-
Bug
-
Resolution: Unresolved
-
Critical
-
None
-
4.19.0
Description of problem:
set up a bastion host as proxy server, then trigger the installation would tell failing to wait for the Kubernetes API, although the installation succeeds finally by itself
Version-Release number of selected component (if applicable):
see the build https://prow.ci.openshift.org/view/gs/test-platform-results/logs/release-openshift-origin-installer-launch-aws-modern/1910510784354455552
How reproducible:
Always
Steps to Reproduce:
1. create VPC network/subnets/etc. (see [1]) 2. create bastion host within the VPC, which will act as http proxy server (see [2]) 3. "create install-config", then insert the interested settings (see [3]) 4. export http_proxy & https_proxy settings (see [4]) 5. "create cluster" (see [5])
Actual results:
Failed on waiting for the Kubernetes API with below error, please note that the IP "34.8.255.171" is the API server's IP, rather than the bastion host's IP. ERROR Attempted to gather ClusterOperator status after installation failure: listing ClusterOperator objects: Get "https://api.jiwei-0411z.qe.gcp.devcluster.openshift.com:6443/apis/config.openshift.io/v1/clusteroperators": proxyconnect tcp: dial tcp 34.8.255.171:3128: i/o timeout
Expected results:
Installer should not take API server as the proxy server, and tell successful installation.
Additional info:
The bootstrap log bundle is available at https://drive.google.com/file/d/17zidRrhJEy1XV6f7fmwfqIOyokWdUcm7/view?usp=drive_link
- blocks
-
HIVE-1812 Support external DNS for OpenShift on GCP
-
- Testing
-
-
OCPSTRAT-990 [Tech Preview] Allow customer managed DNS solutions for GCP: Implementation
-
- Closed
-
- is related to
-
CORS-3992 Allow GCP custom-DNS to be enabled on all installs
-
- Dev Complete
-
- links to