-
Bug
-
Resolution: Done-Errata
-
Normal
-
None
-
4.14
-
Moderate
-
No
-
False
-
Description of problem:
Cloud provider kubevirt will hang indefinitely, causing the creation and destruction of all LoadBalancer services in a KubeVirt guest cluster to get stuck in pending.
Version-Release number of selected component (if applicable):
4.13
How reproducible:
run the "should serve endpoints on same port and different protocol for internal traffic on Type LoadBalancer" ocp conformance test using an AWS infra cluster. The AWS ELB does not support multi protocol traffic, which triggers the cloud provider kubevirt controller to get stuck in an infinite loop waiting for the LB to provision. The infinite polling loop can be found here, https://github.com/kubevirt/cloud-provider-kubevirt/blob/aa46f3cdd0bf2ad7c8a4ad8eaac90e34120cc0f0/pkg/provider/loadbalancer.go#L95
Steps to Reproduce:
1. 2. 3.
Actual results:
cloud provider kubevirt controller hangs indefinitely unable to process new requests
Expected results:
cloud provider kubevirt can continue to process requests despite the incomplete LB on the infra cluster
Additional info:
- links to
-
RHEA-2023:7198 rpm