VectorDiscardEvents #

Meaning #

Vector discarded events because it was unable to keep up with the rate of incoming events.

Impact #

This means that some events were not processed by Vector. This means that some logs were not sent to Loki, which means that they are not available in Grafana.

Diagnosis #

Check the Vector logs for errors. If there are no errors, check the Vector metrics for the vector_events_discarded_total metric. If the metric is increasing, it means that Vector is unable to keep up with the rate of incoming events. Also check the other metrics on the vector dashboard for irregularities.

Mitigation #

To address this alert, you can either scale up the Vector deployment or reduce the rate of incoming events. As we deploy Vector as a DaemonSet, scaling up the deployment will scale up Vector on all nodes.