Join GitHub today
GitHub is home to over 40 million developers working together to host and review code, manage projects, and build software together.Sign up
CoreDNS not started with k8s 1.11 and weave (CentOS 7) #998
Is this a BUG REPORT or FEATURE REQUEST?
kubeadm version 1.11
after kubeadm init the coreos pods stay in Error
The logs of both pods show the following:
@dims , can make sense, I will try
docker version: docker-1.13.1-63.git94f4240.el7.centos.x86_64
coredns pods log:
Same issue for me. Similar setup CentOS 7.4.1708, Docker version 1.13.1, build 94f4240/1.13.1 (comes with CentOS):
Same issue for me.
FYI, Also works for me with SELinux disabled (not permissive, but disabled).
Same issue here
So, it appears as if there is a incompatibility between old versions of docker and SELinux with the allowPrivilegeEscalation directive which has apparently been resolved in later versions of docker.
There appear to be 3 different work-arounds:
There is also answer for that in stackoverflow:
is caused when CoreDNS detects a loop in the resolve configuration, and it is the intended behavior. You are hitting this issue:
Hacky solution: Disable the CoreDNS loop detection
Edit the CoreDNS configmap:
Remove or comment out the line with
Then remove the CoreDNS pods, so new ones can be created with new config:
All should be fine after that.
Preferred Solution: Remove the loop in the DNS configuration
First, check if you are using
If it is, check which
You might see a line like:
The important part is
If it is the
Check the content of
If there is
To get rid of it, you should not edit that file, but check other places to make it properly generated.
Check all files under
delete that record. Also check
After doing all that, restart the systemd services to put your changes into effect:
After that, verify that
Finally, trigger re-creation of the DNS pods
Summary: The solution involves getting rid of what looks like a DNS lookup loop from the host DNS configuration. Steps vary between different resolv.conf managers/implementations.
I have same problem , and another problem
kubectl -n kube-system get deployment coredns -o yaml |
then pod rebuild only have one error
[ERROR] plugin/errors: 2 10594135170717325.8545646296733374240. HINFO: unreachable backend: no upstream host
I don't know if that's normal . maybe
2、the coredns cannot found my api service . error is
kube-dns Failed to list *v1.Endpoints getsockopt: 10.96.0.1:6443 api connection refused
coredns restart again and again ,at last will CrashLoopBackOff
so i have to run coredns on master node i do that
kubectl edit deployment/coredns --namespace=kube-system
I don't know if that's normal
at last give my env
Linux 4.20.10-1.el7.elrepo.x86_64 /// centos 7
docker Version: 18.09.3
[root@k8smaster00 ~]# docker image ls -a
kubenets is 1.13.3
I think this is a bug Expect an official update or a solution
@mengxifl, Those errors are significantly different than the ones reported and discussed in this issue.
Those errors mean that the CoreDNS pod (and probably all other pods) cannot reach your nameservers. This suggests a networking problem in your cluster to the outside world. Possibly flannel misconfiguration or firewalls.
This is also not normal. If I understand you correctly, you are saying that CoreDNS can contact the API from the master node but not other nodes. This would suggest pod to service networking problems between nodes within your cluster - perhaps an issue with flannel configuration or firewalls.
Thank you for your reply
maybe i should put up my yaml file
my config.yaml content is
my fannel yaml is default
so i don't think firewall have issue mybe fannel ? but i use default config . And maybe linux version . i don't know .
OK I run
on all my node that work for me . thanks