Last9 anomaly detection eliminates the need to set alarms on individual components that come together to power a service. But at times, these components are also monitored elsewhere or don't need alerting because of some business context. To avoid redundant alerting, you can disable anomaly detection. This can be done on a per-metric, per-service basis as well.
For example, for the below example service, there are multiple AWS Lambda SLIs for which the system has started analyzing the data and is alerting the service owners
For scenarios where the alerts are important only during a particular time of the day
The channels tab shows us the configured channels for current service. As Last9 sends alerts for a service namespace, we have the ability to snooze alerts for the service as well.
If the service is under maintenance, simply snooze the alerts and un-snooze them later. Most of the times, this is all that is enough to contain the inflow of alerts for a service.
To do so, visit the service settings page of the service by clicking the ⚙️ icon towards the top right of the chart canvas.
Once in the settings page, on the left sidebar click Anomaly detection and disable the metric that you do not wish to be alerted on.
Most of the times, this is all that is enough to contain the inflow of alerts for a service. In rare circumstances, you can actually unset SLOs, or delete the channels as well. But we recommend snoozing as the best way to go about it.
In rare circumstances, you may want us to stop sending you alerts completely. Just contact us and we have a big red button to completely stop all alerts to you on any channel.