-
Bug
-
Resolution: Unresolved
-
Critical
-
None
-
4.14.z
-
Critical
-
No
-
CLOUD Sprint 255, CLOUD Sprint 256, CLOUD Sprint 257
-
3
-
False
-
-
-
Description of problem:
Machine controller looking for incorrect backend service name in GCP UPI resulting in error like below which is causing issues during machine scaleup. ~~~ E0527 15:49:04.656956 1 actuator.go:58] <infra-id>-master-0 error: <infra-id>-master-0: reconciler failed to Update machine: failed to register instance to instance group: failed to ensure that instance group <infra-id>-master-<region>-a is a proper instance group: failed to retrieve the backend service: backendServiceGet request failed: googleapi: Error 404: The resource 'projects/<project>/regions/<region>/backendServices/<infra-id>-api-internal' was not found, notFound E0527 15:49:04.657004 1 controller.go:284] <infra-id>-master-0: error updating machine: <infra-id>-master-0: reconciler failed to Update machine: failed to register instance to instance group: failed to ensure that instance group <infra-id>-master-europe-west3-a is a proper instance group: failed to retrieve the backend service: backendServiceGet request failed: googleapi: Error 404: The resource 'projects/<project>/regions/<region>/backendServices/<infra-id>-api-internal' was not found, notFound, retrying in 30s seconds I0527 15:49:04.657139 1 recorder.go:104] events "msg"="<infra-id>-master-0: reconciler failed to Update machine: failed to register instance to instance group: failed to ensure that instance group <infra-id>-master-<region>-a is a proper instance group: failed to retrieve the backend service: backendServiceGet request failed: googleapi: Error 404: The resource 'projects/<project>/regions/<region>/backendServices/<infra-id>-api-internal' was not found, notFound" "object"={"kind":"Machine","namespace":"openshift-machine-api","name":"<infra-id>-master-0","uid":"xxxxxxxxxxxxxxxxxxxx","apiVersion":"machine.openshift.io/v1beta1","resourceVersion":"425037150"} "reason"="FailedUpdate" "type"="Warning" ~~~ The backend service name should is <infra-id>-api-internal-backend-service but the machine-controller is looking for <infra-id>-api-internal which is incorrect. As per the below docs, the backend service is created with the name <infra-id>-api-internal-backend-service. https://github.com/openshift/installer/blob/8310b1083821260b106a373d2136558bc92c7462/docs/user/gcp/install_upi.md#add-bootstrap-instance-group-to-the-internal-load-balancer-backend-service https://docs.openshift.com/container-platform/4.14/installing/installing_gcp/installing-restricted-networks-gcp.html#installation-creating-gcp-bootstrap_installing-restricted-networks-gcp
Version-Release number of selected component (if applicable):
4.14
How reproducible:
100%
Steps to Reproduce:
1. Install a cluster on GCP with UPI installation method 2. Try scaling a machineset 3. Error can be seen in machine-controller
Actual results:
The machine-controller is looking for backend service name <infra-id>-api-internal
Expected results:
Machine controller should be looking for <infra-id>-api-internal-backend-service name
Additional info:
- relates to
-
OCPBUGS-43437 UPI docs do not reference correct internal api name (Docs update)
- Verified
-
OCPBUGS-35366 UPI docs do not reference correct internal api name
- Closed