UPSTREAM: <carry>: openshift: Remove mutating logic from exist and set vm state in update #62
Conversation
openshift-ci-robot
added
the
size/M
|
Let's wait until |
openshift-ci-robot
added
size/L
I0723 18:15:36.972456 8674 autoscaler.go:461] [10s remaining] Waiting for cluster to reach original machine count of 6; currently have 7
I0723 18:15:40.069120 8674 autoscaler.go:461] [7s remaining] Waiting for cluster to reach original machine count of 6; currently have 7
I0723 18:15:43.166344 8674 autoscaler.go:461] [4s remaining] Waiting for cluster to reach original machine count of 6; currently have 7
I0723 18:15:46.264489 8674 autoscaler.go:461] [0s remaining] Waiting for cluster to reach original machine count of 6; currently have 7
• Failure [646.895 seconds]
[Feature:Machines][Serial] Autoscaler should
/tmp/tmp.eiofjseXFX/src/github.com/openshift/cluster-api-actuator-pkg/pkg/e2e/autoscaler/autoscaler.go:229
scale up and down [It]
/tmp/tmp.eiofjseXFX/src/github.com/openshift/cluster-api-actuator-pkg/pkg/e2e/autoscaler/autoscaler.go:230
Timed out after 540.000s.
Expected
<int>: 7
to equal
<int>: 6/test e2e-azure-operator |
|
/test e2e-azure-operator |
…t vm state in update
enxebre
changed the title
|
@enxebre: The following test failed, say
Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here. |
|
label didn't get applied |
|
@enxebre: you cannot LGTM your own PR. In response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
/lgtm |
|
/approve |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: frobware The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
openshift-ci-robot
added
the
approved
| @@ -283,45 +283,58 @@ func (s *Reconciler) Update(ctx context.Context) error { | |||
| Message: machineCreationSucceedMessage, | |||
| }) | |||
|
|
|||
| vmState := v1beta1.VMState(*vm.ProvisioningState) | |||
There was a problem hiding this comment.
Original purpose of isVMExists was to decide if vm object exists. After the vm object was successfully retrieved, Exists went through additional checking. Exists was true only when vm was in VMStateSucceeded or VMStateUpdating state. Though, after isVMExists returned true, both s.scope.MachineStatus.VMID and s.scope.MachineStatus.VMState were set. Given that your PR got merged, the workflow of setting VMID and VMState is different now. Exists still returns false/true the same way. However, s.scope.MachineStatus.VMID and s.scope.MachineStatus.VMState will not be set until Update op is called.s.scope.MachineStatus.VMState will always be set to Succeeded or Failed. It will be empty for the first 40s. The whole purpose of that field was to report in which state during provisioning a vm is. Now it's broken.
Remove mutating logic from Exist(). Mutating logic on exists is semantically wrong as well as it leads to stale objects api update issues, given the machine controller workflow design. The machine controller workflow always heads towards
if exist then updateany mutating logic other than status should preferably belong toupdateso it will become eventually consistent.Follow up for #61