Details
-
Bug
-
Status: Done
-
Medium
-
Resolution: Fixed
-
2.18.0
-
perconalab/pmm-server-fb:PR-1773-99cfd95
-
Yes
-
1
-
Yes
-
Yes
-
[obsolete] Server Integrations
Description
User impact:
If user tries to register a cluster with operators already installed and higher version it causes Internal Sever error
STR:
Install PSMDB operator v1.8 (
curl -sSf -m 30 https://raw.githubusercontent.com/percona/percona-server-mongodb-operator/v1.8.0/deploy/bundle.yaml \
> | kubectl apply -f -
Register the K8s cluster
Result:
dbaas-controller.log
time="May 18 10:13:18.828541277" level=error msg="RPC /percona.platform.dbaas.controller.v1beta1.PSMDBOperatorAPI/InstallPSMDBOperator done in 4.684852883s with unexpected error: exit status 1\ncmd: /opt/dbaas-tools/bin/kubectl-1.16 --kubeconfig=/tmp/dbaas-controller-kubeconfig-480156622 apply -f -\nstderr: The CustomResourceDefinition \"perconaservermongodbs.psmdb.percona.com\" is invalid: status.storedVersions[0]: Invalid value: \"v1-8-0\": must appear in spec.versions\n" request=a814c508-b7c1-11eb-b1f4-080027d19c11 time="May 18 10:13:18.856333052" level=info msg="Starting RPC /percona.platform.dbaas.controller.v1beta1.KubernetesClusterAPI/CheckKubernetesClusterConnection ..." request=aae3dd2b-b7c1-11eb-b1f4-080027d19c11
pmm-managed.log
WARN[2021-05-18T10:13:18.828+00:00] RPC /dbaas.v1beta1.Kubernetes/RegisterKubernetesCluster done in 10.806145191s with gRPC error: rpc error: code = Internal desc = Internal server error. request=a46ecaf9-b7c1-11eb-a687-080027d19c11
It works as expected when trying to register a cluster with already installed operators which are same or lower version as current version. But downgrade fails.
Attachments
Issue Links
- is caused by
-
PMM-6432
APIs for installing operators
-
- Done
-