SLA Calculator: Uptime and Downtime

When we talk about 99.9% availability, it sounds impressive. But what does it actually mean? How much downtime does it allow per year, per month, per week? These calculations are essential for defining realistic and measurable SLAs.

The difference between 99% and 99.9% may seem minimal (0.9 points), but in practice it's huge: 3.65 days of allowed downtime versus only 8.77 hours. Each additional "9" divides the acceptable downtime by 10.

This calculator helps you visualize what availability percentages really mean, choose an SLA suited to your needs, and understand the technical implications of your commitments.

Understanding SLAs

Key elements of an availability SLA:

• Availability percentage: The rate of time the service must be operational. 99.9% means the service can be unavailable 0.1% of the time.
• Measurement period: The window over which availability is calculated: monthly, quarterly or annually. Impacts allowed downtime.
• Exclusions: Scheduled maintenance, force majeure events, client-side issues. These times are often excluded from calculation.
• Penalties: Compensation in case of SLA breach. Usually service credits (10-25% of monthly amount).

Availability Levels Table

Allowed downtime for each SLA level:

The Rule of Nines

Each additional "9" divides downtime by 10:

• Two 9s (99%): 3.65 days of downtime per year. Acceptable for non-critical services or development environments.
• Three 9s (99.9%): 8.77 hours of downtime per year. Standard for most professional SaaS applications.
• Four 9s (99.99%): 52 minutes of downtime per year. Required for critical services: payment, healthcare, infrastructure.
• Five 9s (99.999%): 5.26 minutes of downtime per year. Extremely difficult to achieve. Reserved for vital systems.

Calculation Formula

How to calculate allowed downtime:

Allowed downtime = Period x (1 - SLA/100)

Examples:
- 99.9% over 1 year: 365 x 24 x 60 x (1 - 0.999) = 525.6 minutes = 8.77 hours
- 99.5% over 1 month: 30 x 24 x 60 x (1 - 0.995) = 216 minutes = 3.6 hours

This simple formula lets you calculate allowed downtime for any percentage and any period.

Factors Affecting Availability

Elements to consider for achieving your SLA:

• Infrastructure: Server redundancy, load balancing, multi-region. More robust infrastructure enables higher SLA.
• Deployments: Each deployment is a risk. Blue-green deployments and canary releases reduce impact.
• Dependencies: Your SLA cannot be better than your dependencies. A 99.9% service depending on another at 99% = 98.9% max.
• Monitoring: Fast detection = fast resolution. Good monitoring reduces MTTR and improves effective SLA.

Composite SLA

When multiple services are in series, total availability decreases:

Composite SLA = SLA1 x SLA2 x SLA3 x ...

Example:
- Frontend: 99.9%
- Backend: 99.9%
- Database: 99.95%
- Composite SLA: 0.999 x 0.999 x 0.9995 = 99.75%

This is why architectures with many components need very high individual SLAs to maintain an acceptable overall SLA.

Best Practices

Tips for defining and meeting your SLAs:

• Be realistic: Don't promise 99.99% if your infrastructure can't support it. An unmet SLA destroys trust.
• Measure accurately: Use external monitoring (like MoniTao) to measure availability as seen by your users.
• Define exclusions: Be explicit about what's excluded: scheduled maintenance, provider issues, etc.
• Plan for margin: Aim internally for a stricter SLO than the promised SLA. 99.95% internal for a 99.9% SLA.

Frequently Asked Questions

Define Realistic SLAs

An SLA isn't just a marketing number. It's a measurable commitment that must reflect your infrastructure's and team's actual capability. Promise what you can deliver, and exceed expectations.

With MoniTao, you have the tools to measure your actual availability, detect incidents quickly, and prove to your clients that you're meeting your commitments. Start free.

Ready to Sleep Soundly?

Start free, no credit card required.

Create My Account See pricing