Console plugin should read the alerts definitions and firing status (or execute related promql) to display:

• a global network health indicator, that is an aggregation of all alert statuses
• per-alert status

It fetches all alerts having the annotation "netobserv.io/health".

The global indicator should work as follow:

• it takes the highest severity of all firing alerts, with color codes (red=critical, yellow=warning etc.)
• it shows a score of number of firing alerts / number of total alerts for each severity

Here's an example showing 3 different states (not final design - must be patternflied) :

Below that, for each firing alert, the list of targets (labels) for which it is firing, and the metric value compared to threshold, in a table.

Table headers are:

• alert name
• source (source-prefixed labels set)
• destination (destination-prefixed labels set)
• labels (other labels set)
• severity
• metric value
• threshold
• trend

The trend indicates the evolution over time.

For instance, this promql:

100 * sum (rate(netobserv_namespace_drop_packets_total[1h])) / sum (rate(netobserv_namespace_ingress_packets_total[1h])) - 100 * sum(rate (netobserv_namespace_drop_packets_total[1h] offset 2h)) / sum(rate(netobserv_namespace_ingress_packets_total[1h] offset 2h)) 

=> value: 1.48

shows %drops now compared to 2h ago (+1.48%)