What uptime should we expect, and what does 99.9% mean?
Uptime percentages are easiest to understand as time. 99% allows about seven hours of downtime per month, 99.9% about forty-three minutes and 99.99% about four minutes. Each additional level of availability costs significantly more to achieve.
Around 99.9% is achievable with standard practices: redundant application servers, a managed database with automatic failover, health checks and zero-downtime deployments.
Higher availability requires multi-region deployment, cross-region data replication and fully automated failover that does not depend on human intervention.
The right target depends on the cost of downtime to your business. An internal tool used during office hours has very different requirements from a payment system or clinical application.
When reviewing a published SLA, check whether it excludes scheduled maintenance, how it treats degraded performance and how compensation is calculated.
Related service
Custom SaaS DevelopmentMulti-tenant platforms with billing, administration and access control built in.
Was this helpful?