Option to disable writing k8s events at workflow level

[tooptoop4]

Summary

Environment variable/config to stop writing k8s events

Use Cases

When would you use this? When i want to stop etcd filling up

These parts of the code are writing k8s events which end up being stored in etcd:

argo-workflows/workflow/controller/operator.go

Lines 2339 to 2344 in 0bfaea6

	case wfv1.WorkflowRunning:
	woc.eventRecorder.Event(woc.wf, apiv1.EventTypeNormal, "WorkflowRunning", "Workflow Running")
	case wfv1.WorkflowSucceeded:
	woc.eventRecorder.Event(woc.wf, apiv1.EventTypeNormal, "WorkflowSucceeded", "Workflow completed")
	case wfv1.WorkflowFailed, wfv1.WorkflowError:
	woc.eventRecorder.Event(woc.wf, apiv1.EventTypeWarning, "WorkflowFailed", message)

argo-workflows/workflow/controller/operator.go

Line 783 in 0bfaea6

woc.recordNodePhaseChangeEvents(woc.orig.Status.Nodes, woc.wf.Status.Nodes)

With a lot of workflows being churned through these events add a lot of storage pressure to etcd. At the same time i have not found these events useful as i can always look at logs in the controller pod

[terrytangyuan]

argo-workflows/docs/workflow-controller-configmap.yaml

Lines 36 to 42 in 026b14e

	# Whether or not to emit events on node completion. These can take a up a lot of space in
	# k8s (typically etcd) resulting in errors when trying to create new events:
	# "Unable to create audit event: etcdserver: mvcc: database space exceeded"
	# This config item allows you to disable this.
	# (since v2.9)
	nodeEvents: |
	enabled: true

[tooptoop4]

@terrytangyuan i already have that as false but it still sends events!! seems i am getting events at workflow level