Understanding SLA Uptime and Availability
| SLA Level | Uptime % | Downtime/Year | Downtime/Month | Downtime/Week |
|---|---|---|---|---|
| Two nines | 99% | 3d 15h 36m | 7h 18m | 1h 41m |
| Three nines | 99.9% | 8h 45m | 43m 49s | 10m 4s |
| Four nines | 99.99% | 52m 33s | 4m 22s | 1m 0s |
| Five nines | 99.999% | 5m 15s | 26s | 6s |
| Six nines | 99.9999% | 31s | 2.6s | 0.6s |
| Use Case | Typical SLA | Notes |
|---|---|---|
| Enterprise cloud (AWS, Azure, GCP) | 99.9–99.99% | Core services; region-specific |
| E-commerce platform | 99.9–99.95% | Peak traffic may have stricter SLA |
| Internal business apps | 99–99.9% | Business hours sometimes excluded |
| Healthcare/financial systems | 99.99–99.999% | Regulatory and compliance driven |
| Consumer web apps | 99.5–99.9% | Less strict; users more tolerant |
| Critical infrastructure | 99.999%+ | Telecom, payment processors |
| Revenue/Year | Cost/Hour of Downtime | Cost/Minute | Notes |
|---|---|---|---|
| $1M | ~$114 | ~$1.90 | Small business |
| $10M | ~$1,140 | ~$19 | Mid-market |
| $100M | ~$11,400 | ~$190 | Enterprise |
| $1B | ~$114,000 | ~$1,900 | Large enterprise |
| Amazon (peak) | ~$220,000 | — | Estimated per hour |
| Average enterprise (2023) | — | ~$5,600/min | Gartner estimate |
Service Level Agreements (SLAs) define the expected availability of a service and the consequences of not meeting those expectations. Understanding how uptime percentages translate to actual downtime is crucial for both service providers and consumers. This guide explains SLA calculations, composite availability, and strategies for achieving high availability.
The "Nines" of Availability
Common SLA Tiers
| Availability | Name | Downtime/Year | Downtime/Month |
|---|---|---|---|
| 99% | Two Nines | 3.65 days | 7.31 hours |
| 99.5% | Two and a Half Nines | 1.83 days | 3.65 hours |
| 99.9% | Three Nines | 8.77 hours | 43.8 minutes |
| 99.95% | Three and a Half Nines | 4.38 hours | 21.9 minutes |
| 99.99% | Four Nines | 52.6 minutes | 4.38 minutes |
| 99.999% | Five Nines | 5.26 minutes | 26.3 seconds |
Calculating Downtime
Basic Formula
Downtime = Total Time x (1 - Uptime Percentage)
- Minutes per year: 525,600 (365 days x 24 hours x 60 minutes)
- Minutes per month: 43,800 (30.42 days average)
- Minutes per week: 10,080
- Minutes per day: 1,440
Service Window Considerations
SLAs may apply only during specific service windows:
- 24x7: Full calendar time (8,760 hours/year)
- 24x5: Weekdays only (6,240 hours/year)
- Business hours: 9-5 weekdays (2,080 hours/year)
Composite Availability
Serial Dependencies
When components are in series (all must work):
Composite SLA = SLA1 x SLA2 x SLA3 x ...
Example: Three 99.9% services in series:
- 0.999 x 0.999 x 0.999 = 0.997 (99.7%)
- Results in ~26 hours downtime/year instead of ~8.7 hours
Parallel Dependencies (Redundancy)
When components have redundancy:
Composite SLA = 1 - (1 - SLA1) x (1 - SLA2)
Example: Two 99% services in parallel:
- 1 - (0.01 x 0.01) = 0.9999 (99.99%)
- Redundancy dramatically improves availability
Impact of Dependencies
| Dependencies (99.9% each) | Composite SLA | Downtime/Year |
|---|---|---|
| 1 | 99.9% | 8.77 hours |
| 3 | 99.7% | 26.3 hours |
| 5 | 99.5% | 43.8 hours |
| 10 | 99.0% | 87.6 hours |
Achieving High Availability
Strategies for Each Level
| Target | Requirements | Complexity |
|---|---|---|
| 99% | Basic monitoring, manual recovery | Low |
| 99.9% | Redundancy, automated failover, tested procedures | Medium |
| 99.99% | Multi-AZ, no single points of failure, chaos engineering | High |
| 99.999% | Multi-region, active-active, extensive automation | Very High |
Key Components for HA
- Redundancy: Multiple instances of every component
- Load balancing: Distribute traffic and detect failures
- Health checks: Continuous monitoring of component health
- Auto-scaling: Respond to load and replace failed instances
- Data replication: Synchronous or asynchronous replication
- Automated failover: Minimize human intervention
SLA Financial Terms
Common Credit Structures
| Uptime Level | Typical Credit | Max Credit |
|---|---|---|
| Below target but > 99% | 10% | 10% |
| Below 99% but > 95% | 25% | 25% |
| Below 95% | 50% | 100% |
Exclusions
Most SLAs exclude certain events:
- Scheduled maintenance windows
- Force majeure events
- Customer-caused issues
- External network problems
- Beta or preview features
Measuring Availability
Calculation Methods
- Time-based: (Total time - Downtime) / Total time
- Request-based: Successful requests / Total requests
- Composite: Combined metrics with weights
What Counts as Downtime?
- Complete service unavailability
- Error rates above threshold (e.g., >5%)
- Response times above SLA (e.g., >3 seconds)
- Partial functionality loss (weighted)
Industry Benchmarks
Typical SLAs by Service Type
| Service Type | Typical SLA | Premium SLA |
|---|---|---|
| Cloud Compute (AWS, GCP, Azure) | 99.99% | 99.999% |
| Cloud Storage | 99.9% | 99.99% |
| CDN | 99.9% | 99.99% |
| Database (managed) | 99.95% | 99.99% |
| SaaS Applications | 99.5% | 99.9% |
Best Practices
For Service Providers
- Define clear measurement methodology
- Specify maintenance windows upfront
- Be transparent about composite dependencies
- Provide status pages and incident communication
- Automate credit calculations
For Consumers
- Understand what's actually covered
- Calculate composite SLA for your architecture
- Monitor independently - don't rely only on provider metrics
- Document SLA breaches promptly
- Consider multi-provider strategies for critical services
Conclusion
SLA uptime percentages may seem similar at first glance, but the difference between 99.9% and 99.99% is enormous in practice. Understanding how these numbers translate to real downtime, how composite availability affects your system, and how to architect for your target availability is essential for both service providers and consumers. Use our calculator to explore different SLA scenarios and plan your availability strategy.
Frequently Asked Questions
How accurate are the results?
The SLA Uptime applies a standard formula to your inputs — accuracy depends on how precisely you measure those inputs. For planning and estimation, results are reliable. For high-stakes or professional decisions, cross-check the output with a domain expert or primary source.
Can I use this on mobile?
Yes — the calculator is designed to work on any device. For complex multi-input calculations on small screens, landscape orientation gives more room to see all fields and results simultaneously.