Alerting & On-Call

10 templates, custom PromQL rules, and 7 notification channels — in one place.

How It Works

How Alerting Works

Prometheus-powered alerting with smart deduplication, context enrichment, and 7 notification channels.

EventDetected

Dedup70% noise cut

EnrichContext added

Notify7 channels

Active Alerts — prod-eks-us-east3 firing · 1 pending

3Firing

1Pending

24Resolved (24h)

2Acknowledged

High CPU Usage — api-gateway (prod)92%

Pod Restart Loop — payments-worker (prod)85%

Container OOM Killed — batch-processor (staging)100%

PVC Storage Almost Full — postgres-data (prod)83%

10Templates

7Channels

5Severity Levels

72hAlert History

Lifecycle

From Detection to Resolution

Pending → Firing → Resolved

PromQL condition matches → Pending. Persists for evaluation window → Firing + notifications. Metric recovers → Resolved.

3 states

Acknowledge & Silence

Acknowledge stops repeat notifications. Silence mutes alerts with label matchers — ideal for maintenance windows. Both are audit-logged.

Audit logged

Test & Pause

Test rules against live data before saving. Pause evaluation globally during cluster upgrades to avoid notification floods.

Maintenance-safe

10 Pre-Configured Templates

Alert Templates Ready to Use

Start monitoring immediately. Every threshold, duration, and severity is fully customizable.

Alert Rules — 10 templates across 4 categories

Category	Alert Name	Threshold	Duration	Severity
Resource	High CPU Usage	`> 80%`	5 min	High
Resource	High Memory Usage	`> 85%`	5 min	High
Resource	High Disk Usage	`> 85%`	10 min	High
Resource	Container OOM Killed	`> 0`	1 min	Critical
Availability	Pod Restart Loop	`> 5 restarts`	10 min	Critical
Availability	Pod Not Ready	`= false`	5 min	High
Availability	Deployment Replicas Mismatch	`> 0`	5 min	High
Infra	Node Not Ready	`!= Ready`	5 min	Critical
Infra	PVC Storage Almost Full	`> 80%`	10 min	High
App	High HTTP Error Rate	`> 5%`	5 min	Critical

Full Flexibility

Custom PromQL Rules

Rule Editor

Write any valid PromQL expression with 6 comparison operators: > < >= <= == !=. Set evaluation duration from 1 min to 2 hours, assign severity, and scope to specific clusters or namespaces.

Test Before Deploy

Execute your PromQL query against live Prometheus data before saving. See whether the alert would currently fire — eliminating guesswork and helping you tune thresholds before they reach your channels.

Notification Channels

7 Ways to Get Notified

Severity routing, quiet hours, rate limiting, and test delivery — built into every channel.

Slack

Discord

Teams

PagerDuty

Webhook

Severity Routing

Route Critical + High to PagerDuty, everything to Slack. Filter by cluster and namespace so teams only see their alerts.

Rate Limiting & Quiet Hours

60 notifications/hour per channel. Quiet hours suppress Low/Medium during off-hours. Critical always delivers immediately.

Test Before Save

Send a test notification through the full pipeline — verify webhook, SMTP, bot token, or integration key before going live.

Alert States

Five States, Full Visibility

Firing

Pending

Resolved

Acknowledged

Silenced

Documentation

Go deeper in the SRExpert docs:

Ready to simplify your workflows?

Start with SRExpert today

Connect your first cluster in 5 minutes. See every workload, enforce security policies, and let AI handle the noise.

Replace your monitoring stack, alerting tools, and compliance spreadsheets with one platform built for Kubernetes teams.

Free 30-day trial — no credit card needed.

Start Free Book a Call