-
Bug
-
Resolution: Done
-
Normal
-
4.20
-
Quality / Stability / Reliability
-
False
-
-
None
-
Critical
-
None
-
None
-
Rejected
-
Installer Sprint 267, Installer Sprint 268, Installer Sprint 269, Installer Sprint 270, Installer Sprint 271
-
5
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Description of problem:
Create a cluster with custom DNS enabled on AWS, the bootstrap failed, checking pods on control plane, the coredns was not running:
[core@ip-10-0-3-86 ~]$ sudo crictl ps
CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID POD NAMESPACE
<empty>
install-config:
platform:
aws:
userProvisionedDNS: Enabled
featureSet: TechPreviewNoUpgrade
Version-Release number of selected component (if applicable):
4.19.0-0.nightly-2025-03-09-063419
How reproducible:
Always
Steps to Reproduce:
1.Config install-config.yaml as described in the description
2.Create cluster
3.
Actual results:
No coredns pod on control plane machines, bootstrap failed.
Expected results:
Except coredns running and work well on control plane machines, bootstrap process succeeds.
Additional info:
Checking kubelet log on master
Mar 11 03:06:00 ip-10-0-3-86 kubenswrapper[2485]: I0311 03:06:00.200021 2485 csi_plugin.go:887] Failed to contact API server when waiting for CSINode publishing: Get "https://api-int.yunjiang-dns12.qe.devcluster.openshift.com:6443/apis/storage.k8s.io/v1/csinodes/ip-10-0-3-86.us-east-2.compute.internal?resourceVersion=0": dial tcp: lookup api-int.yunjiang-dns12.qe.devcluster.openshift.com on 10.0.0.2:53: no such host
it looks looks like api-int is not resolvable, the name servers on control plane and bootstrap are:
find . -name resolv.conf -printf "%P\n" -exec cat {} \;
control-plane/10.0.66.134/network/resolv.conf
# Generated by NetworkManager
search us-east-2.compute.internal
nameserver 10.0.0.2
control-plane/10.0.57.233/network/resolv.conf
# Generated by NetworkManager
search us-east-2.compute.internal
nameserver 10.0.0.2
control-plane/10.0.3.86/network/resolv.conf
# Generated by NetworkManager
search us-east-2.compute.internal
nameserver 10.0.0.2
bootstrap/network/resolv.conf
# Generated by NetworkManager
nameserver 127.0.0.1
search us-east-2.compute.internal
nameserver 10.0.0.2
The cluster-infrastructure-02-config.yml:
status:
apiServerInternalURI: https://api-int.yunjiang-dns12.qe.devcluster.openshift.com:6443
apiServerURL: https://api.yunjiang-dns12.qe.devcluster.openshift.com:6443
controlPlaneTopology: HighlyAvailable
cpuPartitioning: None
etcdDiscoveryDomain: ""
infrastructureName: yunjiang-dns12-pd4mm
infrastructureTopology: HighlyAvailable
platform: AWS
platformStatus:
aws:
cloudLoadBalancerConfig:
clusterHosted:
apiIntLoadBalancerIPs:
- 10.0.46.213
- 10.0.16.77
- 10.0.76.110
apiLoadBalancerIPs:
- 3.142.117.227
- 3.135.145.102
- 3.146.246.29
dnsType: ClusterHosted
region: us-east-2
type: AWS
- blocks
-
CORS-3292 Start in-cluster DNS to resolve API and API-Int URLs by starting CoreDNS pods on bootstrap and Control plane nodes
-
- Closed
-
-
OCPSTRAT-992 [Dev Preview]: Allow customer managed DNS solutions for AWS: Implementation
-
- Closed
-
- links to