Description of problem: The catalogsource & cm can’t be created by marketplace Version-Release number of selected component (if applicable): clusterversion:4.0.0-0.2 marketplace image:quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:b39fe61df5582a5de0de4917dd61546be75d197962fbb46063ae407de9444d5f olm image: quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:279937a595bea87b70e5618783b0df6022aac31a9fba7bb104bb88e1a3f6bf79 olm commit: io.openshift.build.commit.id=1e295784b30a7d54eb2db82b99a9c6307133ebbf How reproducible:always Steps to Reproduce: 1. install the cluster Actual results: 1.no default catalogsource & cm created,and the marketplace has error message #oc logs marketplace-operator-5878c7986c-f8kv4 “E0128 09:07:48.516729 1 streamwatcher.go:109] Unable to decode an event from the watch stream: http2: server sent GOAWAY and closed the connection; LastStreamID=43, ErrCode=NO_ERROR, debug=""”E0128 09:07:48.516738 1 streamwatcher.go:109] Unable to decode an event from the watch stream: http2: server sent GOAWAY and closed the connection; LastStreamID=43, ErrCode=NO_ERROR, debug="" Expected results: 1.default catalogsource & cm will be created default,and no error message in marketplace’s pod #oc get catalogsource -n openshift-marketplaceNAME NAME TYPE PUBLISHER AGE certified-operators Certified Operators internal Red Hat 6h community-operators Community Operators internal Red Hat 6h redhat-operators Red Hat Operators internal Red Hat 6h #oc get cm -n openshift-marketplace NAME DATA AGE certified-operators 3 6h community-operators 3 6h redhat-operators 3 6h Additional info:
Jia, this is happening because is CVO is processing other higher priority operators and has not gotten to handling Marketplace yet. There is not much we can do to fix this on the Marketplace side.
>“E0128 09:07:48.516729 1 streamwatcher.go:109] Unable to decode an event from the watch stream: http2: server sent GOAWAY and closed the connection; LastStreamID=43, ErrCode=NO_ERROR, debug=""”E0128 09:07:48.516738 1 streamwatcher.go:109] Unable to decode an event from the watch stream: http2: server sent GOAWAY and closed the connection; LastStreamID=43, ErrCode=NO_ERROR, debug="" This indicates that something basic in the cluster is not working. For marketplace to work, basic networking, dns, apiserver etc need to functioning.
The same issue as https://bugzilla.redhat.com/show_bug.cgi?id=1666225#c11
Jian Zhang, Fan Jia, like I mentioned in my previous comment, the root cause of this issue is not the Marketplace operator but some basic functionality in the cluster is not working. There is nothing we can do on the Marketplace side regarding this.
Aravindh, Many thanks for your quick response. Yeah, I saw a similar fixed PR on Kubernetes: https://github.com/kubernetes/kubernetes/pull/73277. We hit this issue more often recently. I already @Clayton on bug 1666225, but no response by now. Do you know who responsible for this kind of issue? I'd like to label this bug as a duplicate of bug 1666225 for now. *** This bug has been marked as a duplicate of bug 1666225 ***
I am not sure which team is responsible for these sort of bugs. I suggest posting on #4-dev-triage.