Alert types
Kloudfuse supports many alert types that support your observability experience.
Kloudfuse supports these specific alerting types:
- Metric Alerts
-
Metric monitors analyze continuous streams of data. When collecting metrics through Kloudfuse or APIs, you can configure alerts to trigger on specified conditions:
- Threshold Alert
-
An alert triggered when a metric crosses a threshold over a specified period of time.
Tracks when a metric crosses a specified boundary value, to monitor signals that exceed known acceptable ranges.
- Change Alert
-
An alert triggered when the change between values is greater than the specified tolerance over a specified period of time.
Tracks how quickly a metric changes by comparing current and past values.
- Outlier Alert
-
An alert triggered when a single observed data point has a markedly different value from the rest of the sample.
Identifies statistically significant deviations within the dataset.
- Anomaly Detection Alert
-
An alert triggered when a pattern in the data does not conform to expected behavior.
Uses ML Models to learn normal patterns in your metrics, accounting for seasonality, to identify unexpected patterns.
- Forecast Alert
-
An alert triggered when the forecast algorithm predicts that in the future, a metric will cross a threshold over a specified period of time.
Predicts future metric values to enable proactive deflection, before problems occur.
- Log Alerts
-
Log monitors generate alerts when a specified a specific pattern or condition is detected within log data — when something unusual or potentially problematic happens within the system, based on information recorded in logs.
- Threshold Alert
-
An alert triggered when a metric reported by a log crosses a threshold over a specified period of time.
Counts log patterns for frequency, and tracks errors, warning, and other important events in the log signals.
- Outlier Alert
-
An alert triggered when a single observed data point has a markedly different value from the rest of the sample.
Identifies when log patterns in one source show statistically significant deviations from other sources.
- Anomaly Detection Alert
-
An alert triggered when a pattern in the log data does not conform to expected behavior.
Uses ML Models to learn normal patterns in your log, detecting unexpected changes in log frequency or content.
- Forecast Alert
-
An alert triggered when the forecast algorithm predicts that in the future, an aggregation on logs will cross a threshold over a specified period of time.
Predicts future patterns to enable proactive deflection before potential issues become critical.
See Create a logs alert.
- APM Alerts
-
Monitor features of APM Service metrics, when the system detects significant issues within a monitored service: high latency, excessive errors, resource exhaustion, and so on.
- Service Threshold Alert
-
An alert triggered when an RED (request, error, duration) service metric crosses a threshold over a specified period of time.
Monitors service performance metrics to track response time, error rates, and request rates against defined acceptable ranges.
- Service Anomaly Alert
-
An alert triggered when a service metric pattern in the data does not conform to expected behavior.
Uses ML Models to detect unusual patterns in your service performance data.
- Span Metric Alert
-
An alert triggered when a trace metric crosses a specified threshold over a particular period of time.
Track detailed transaction performance through distributed tracing data to monitor operations and requests.
See Create an APM alert.
- RUM Alerts
-
Monitor the performance and user experience of real users, in real time, to ensure a seamless client experience.
- RUM Metric Alert
-
An alert triggered when a client-side performance metric crosses a specified threshold over a particular period of time.
Monitor user experience metrics: page load times, user interactions, JavaScript errors, and so on.
- SLO Alerts
-
SLO Alerts notify you when there is a risk of violating your service-level objectives (SLOs).
- SLO Latency Alert
-
An alert triggered when the latency of a service doesn’t meet the specified targets over a period of time.
Track service response time to ensure that the services meet performance commitments of the SLO.
- SLO Availability Alert
-
An alert triggered when the availability of the service doesn’t meet the specified targets over a period of time.
Monitor service uptime and reliability, to ensure that services satisfy the availability commitments of the SLO.
- SLO Metric Alert
-
An alert triggered when an error rate trace metric crosses a specified threshold over a particular period of time.
Track detailed transaction performance through distributed tracing data to monitor operations and requests.
- Event Alerts
-
Event alerts help identify and address issues before they become major problems. Events are discrete occurrences that can be generated by software or hardware.
- Event Metric Alert
-
An alert triggered when an event metric crosses a threshold over a specified period of time.
Monitor system events and changes: deployments, application performance, security threats,database activity, infrastructure health, business transactions, compliance with regulations, and so on.
