # This will show healthy=0

output:

curl -s http://localhost:9901/stats | grep endpoint | grep healthy
cluster.test_cluster.endpoint.127.0.0.1_8443.healthy: 1
cluster.test_cluster.endpoint.127.0.0.1_8443.healthy: 0

[wbpcode]

duplicated endpoint in the envoy is unexpected in lots of case, because in lots of features, we will use the endpoint address as the unique key.

But for the problem self, let's see could we add additional tag or adding a new alt_stat_name for the host to address this problem.