Current Behavior

APISIX is throwing an error related to port being 0 when upstream has discovery_type "dns".
same exact setup works fine with clusters from other kubernetes providers.

Expected Behavior

Should not throw the error http_access_phase(): failed to set upstream: invalid nodes format: failed to validate item 1: property "port" validation failed: expected 0 to be at least 1

Error Logs

in web response -
503 Service Temporarily Unavailable
in apisix pod -
init.lua:520: http_access_phase(): failed to set upstream: invalid nodes format: failed to validate item 1: property "port" validation failed: expected 0 to be at least 1

Steps to Reproduce

1. Create a standard Google Kubernetes Engine cluster.
2. Install APISIX with ingress controller and dashboard via api7 helm chart version 0.8.4 using this values.yaml file (in ingress-apisix namespace).
3. create this upstream via admin api - { "id": "398808046194655761", "create_time": 1647237719, "update_time": 1647237719, "timeout": { "connect": 6, "send": 6, "read": 6 }, "type": "roundrobin", "scheme": "http", "discovery_type": "dns", "pass_host": "node", "name": "apisix-dashboard", "service_name": "apisix-dashboard.ingress-apisix.svc.cluster.local", "keepalive_pool": { "idle_timeout": 60, "requests": 1000, "size": 320 } }
   4. create this route via admin api - {"id":"398808266596942353","create_time":1647237850,"update_time":1647441243,"uri":"/*","name":"apisix-dashboard","priority":50,"methods":["GET","POST","PUT","DELETE","PATCH","HEAD","OPTIONS","CONNECT","TRACE"],"host":"apisix-dashboard.example.com","plugins":{"redirect":{"http_to_https":true}},"upstream_id":"398808046194655761","labels":{"API_VERSION":"v1"},"status":1}
   5. visit the host in the route and then check the logs in the apisix pod.

Environment

• APISIX version (run apisix version): 2.12.1
• Operating system (run uname -a): kubernetes (GKE)
  Rest of the values are whatever is bundled in apisix/apisix version 0.8.4 helm chart.
• OpenResty / Nginx version (run openresty -V or nginx -V):
• etcd version, if relevant (run curl http://127.0.0.1:9090/v1/server_info):
• APISIX Dashboard version, if relevant:
• Plugin runner version, for issues related to plugin runners:
• LuaRocks version, for installation issues (run luarocks --version):