Change IsPlatformAlertRule implementation #1
Conversation
| return ni, err | ||
| } | ||
|
|
||
| func namespaceListWatch(client corev1client.CoreV1Interface) *cache.ListWatch { |
There was a problem hiding this comment.
maybe set a LabelSelector for openshift.io/cluster-monitoring=true to reduce watch/list volume
There was a problem hiding this comment.
nice catch, thought i added it, but forgot
| ) | ||
|
|
||
| const ( | ||
| // ClusterMonitoringLabel is the label used to identify namespaces with cluster monitoring enabled |
There was a problem hiding this comment.
This label is used to define Platform alerts.
Alerts defined in a PrometheusRule in a namespace that has this label will be evaluated by the platform stack.
@machadovilaca This should say that we set the For |
machadovilaca
force-pushed
the
change-IsPlatformAlertRule-implementation
branch
from
2e122a4 to
24f4d76
Compare
not yet, next PR |
| } | ||
| ni.updateMonitoringNamespace(ns) | ||
| }, | ||
| DeleteFunc: func(obj interface{}) { |
There was a problem hiding this comment.
consider checking also obj.(cache.DeletedFinalStateUnknown) in case of fast create and delete
Signed-off-by: machadovilaca <machadovilaca@gmail.com>
machadovilaca
force-pushed
the
change-IsPlatformAlertRule-implementation
branch
from
24f4d76 to
3c503cb
Compare
This PR refactors how platform managed alert rules are identified in management API and Kubernetes client. It removes hard-coded namespace naming assumptionson platform namespace, and now relies on informers for to check the cluster monitoring label on the namespaces
https://issues.redhat.com/browse/CNV-71645