Unified alerting HTTP API in Grafana?

Monitoring and Observability Grafana
#1

Is there any documentation for the unified alerting HTTP API? Aka, how can I register a new rule by posting the rule yaml payload to an endpoint, same question for the AlertManager configuration.

#2

You can view the specification here
an example for creating a grafana-managed rule querying a prometheus datasource and having 2 expressions reduce and math (the math expression is the rule condition)
the namespace URL parameter should point to an existing folder (under which the rule will be created)

#3

And an example for POSTing an alertmanager configuration

#4

Could I ask if there is a way to drop in prometheus rules yaml into one of these endpoints? Or the rules need to be in specific format that is specicifc to Grafana?

#5

If you use cortex as prometheus datasource you can use POST​ /api​/ruler​/{Recipient}​/api​/v1​/rules​/{Namespace} where the Recipient should be the datasource id but this route expects JSON (not YAML).
An example payload is:

    "name": "group42",
    "rules": [
        {
            "alert": "prom",
            "expr": "(node_filesystem_avail_bytes{job=\"integrations/node_exporter\",fstype!=\"\"} / node_filesystem_size_bytes{job=\"integrations/node_exporter\",fstype!=\"\"} * 100 < 5 and node_filesystem_readonly{job=\"integrations/node_exporter\",fstype!=\"\"} == 0)",
            "for": "1m"
        }
    ]
}```
#6

In that case grafana proxies the request to this route