Simon Pasquier

I'm not sure to understand your issue. Can you show us an example of notification that you get and what you would expect instead?

Sorry but it isn't very clear from the screenshot what is happening here. It would be easier to share Alertmanager's logs in debug mode.

Ok I understand now. What happens is that the `Lost connection with localNode` alert has been firing and resolved between 11:22 and 11:23. Though it was never notified as firing,...

I suppose that it's more a question for the Helm chart. From what I can tell, your `config` AlertmanagerConfig resource is used in this [scenario](https://prometheus-operator.dev/docs/developer/alerting/#using-alertmanagerconfig-resources) while you expected https://prometheus-operator.dev/docs/developer/alerting/#using-alertmanagerconfig-for-global-configuration

Please share the Alertmanager manifest.

The semantics of ExceptLast (which is a bad name tbh) would be that when an alert with namespace="X" is received, it is dispatched to all AlertmanagerConfig objects in that namespace...

> should the ExceptLast strategy update the child routes to continue: true as well? Or should change only the first-level routes and stop in the last first-level route? The latter,...

@vince-weka can you check if you see `repeat_interval is greater than the data retention period. It can lead to notifications being repeated more often than expected` in the logs? @RandellP...

Yes it looks like something useful to have. We would have to look into the details to see how easy/difficult it is to leverage the existing code.

Thanks! > but I'm sure this will be a significant time saver for a lot of users and "wait for config to get applied" tests/checks. I haven't looked into the...