This document explains why alerts are affected by enabling remote monitoring in the Apache, Cassandra, MySQL, Nginx, and Redis integrations, and how to proceed if this happens.
We strongly recommend reviewing the configuration of your alerts after activating the
Activating remote monitoring in the Apache, Cassandra, MySQL, Nginx, and Redis integrations modifies the entity reporting data. As a result, any alert set using entity-related parameters could stop behaving as expected and may need to be reconfigured.
Let's say you have an integration running with disabled remote monitoring and you set an alert to monitor the number of connected clients filtered by the entity key, in this case,
You decide to enable remote monitoring in your integration, and set the parameter
remote_monitoring to true. After restarting the infrastructure agent, the alert stops monitoring. This happens because the entity resulting from activating remote monitoring has a different key, and therefore the current filter does not apply.
Alert before and after activating remote monitoring.
By filtering by the new entity key value, in this example
instance:127.0.0.1:6379, your alert will be working as expected again.
With the new filter the alert behaves as expected.