When the maximum number of connections of a cluster is reached, the TemBoard agent can no longer connect to the PostgreSQL cluster to retrieve metrics. Once the situation returns to normal (when it can connect again to collect metrics), the client sessions graph displays an average between the previous value and the new value, without indicating that the maximum value was reached, and therefore without triggering an alert.
However, the agent should be aware that it was unable to connect to the cluster for this specific reason because the returned error message states: "sorry, too many clients already".
In this specific case, is it possible to make the number of connections in the graph rise to 100%, thus triggering an alert?
When the maximum number of connections of a cluster is reached, the TemBoard agent can no longer connect to the PostgreSQL cluster to retrieve metrics. Once the situation returns to normal (when it can connect again to collect metrics), the client sessions graph displays an average between the previous value and the new value, without indicating that the maximum value was reached, and therefore without triggering an alert.
However, the agent should be aware that it was unable to connect to the cluster for this specific reason because the returned error message states: "sorry, too many clients already".
In this specific case, is it possible to make the number of connections in the graph rise to 100%, thus triggering an alert?