Alerting Overview

Alerting adds macro-level, derived metrics to the Cisco Provider Connectivity Assurance (formerly Skylight performance analytics) platform. This topic is part of a series, with additional articles as use cases expand. For more information, see Alert Policies.

Overview

Alerting provides new entry points into the Cisco Provider Connectivity Assurance data set:

• Preemptive identification of issues.
• Metric noise reduction.
• Macro trending.
• Alerting on third-party data.
• Upstream notification (via SNMP).

Cisco Provider Connectivity Assurance treats alerts as a peer data source to ingested metrics. Alert-based metrics such as Alerts raised can be used alongside typical metrics such as Delay or Traffic in dashboards, reports, and analysis.

Alert Policies

Alerts are derived from conditions set on metrics within filtered data sets.

• Session-based alerts support all categorical metadata for filtering and one ingested metric per policy.
• Capture-based alerts support application- or zone-based filtering and multiple ingested metrics per policy.

Policies specify raise and clear conditions for defined time frames and metric thresholds.

Alert Conditions

You can configure two types of alert conditions:

• Sustained violations. Raises an alert when a metric breaches a threshold for the entire interval. The minimum interval is 1 minute.
• Percent of interval in violation. Raises an alert when a metric breaches a threshold for a specified percentage of the interval. The minimum interval is 1 minute.

Alerts as a Data Source

Because policies can produce data only under defined conditions, alerts help reduce noise (data appears only when underlying metrics are significant) and improve trending (focused data supports clearer trends).

Use alert metrics in the same places as traditional metrics. Treat alert metrics as supplementary, higher-order data. After an alert identifies where to investigate, use the underlying source metrics for detailed analysis.

Further Discussion

To determine how alerts can help your use cases, review the articles in this series or contact the Customer Success Team. As we expand alerting use cases in Provider Connectivity Assurance, we welcome input to help shape the product.