Applies to paid plans. Free tier is best-effort, no SLA. Enterprise customers may negotiate a higher uptime target via the MSA.
1. Uptime commitment
| Plan | Monthly Uptime % |
|---|---|
| Free | none (best-effort) |
| Pro | 99.9% (≤ 43.2 min downtime / month) |
| Enterprise | 99.95% (≤ 21.6 min) or as negotiated |
Uptime = fraction of 5-minute windows in the calendar month during which the API endpoint at api.genedata.io returns a 2xx response to a properly-authenticated GET /health probe.
2. Excluded events
The following do NOT count against the uptime calculation:
- Scheduled maintenance with at least 7 days' notice
- Force majeure (natural disasters, cyber-attacks, government action)
- Issues caused by Customer (misuse, integration bugs, exceeded quotas)
- Issues in subprocessor services where root cause is theirs (we'll still try, but it's not our SLA breach)
- Beta / preview features (explicitly excluded)
3. Service credits
If we miss the commitment, Customer gets a credit on the next bill:
| Monthly Uptime % | Credit % of monthly fee |
|---|---|
| < commitment but ≥ 99.0% | 10% |
| < 99.0% but ≥ 95.0% | 25% |
| < 95.0% | 50% |
Credits are capped at one month's fees. Credits are the sole remedy for SLA breach. Customer must request the credit within 30 days of the breach.
4. Support response times
| Severity | Definition | Initial response (business hours: M-F 9am-6pm Pacific) | Initial response (24x7, Enterprise only) |
|---|---|---|---|
| P1 | Production down, no workaround | 1 hour | 30 minutes |
| P2 | Production degraded; major feature broken | 4 hours | 2 hours |
| P3 | Non-urgent bug or question | 1 business day | 8 hours |
| P4 | Feature request, documentation, low-priority | 3 business days | 2 business days |
Customer initiates support via in-app ticket or support@genedata.io. Status updates every 4 hours for active P1.
5. Status page
Real-time service health: status.genedata.io. Customers can subscribe to incident notifications by email or RSS.
6. Backups + recovery
| Metric | Target |
|---|---|
| RPO (max acceptable data loss) | 5 minutes for managed-Postgres tier; 24 hours for daily-backup tier |
| RTO (recovery time) | 60 minutes |
| Backup retention | 14 daily / 8 weekly / 12 monthly / 7 yearly |
| Drill cadence | Weekly automated; quarterly manual region-failover |
Implementation: infrastructure/k8s/helm/genedata/templates/dr/, runbook in docs/runbooks/01-disaster-recovery.md.
7. Maintenance windows
Scheduled maintenance is announced at least 7 days in advance, in advance and during the window:
- Preferred window: Saturdays 10pm-2am Pacific
- Status page banner + email to subscribed customers + in-app announcement
Emergency security patches may be applied without the 7-day notice; we'll communicate via the status page in real-time.
8. Security incidents
For incidents that involve unauthorized access to Customer Data:
- We notify the affected customer within 72 hours of confirming the incident
- We provide a written incident report within 30 days of resolution
- We assist Customer in fulfilling its own breach-notification obligations
See docs/runbooks/ for the incident-response process.
9. Communications
- General: support@genedata.io
- Security: security@genedata.io
- Privacy: privacy@genedata.io
- Legal: legal@genedata.io
- Emergency (P1, after-hours, Enterprise): dedicated phone line + PagerDuty rotation in MSA
10. Modification
This SLA may be updated; material changes take effect 30 days after notice, and Customer may terminate if the change is materially adverse and unresolved.