In short
- Definition: The proportion of time a system is operational and accessible when users need it, often defined in a service level agreement (SLA).
- Goal: Keep the service usable during agreed hours and limit the business impact of downtime.
- Example S.M.A.R.T. requirement: 99.9% monthly availability, i.e., no more than 43 minutes of unplanned downtime per month, measured by external monitoring.
1. Definition
As a non-functional requirement, availability describes the share of time a system is operational and usable when people need it [1]. It is usually expressed as a percentage, such as 99.9% over a month [2].
Each extra nine divides the allowed downtime by ten, and usually multiplies the cost.Key takeaway
In ISO/IEC 25010:2023, availability is a sub-characteristic of reliability. It depends on how often the system fails and how quickly it is restored after each failure [3].
Availability must be defined precisely: over which period, during which hours, for which functions, and measured from where. “Up” on a server is not the same as “usable” for a customer.
2. Goal
The goal is to keep the service usable during the hours that matter, and to limit the business impact of downtime to a level agreed in advance.
Without an explicit target, teams either over-engineer at high cost or discover in production that the service is down more than the business can accept.
3. How to Measure It (S.M.A.R.T.)
| Metric | Unit | Typical target | How to measure |
|---|---|---|---|
| Availability | % of time or of successful requests | 99.9% monthly (about 43 minutes of downtime) | External monitoring probes or request success rate |
| Unplanned downtime | Minutes per month | < 43 min | Incident records with start and end times |
| Mean time to recovery (MTTR) | Minutes | < 30 min | Incident management tool |
| Planned maintenance windows | Hours per month | < 4 h, outside business hours | Change calendar |
Rough downtime allowances per 30-day month: 99% is about 7.2 hours, 99.9% about 43 minutes, 99.99% about 4.3 minutes.
4. Example Requirements
Each requirement below is ready to copy into a specification: specific, measurable, achievable, relevant and time-bound.
- E-commerce: “The checkout service shall be available 99.9% of the time each calendar month, measured by external probes every minute, excluding announced maintenance of at most 2 hours on Sunday nights.”
- Internal tool: “The timesheet application shall be available 99% of business hours (Monday to Friday, 7:00 to 19:00) each quarter.”
- API: “At least 99.95% of API requests over every rolling 30 days shall return a non-error response, as measured at the load balancer.”
5. Questions to Ask Stakeholders
- When must the system be available: 24/7, business hours, peak seasons?
- What does one hour of downtime cost in money, reputation or safety?
- Which functions are critical, and which can be unavailable for a while?
- Is planned maintenance downtime acceptable, and when?
- Is there a contractual SLA with customers?
- From where will availability be measured: server, network or user’s point of view?
6. Common Pitfalls
- Asking for 100%. No system achieves it, and chasing it is very costly. Agree on a realistic target and an error budget [4], [5].
- Measuring only server uptime. Users can be blocked while servers are “up”. Measure from the user’s side or by request success rate.
- Forgetting dependencies. A service cannot be more available than the databases and third parties it depends on, unless it degrades gracefully.
7. Related NFRs
- Reliability — how often the system fails; availability also depends on recovery time.
- Fault Tolerance — keeps the service up when components fail.
- Disaster Recovery — restores availability after major incidents.
- Resilience — keeps essential functions running in degraded mode.
- Maintainability — faster repairs mean shorter outages.
8. FAQ
It allows about 43 minutes of downtime in a 30-day month, or about 8.8 hours per year.
An SLO is an internal target. An SLA is a contract with customers, usually with penalties, and is generally set lower than the SLO.
It depends on the definition you agree on. Many SLAs exclude announced maintenance windows, but users still feel them, so keep them short.
9. Acronyms, Initialisms & Terms
Acronyms are abbreviations pronounced as words (like NASA); initialisms are spelled out letter by letter (like FBI). The classification below follows common pronunciation.
Acronyms
| Abbreviation | Stands for | Definition | Ref. |
|---|---|---|---|
| ISO | International Organization for Standardization | Independent international body that publishes standards, such as ISO/IEC 25010 for software quality. Its short name comes from the Greek “isos” (equal) and is said as a word. | [6] |
| SMART | Specific, Measurable, Achievable, Relevant, Time-bound | Criteria for writing a goal or requirement so that anyone can check whether it has been met. | [7] |
Initialisms
| Abbreviation | Stands for | Definition | Ref. |
|---|---|---|---|
| API | Application programming interface | Defined way for one program to request data or services from another. | [8] |
| FAQ | Frequently asked questions | List of common questions with their answers. | [9] |
| IEC | International Electrotechnical Commission | International standards body for electrical and electronic technologies; it publishes joint “ISO/IEC” standards with ISO. | [10] |
| MTTR | Mean time to repair (or recover) | Average time needed to restore a service after a failure. | [11] |
| NFR | Non-functional requirement | Requirement that describes how well a system must work (its qualities, such as speed or security) rather than what it does. | [1] |
| SLA | Service level agreement | Contract that commits a provider to a level of service, with remedies if it is missed. | [12] |
| SLO | Service level objective | Internal, measurable target for a service, such as “99.9% of requests succeed over 30 days”. | [13] |
Terms
Plain-language definitions of technical and project-management terms used in this article.
| Term | Plain-language definition | Ref. |
|---|---|---|
| Database | Organized store of data that software can query and update. | [14] |
| Error budget | Amount of unreliability allowed by an SLO (for example, 0.1% failed requests), which a team can “spend” on changes. | [5] |
| Incident | Unplanned interruption or degradation of a service. | [15] |
| Load balancer | Component that spreads incoming requests across several servers. | [16] |
| Production environment | The live system that real users use, as opposed to test environments. | [17] |
| Stakeholder | Person or group with an interest in the system: users, sponsors, operators, regulators. | [18] |
| Uptime | Time during which a system is working and available. | [19] |
10. Sources and Further Reading
- [1] “Non-functional requirement.” Wikipedia. https://en.wikipedia.org/wiki/Non-functional_requirement (accessed Nov. 18, 2026).
- [2] “High availability.” Wikipedia. https://en.wikipedia.org/wiki/High_availability (accessed Nov. 18, 2026).
- [3] “ISO/IEC 25010 product quality model.” ISO 25000. https://iso25000.com/index.php/en/iso-25000-standards/iso-25010 (accessed Nov. 18, 2026).
- [4] Google. “Site Reliability Engineering.” Google SRE. https://sre.google/sre-book/table-of-contents/ (accessed Nov. 18, 2026).
- [5] Google. “The Site Reliability Workbook, Implementing SLOs.” Google SRE. https://sre.google/workbook/implementing-slos/ (accessed Nov. 18, 2026).
- [6] “International Organization for Standardization.” Wikipedia. https://en.wikipedia.org/wiki/International_Organization_for_Standardization (accessed Nov. 19, 2026).
- [7] “SMART criteria.” Wikipedia. https://en.wikipedia.org/wiki/SMART_criteria (accessed Nov. 19, 2026).
- [8] “API.” Wikipedia. https://en.wikipedia.org/wiki/API (accessed Nov. 19, 2026).
- [9] “FAQ.” Wikipedia. https://en.wikipedia.org/wiki/FAQ (accessed Nov. 19, 2026).
- [10] “International Electrotechnical Commission.” Wikipedia. https://en.wikipedia.org/wiki/International_Electrotechnical_Commission (accessed Nov. 19, 2026).
- [11] “Mean time to repair.” Wikipedia. https://en.wikipedia.org/wiki/Mean_time_to_repair (accessed Nov. 19, 2026).
- [12] “Service-level agreement.” Wikipedia. https://en.wikipedia.org/wiki/Service-level_agreement (accessed Nov. 19, 2026).
- [13] “Service-level objective.” Wikipedia. https://en.wikipedia.org/wiki/Service-level_objective (accessed Nov. 19, 2026).
- [14] “Database.” Wikipedia. https://en.wikipedia.org/wiki/Database (accessed Nov. 19, 2026).
- [15] “Incident management.” Wikipedia. https://en.wikipedia.org/wiki/Incident_management (accessed Nov. 19, 2026).
- [16] “Load balancing (computing).” Wikipedia. https://en.wikipedia.org/wiki/Load_balancing_(computing) (accessed Nov. 19, 2026).
- [17] “Deployment environment.” Wikipedia. https://en.wikipedia.org/wiki/Deployment_environment (accessed Nov. 19, 2026).
- [18] “Project stakeholder.” Wikipedia. https://en.wikipedia.org/wiki/Project_stakeholder (accessed Nov. 19, 2026).
- [19] “Uptime.” Wikipedia. https://en.wikipedia.org/wiki/Uptime (accessed Nov. 19, 2026).
Written by Philippe Huard · LinkedIn · Last updated: November 18, 2026 · Header photo by @rodrigotypo; inline photo by @jontyson, both on Unsplash.




