Overview Warrn’s alert management system provides comprehensive incident tracking and resolution capabilities with AI-powered triage, real-time updates, and collaborative features for modern DevOps teams.
Core Features Alert States & Lifecycle Alerts in Warrn follow a structured lifecycle with clear state transitions:
type AlertStatus = "open" | "acknowledged" | "resolved" ; type WarrnStatus = "pending" | "triage_in_progress" | "triage_success" | "triage_failed" ;
Alert Creation
Alerts can be created manually or automatically collected from external monitoring systems.
Acknowledgment
Team members can acknowledge alerts to indicate they’re being investigated.
AI Triage
Trigger intelligent triage for automated analysis and resolution recommendations.
Resolution
Mark alerts as resolved once the underlying issue is fixed.
AI-Powered Triage Warrn includes an advanced AI triage system that automatically analyzes alerts and provides intelligent recommendations.
const handleAutoTriage = async ( alertId : string ) => { setActiveTriageId ( alertId ); // Start AI triage process await triageAlertMutation . mutateAsync ( alertId ); // Monitor progress with smart polling const pollInterval = setInterval ( async () => { const updatedAlert = await refetchAlerts (); if ( updatedAlert ?. warrn_status === 'triage_success' ) { // Triage completed successfully toast . success ( "Agentic Triage completed successfully!" ); } }, 5000 ); };
Real-time Features The alert system includes smart real-time capabilities:
Active Polling : 15-second intervals when tab is activeBackground Polling : 1-minute intervals when tab is hiddenExponential Backoff : Up to 5-minute maximum intervalsPage Visibility API : Automatic adjustment based on user focus
useSmartPolling ({ onPoll : async () => { await refetchAlerts (); return true ; // Indicates successful update }, activeInterval: 15000 , // 15 seconds when active backgroundInterval: 60000 , // 1 minute when hidden maxBackgroundInterval: 300000 , // 5 minute maximum enabled: ! isLoading });
Alert Filtering & Search Filter Options Status Filters
All alerts
Not resolved (open + acknowledged)
Open only
Acknowledged only
Resolved only
Severity Filters
All severities
Critical
High
Medium
Low
Search Capabilities The alert search supports multiple fields for comprehensive filtering:
const filteredAlerts = alerts . filter (( alert ) => { const matchesSearch = alert . name . toLowerCase (). includes ( searchQuery . toLowerCase ()) || alert . service_name . toLowerCase (). includes ( searchQuery . toLowerCase ()) || alert . team_name . toLowerCase (). includes ( searchQuery . toLowerCase ()); return matchesSearch && matchesStatus && matchesSeverity ; });
Alert Actions Primary Actions
Acknowledge Mark alert as acknowledged to indicate investigation has started.
Auto Triage Trigger AI-powered analysis for intelligent resolution recommendations.
Resolve Mark alert as resolved when the underlying issue is fixed.
Bulk Operations Warrn supports efficient bulk operations for managing multiple alerts:
Bulk Acknowledge : Acknowledge multiple alerts simultaneouslyBulk Resolve : Resolve multiple alerts at onceBulk Assignment : Assign multiple alerts to team members
Each alert includes a collaborative comment system for team communication:
Alert Detail View The alert detail view provides comprehensive information for incident investigation:
Integration Patterns Multi-tenant Architecture All alert operations respect multi-tenant boundaries and ensure data isolation between organizations.
// Example: Alert creation respects organization context const createAlert = async ( alertData : CreateAlertData ) => { // Organization ID is automatically injected based on user context const response = await api . post ( '/alerts' , { ... alertData , organization_id: getCurrentOrganization (). id }); return response . data ; };
Optimistic Updates Warrn implements optimistic updates for improved user experience:
const acknowledgeAlertMutation = useMutation ({ mutationFn: acknowledgeAlert , onMutate : async ( alertId ) => { // Optimistically update the UI immediately queryClient . setQueryData ([ 'alerts' ], ( old ) => old ?. map ( alert => alert . id === alertId ? { ... alert , status: 'acknowledged' } : alert ) ); }, onError : ( error , alertId , context ) => { // Revert optimistic update on error queryClient . invalidateQueries ([ 'alerts' ]); } });
Best Practices
Smart Polling
Use the built-in smart polling instead of aggressive real-time updates to conserve resources.
Pagination
Implement pagination for large alert datasets to maintain performance.
Filtering
Apply filters at the API level when possible to reduce data transfer.
Caching
Leverage TanStack Query’s caching for entity mappings and static data.
User Experience
Loading States : Always show appropriate loading indicators during operationsError Handling : Provide clear error messages with actionable guidanceKeyboard Shortcuts : Implement shortcuts for power users (Ctrl+K for search)Progressive Enhancement : Ensure core functionality works without JavaScript
API Integration Key Endpoints # List alerts with filtering GET /api/alerts?status=open & severity = critical # Get alert details GET /api/alerts/{id} # Update alert status PATCH /api/alerts/{id} { "status" : "acknowledged" } # Trigger AI triage POST /api/alerts/{id}/triage # Add comment POST /api/alerts/{id}/comments { "content" : "Investigation started", "type" : "comment" }
See the Alerts API documentation for complete endpoint details and examples. 
