The pods can't reach the network

[Tehacz]

Hi,

I have a mixed cluster. The pods on linux machines are working properly but the pods scheduled on windows machine cannot reach outside world. They only can ping themselves and master node IP. They cant reach linux nodes or for example a 8.8.8.8. Dns names are resolved though.

In the flannel 0.12 there is constant error log:
Failed to list *v1.Node: Get https://xxxx:6443/api/v1/nodes?resourceVersion=0: http2: no cached connection was available\n","stream":"stderr"

Any solution to this problem?

[oe-hbk]

Ping (ICMP) will not work from within a windows pod to connect to service IPs (like the DNS server) To do actual tests against services, you should to a curl or Test-NetConnection -Port XX type of call.

[xunchangguo]

i have the same error. but still now do not found a solution to this problem

[jsturtevant]

This is a duplicate of #103. Closing this in favor of that issue which is currently active and has a few work around listed.

/close

[k8s-ci-robot]

@jsturtevant: Closing this issue.

In response to this:

This is a duplicate of #103. Closing this in favor of that issue which is currently active and has a few work around listed.

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.