Service Level Agreements & Objectives
This document outlines Anava's service level agreements (SLAs) and service level objectives (SLOs) for enterprise customers.
Overview
Anava provides enterprise-grade reliability with clearly defined service levels across all platform components.
Availability Targets
Cloud Services
| Service | SLO | SLA |
|---|---|---|
| Events Viewer (SPA) | 99.9% | 99.5% |
| Cloud Functions | 99.9% | 99.5% |
| Firestore Database | 99.99% | 99.9% |
| Firebase Authentication | 99.95% | 99.9% |
Edge Services
| Service | SLO | Description |
|---|---|---|
| MQTT Broker | 99.9% | Device-to-cloud communication |
| Edge Detections | 99.5% | On-camera AI processing |
| Local Failover | 100% | Offline operation capability |
Performance Targets
Response Times
| Operation | P50 | P95 | P99 |
|---|---|---|---|
| API Requests | 100ms | 300ms | 500ms |
| Authentication | 150ms | 400ms | 800ms |
| Real-time Events | 50ms | 150ms | 300ms |
| Image Generation | 5s | 15s | 30s |
| Video Generation | 30s | 60s | 120s |
Throughput
| Metric | Capacity |
|---|---|
| Events per device | 100/minute |
| Concurrent devices | 10,000+ |
| Real-time subscriptions | 1,000 per customer |
| API rate limit | 1,000 req/min per customer |
Data Durability
| Data Type | Durability | Backup Frequency |
|---|---|---|
| Device Configurations | 99.999999999% (11 9s) | Real-time |
| Detection History | 99.999999999% (11 9s) | Real-time |
| Audit Logs | 99.999999999% (11 9s) | Real-time |
| Generated Media | 99.99999999% (10 9s) | Daily |
Recovery Objectives
| Metric | Target | Description |
|---|---|---|
| RTO | 4 hours | Time to restore service |
| RPO | 1 hour | Maximum data loss window |
| MTTR | 30 minutes | Mean time to repair |
Maintenance Windows
Scheduled Maintenance
- Frequency: Monthly
- Duration: Up to 2 hours
- Notice: 7 days minimum
- Window: Weekends, 02:00-06:00 UTC
Emergency Maintenance
- Notice: Best effort (minimum 1 hour)
- Duration: As needed
- Communication: Real-time status updates
Exclusions
SLA does not apply during:
- Scheduled maintenance - Pre-announced windows
- Force majeure - Natural disasters, acts of war
- Customer actions - Self-inflicted issues
- Third-party failures - Upstream provider outages
- Beta features - Explicitly marked as beta
Incident Response
Severity Levels
| Severity | Description | Response Time | Resolution Target |
|---|---|---|---|
| Critical | Service unavailable | 15 minutes | 4 hours |
| High | Major feature impaired | 1 hour | 8 hours |
| Medium | Minor feature impaired | 4 hours | 24 hours |
| Low | Cosmetic or documentation | 24 hours | 5 business days |
Communication
| Severity | Update Frequency | Channels |
|---|---|---|
| Critical | Every 30 minutes | Status page, Email, Phone |
| High | Every 2 hours | Status page, Email |
| Medium | Daily | |
| Low | Resolution only |
Monitoring
Health Endpoints
# Public status
https://status.anava.ai
# API health check
https://api.anava.ai/health
# MQTT broker status
mqtts://mqtt.anava.ai:8883
Metrics Collected
- Request latency (P50, P95, P99)
- Error rates by endpoint
- Device connection count
- MQTT message throughput
- Authentication success rate
Credits
SLA Credit Schedule
| Monthly Uptime | Credit |
|---|---|
| 99.0% - 99.5% | 10% |
| 95.0% - 99.0% | 25% |
| < 95.0% | 50% |
Claiming Credits
- Submit request within 30 days of incident
- Include incident dates and times
- Credits applied to next billing cycle
- Maximum credit: 50% of monthly fees
Compliance
Our SLAs are designed to meet:
- SOC 2 Type II - Security, availability, processing integrity
- ISO 27001 - Information security management
- GDPR - Data protection requirements
Related Documents
- Disaster Recovery - Recovery procedures
- Security Overview - Security architecture
- Compliance - Regulatory compliance
- Troubleshooting - Issue resolution guide