Alerts & Monitoring

Alerts & Monitoring is Breeze’s built-in alerting engine. You define what to watch with rules and templates, set severity to drive triage, monitor availability with network checks, and let correlation group related alerts so your team chases root causes instead of noise.

Alert Rules

Rules define the conditions to watch and the devices to monitor. Each condition specifies a type (metric threshold, status change, or custom), a metric like CPU, memory, disk, or network usage, an operator, a threshold value, and a duration the condition must persist before firing. Target scope ranges from all devices to specific sites, groups, or individual devices, with a configurable cooldown (0–10,080 minutes) and optional auto-resolve when the condition clears.

Alert Templates

Breeze ships five built-in templates covering common scenarios: CPU High, Disk Space Low, Service Stopped, Memory Pressure, and Network Latency High. Templates bundle conditions, severity, target scope, and default cooldown into reusable definitions you can instantiate as rules across the fleet. You can also build custom templates scoped to your organization for environment-specific checks like database connection spikes.

Rules from Templates

Rules are concrete instances of a template, bound to your organization. Create a rule with just a template and a name to inherit every default, or override severity, conditions, targets, and cooldown to tune the same base template per team or site. Rules can be toggled on or off without touching any other field, and every change is written to an audit trail.

Severity Levels & Lifecycle

Every alert carries one of five severity levels (Critical, High, Medium, Low, and Info) that control how it surfaces in the dashboard and how routing handles it. Alerts move through a clear lifecycle: Active, Acknowledged, Resolved, or Suppressed, so technicians can claim an alert, record a resolution note, or silence a noisy condition during planned maintenance.

Alert Correlation

When many related alerts fire at once, such as a network outage knocking dozens of devices offline, the correlation engine groups them automatically. It surfaces the most likely root cause with a confidence score, classifies related alerts as causal, symptom, or duplicate, and presents a chronological timeline. Correlation groups carry an overall score and an optional root-cause hint to cut alert fatigue.

Network Monitors

Network Monitors run recurring availability checks across four types: ICMP ping, TCP port, HTTP/endpoint, and DNS resolution. Checks dispatch through a connected agent on a polling interval from 10 seconds to 24 hours, recording response times and tracking each monitor through unknown, online, degraded, and offline states. Per-monitor alert rules fire on conditions like offline status, degraded performance, response time over a threshold, or consecutive failures.