kubedns addon always report readines failed

[typhoonzero]

Use latest sextant generated kubedns addon files, then kubectl --namespace=kube-system describe po kube-dns-v20-qnm58 shows:

  7m		7m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Started		Started container with docker id a4112ffaedea
  7m		7m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Killing		Killing container with docker id 28d8f68674f8: pod "kube-dns-v20-qnm58_kube-system(5fbb721e-fcd5-11e6-ae9a-6c92bf04de02)" container "kubedns" is unhealthy, it will be killed and re-created.
  7m		7m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{dnsmasq}	Normal		Created		Created container with docker id f1cc72776fb2; Security:[seccomp=unconfined]
  5m		5m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Killing		Killing container with docker id a4112ffaedea: pod "kube-dns-v20-qnm58_kube-system(5fbb721e-fcd5-11e6-ae9a-6c92bf04de02)" container "kubedns" is unhealthy, it will be killed and re-created.
  5m		5m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Created		Created container with docker id 8ccb308c81b4; Security:[seccomp=unconfined]
  5m		5m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{dnsmasq}	Normal		Killing		Killing container with docker id f1cc72776fb2: pod "kube-dns-v20-qnm58_kube-system(5fbb721e-fcd5-11e6-ae9a-6c92bf04de02)" container "dnsmasq" is unhealthy, it will be killed and re-created.
  5m		5m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{dnsmasq}	Normal		Started		Started container with docker id 4e868eaf19a4
  5m		5m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{dnsmasq}	Normal		Created		Created container with docker id 4e868eaf19a4; Security:[seccomp=unconfined]
  5m		5m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Started		Started container with docker id 8ccb308c81b4
  3m		3m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{dnsmasq}	Normal		Killing		Killing container with docker id 4e868eaf19a4: pod "kube-dns-v20-qnm58_kube-system(5fbb721e-fcd5-11e6-ae9a-6c92bf04de02)" container "dnsmasq" is unhealthy, it will be killed and re-created.
  3m		3m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Killing		Killing container with docker id 8ccb308c81b4: pod "kube-dns-v20-qnm58_kube-system(5fbb721e-fcd5-11e6-ae9a-6c92bf04de02)" container "kubedns" is unhealthy, it will be killed and re-created.
  11m		1m		18	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Warning		Unhealthy	Liveness probe failed: HTTP probe failed with statuscode: 503
  1m		1m		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{dnsmasq}	Normal		Killing		Killing container with docker id 4c0133b1e2d6: pod "kube-dns-v20-qnm58_kube-system(5fbb721e-fcd5-11e6-ae9a-6c92bf04de02)" container "dnsmasq" is unhealthy, it will be killed and re-created.
  52s		52s		1	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Killing		(events with common reason combined)
  9m		52s		5	{kubelet 6c-92-bf-13-46-77}	spec.containers{dnsmasq}	Normal		Pulled		Container image "yq01-idl-gpu-jpaas-let00.yq01.baidu.com:5000/pineking/kube-dnsmasq-amd64:1.4" already present on machine
  9m		51s		5	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Pulled		Container image "yq01-idl-gpu-jpaas-let00.yq01.baidu.com:5000/pineking/kubedns-amd64:1.9" already present on machine
  3m		51s		4	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Created		(events with common reason combined)
  3m		51s		4	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Normal		Started		(events with common reason combined)
  12m		2s		72	{kubelet 6c-92-bf-13-46-77}	spec.containers{kubedns}	Warning		Unhealthy	Readiness probe failed: Get http://10.1.97.2:8081/readiness: dial tcp 10.1.97.2:8081: getsockopt: connection refused