api: add error flag no_cluster_found (#15431)

As of today no route found and route is found but the cluster in that route is not ready are sharing the same error detail.
However, this make big difference. The latter is likely a RDS config problem, or explicit blackhole.
The latter might be a slow CDS.

Address #15448

Signed-off-by: Yuchen Dai <silentdai@gmail.com>

Mirrored from https://github.com/envoyproxy/envoy @ a11d48368c62eb2c238d6d3087c16ffa355243d1

envoy/config/accesslog/v3/accesslog.proto

@@ -245,6 +245,7 @@ message ResponseFlagFilter {
         in: "NFCF"
         in: "DT"
         in: "UPE"
+        in: "NC"
       }
     }
   }];

envoy/config/accesslog/v4alpha/accesslog.proto

@@ -244,6 +244,7 @@ message ResponseFlagFilter {
         in: "NFCF"
         in: "DT"
         in: "UPE"
+        in: "NC"
       }
     }
   }];

envoy/data/accesslog/v3/accesslog.proto

@@ -186,7 +186,7 @@ message AccessLogCommon {
 }
 
 // Flags indicating occurrences during request/response processing.
-// [#next-free-field: 25]
+// [#next-free-field: 26]
 message ResponseFlags {
   option (udpa.annotations.versioning).previous_message_type =
       "envoy.data.accesslog.v2.ResponseFlags";
@@ -278,6 +278,9 @@ message ResponseFlags {
 
   // Indicates there was an HTTP protocol error in the upstream response.
   bool upstream_protocol_error = 24;
+
+  // Indicates no cluster was found for the request.
+  bool no_cluster_found = 25;
 }
 
 // Properties of a negotiated TLS connection.