Add support for operational alerts

[ravjotbrar]

Users currently need to manually check dashboards to notice problems. Threshold-based alerts provide proactive notification of issues without requiring constant monitoring.

Surface operational issues through alerting-oriented views and configurable thresholds. Users should be able to define conditions (e.g., memory usage > 80%, slow command spike, node unreachable) and see active alerts in the UI.

[antonin-suzor]

Hello @ravjotbrar ,
When you say "operational alerts", do you mean alerting via external systems, like slack channels, pagerduty, that sort of thing ?
Or is this just in the UI ?
If it's just in the UI, we could do any of the following:

• having toasts pop up (and probably stay until dismissed) to warn the user
• integrate alerting into the current Dashboard view
• create a new Alerting view (though that may be overkill given that Dasboard already has all the relevant info)