Multiple TCP listener definitions for 0.0.0.0:6379 in Pilot (kubernetes external services)

[Lookyan]

Is this a BUG or FEATURE REQUEST?: BUG

What Version of Istio and Kubernetes are you using, where did you get Istio from, Installation details Istio from 0.8 release branch built from source.

istioctl version
Version: 6dbbacac0b478017179480778637c9d8d781ac25-dirty
GitRevision: 6dbbacac0b478017179480778637c9d8d781ac25-dirty
User: lookyan
Hub: docker.io/istio
GolangVersion: go1.9.3
BuildStatus: Modified

kubectl version
Client Version: version.Info{Major:"1", Minor:"8", GitVersion:"v1.8.0", GitCommit:"6e937839ac04a38cac63e6a7a306c5d035fe7b0a", GitTreeState:"clean", BuildDate:"2017-09-28T22:57:57Z", GoVersion:"go1.8.3", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"8", GitVersion:"v1.8.11", GitCommit:"1df6a8381669a6c753f79cb31ca2e3d57ee7c8a3", GitTreeState:"clean", BuildDate:"2018-04-05T17:16:46Z", GoVersion:"go1.8.3", Compiler:"gc", Platform:"linux/amd64"}

Is Istio Auth enabled or not ? I used istio.yaml with disabled Auth option.

What happened: I have such messages in Pilot logs:

2018-05-29T14:48:49.006986Z	warn	Multiple TCP listener definitions for 0.0.0.0:6379 name:"tcp_0.0.0.0_6379" address:<socket_address:<address:"0.0.0.0" port_value:6379 > > filter_chains:<filter_chain_match:<> filters:<name:"envoy.tcp_proxy" config:<fields:<key:"deprecated_v1" value:<bool_value:true > > fields:<key:"value" value:<struct_value:<fields:<key:"route_config" value:<struct_value:<fields:<key:"routes" value:<list_value:<values:<struct_value:<fields:<key:"cluster" value:<string_value:"outbound|6379||external-app-redis00.default.svc.cluster.local" > > > > > > > > > > fields:<key:"stat_prefix" value:<string_value:"outbound|tcp|6379" > > > > > > > > deprecated_v1:<bind_to_port:<> >  external-autoload-redis01.default.svc.cluster.local
2018-05-29T14:48:49.007070Z	warn	Multiple TCP listener definitions for 0.0.0.0:6379 name:"tcp_0.0.0.0_6379" address:<socket_address:<address:"0.0.0.0" port_value:6379 > > filter_chains:<filter_chain_match:<> filters:<name:"envoy.tcp_proxy" config:<fields:<key:"deprecated_v1" value:<bool_value:true > > fields:<key:"value" value:<struct_value:<fields:<key:"route_config" value:<struct_value:<fields:<key:"routes" value:<list_value:<values:<struct_value:<fields:<key:"cluster" value:<string_value:"outbound|6379||external-app-redis00.default.svc.cluster.local" > > > > > > > > > > fields:<key:"stat_prefix" value:<string_value:"outbound|tcp|6379" > > > > > > > > deprecated_v1:<bind_to_port:<> >  external-autoload-redis02.default.svc.cluster.local
2018-05-29T14:48:49.007149Z	warn	Multiple TCP listener definitions for 0.0.0.0:6379 name:"tcp_0.0.0.0_6379" address:<socket_address:<address:"0.0.0.0" port_value:6379 > > filter_chains:<filter_chain_match:<> filters:<name:"envoy.tcp_proxy" config:<fields:<key:"deprecated_v1" value:<bool_value:true > > fields:<key:"value" value:<struct_value:<fields:<key:"route_config" value:<struct_value:<fields:<key:"routes" value:<list_value:<values:<struct_value:<fields:<key:"cluster" value:<string_value:"outbound|6379||external-app-redis00.default.svc.cluster.local" > > > > > > > > > > fields:<key:"stat_prefix" value:<string_value:"outbound|tcp|6379" > > > > > > > > deprecated_v1:<bind_to_port:<> >  external-avi-antifraud-redis01.default.svc.cluster.local

There are many entries like this in logs for each external service. And Pilot tries to reload them permanently and write it in logs. Pilot uses 0.0.0.0 but these external services have different hosts and addr:port pair shouldn't be the same for them.

What you expected to happen: Don't have these messages.

How to reproduce it: Create several kubernetes external services with the same ports and run pilot.