In short
- Definition: The ability of a system to perform its functions without failure for a specified period, often expressed as mean time between failures (MTBF).
- Goal: Give users a system they can depend on.
- Example S.M.A.R.T. requirement: MTBF above 720 hours in production, and fewer than 2 critical incidents per quarter.
1. Definition
As a non-functional requirement, reliability describes how long a system keeps doing its job correctly before something goes wrong. A reliable system does not just work on launch day: it keeps working, day after day, under the conditions it was designed for.
A system “performs specified functions under specified conditions for a specified period of time.”
ISO/IEC 25010:2023, definition of reliability
The international quality model ISO/IEC 25010 splits reliability into four sub-characteristics: faultlessness (working without fault in normal operation), availability (being operational when needed), fault tolerance (working despite faults) and recoverability (restoring data and state after a failure) [1].
Reliability is not the same thing as availability. A service that crashes every hour but restarts in two seconds can show excellent availability while being very unreliable: users still lose their work each time. Reliability is about how often failures happen; availability is about how much of the time the service is up.
2. Goal
Reliability protects trust. Users plan their work around a system they believe will be there and behave the same way tomorrow. Each unexpected failure costs lost transactions, support calls, overtime for the operations team and, over time, customers.
When reliability is not written down as a requirement, it is usually discovered late: in production, during the first busy period, when fixing the architecture is most expensive. Stating a measurable target early lets the team design, test and budget for it.
3. How to Measure It (S.M.A.R.T.)
| Metric | Unit | Typical target | How to measure |
|---|---|---|---|
| Mean time between failures (MTBF) | Hours | > 720 h (about one failure a month or less) | Total operating time divided by the number of failures, from the incident log |
| Critical incidents | Count per quarter | < 2 | Incident management tool (e.g., Jira Service Management), severity 1 and 2 only |
| Request success rate (SLO) | % of requests | ≥ 99.9% over 30 days | Good requests divided by total requests, from monitoring or load balancer logs |
| Error budget consumed | % of budget | < 100% per 30-day window | Share of allowed failures (1 − SLO) already used |
| Mean time to recovery (MTTR) | Minutes | < 30 min | Total downtime divided by the number of incidents |
MTBF and MTTR together give availability: availability = MTBF ÷ (MTBF + MTTR). For components that are replaced rather than repaired, such as a disk, use mean time to failure (MTTF) instead of MTBF [2], [3].
4. Example Requirements
Each requirement below is ready to copy into a specification: specific, measurable, achievable, relevant and time-bound.
- Customer-facing web application: “From go-live, the online ordering service shall achieve a request success rate of at least 99.9% over every rolling 30-day period, and shall have no more than 2 severity-1 incidents per quarter, as measured by the production monitoring platform.”
- Internal business tool: “During the first 6 months after deployment, the expense-approval tool shall run for at least 720 hours on average between failures during business hours, as recorded in the IT incident log.”
- Batch processing: “The nightly payroll export shall complete without failure on at least 99% of scheduled runs per quarter; any failed run shall be detected and reported to the on-call team within 15 minutes.”
5. Questions to Ask Stakeholders
- What happens to the business, in money or reputation, when this system fails for one hour?
- Which functions must never fail, and which can tolerate an occasional error?
- How many failures per month would users accept before losing trust in the system?
- Are there peak periods (month-end, holidays, campaigns) when a failure would be much more costly?
- Do contracts or regulations impose a level of service (SLA) on this system?
- How is a “failure” defined: a crash, a wrong result, a timeout, a lost message?
- Who is notified when a failure happens, and how quickly must they react?
6. Common Pitfalls
- Asking for 100%. No real system is perfectly reliable, and each extra “nine” costs far more than the previous one. Agree on a realistic target and treat the remaining gap as an error budget the team can spend on changes [4], [5].
- Confusing reliability with availability. A high uptime percentage can hide frequent short failures. Measure how often failures happen (MTBF, incident count), not only total uptime.
- Not defining “failure”. Without an agreed definition, every team counts differently and the metric becomes meaningless. Write the definition into the requirement.
7. Related NFRs
- Availability — the share of time the system is up; it depends on both how often it fails and how fast it recovers.
- Fault Tolerance — keeps the service running when a component fails, so faults do not become user-visible failures.
- Resilience — how well the system absorbs disruptions and returns to normal.
- Maintainability — drives MTTR: the easier the system is to repair, the shorter each outage.
- Stability — consistent behavior over time, without crashes or slow degradation.
8. FAQ
Reliability measures how long a system runs without failing; availability measures the share of time it is usable. A system that fails often but restarts quickly can have high availability and poor reliability.
Divide the total operating time by the number of failures in the same period. For example, 2,160 hours of operation with 3 failures gives an MTBF of 720 hours.
It is the amount of unreliability a service is allowed: 1 minus its service level objective. A 99.9% success objective leaves a 0.1% budget, which the team can spend on releases and experiments until it runs out.
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. |
|---|---|---|---|
| FAQ | Frequently asked questions | List of common questions with their answers. | [8] |
| IEC | International Electrotechnical Commission | International standards body for electrical and electronic technologies; it publishes joint “ISO/IEC” standards with ISO. | [9] |
| IT | Information technology | Use of computers, software and networks to store, process and exchange information; also the department that runs them. | [10] |
| MTBF | Mean time between failures | Average operating time between two failures of a repairable system. | [11] |
| MTTF | Mean time to failure | Average operating time before a non-repairable component (such as a disk) fails. | [11] |
| MTTR | Mean time to repair (or recover) | Average time needed to restore a service after a failure. | [12] |
| 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. | [13] |
| SLA | Service level agreement | Contract that commits a provider to a level of service, with remedies if it is missed. | [14] |
| SLO | Service level objective | Internal, measurable target for a service, such as “99.9% of requests succeed over 30 days”. | [15] |
Terms
Plain-language definitions of technical and project-management terms used in this article.
| Term | Plain-language definition | Ref. |
|---|---|---|
| Batch processing | Processing large groups of records together, often at night, rather than one at a time. | [16] |
| Deployment | Installing a new version of software in the environment where people use it. | [17] |
| 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. | [18] |
| Load balancer | Component that spreads incoming requests across several servers. | [19] |
| Log | Time-stamped record of events written by software, used for troubleshooting and audit. | [20] |
| Production environment | The live system that real users use, as opposed to test environments. | [21] |
| Stakeholder | Person or group with an interest in the system: users, sponsors, operators, regulators. | [22] |
| Transaction | Unit of work, such as a payment or an order, that must complete fully or not at all. | [23] |
| Uptime | Time during which a system is working and available. | [24] |
10. Sources and Further Reading
- [1] “ISO/IEC 25010 — Product quality model: reliability.” ISO 25000. https://iso25000.com/index.php/en/iso-25000-standards/iso-25010 (accessed Nov. 18, 2026).
- [2] “Common incident management metrics: MTBF, MTTR, MTTA and MTTF.” Atlassian. https://www.atlassian.com/incident-management/kpis/common-metrics (accessed Nov. 18, 2026).
- [3] “Reliability vs. availability.” Atlassian. https://www.atlassian.com/incident-management/kpis/reliability-vs-availability (accessed Nov. 18, 2026).
- [4] Google. “Site Reliability Engineering, chapter 3 ‘Embracing Risk’ and chapter 4 ‘Service Level Objectives’.” Google SRE. https://sre.google/sre-book/table-of-contents/ (accessed Nov. 18, 2026).
- [5] Google. “The Site Reliability Workbook, chapter 2 ‘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] “FAQ.” Wikipedia. https://en.wikipedia.org/wiki/FAQ (accessed Nov. 19, 2026).
- [9] “International Electrotechnical Commission.” Wikipedia. https://en.wikipedia.org/wiki/International_Electrotechnical_Commission (accessed Nov. 19, 2026).
- [10] “Information technology.” Wikipedia. https://en.wikipedia.org/wiki/Information_technology (accessed Nov. 19, 2026).
- [11] “Mean time between failures.” Wikipedia. https://en.wikipedia.org/wiki/Mean_time_between_failures (accessed Nov. 19, 2026).
- [12] “Mean time to repair.” Wikipedia. https://en.wikipedia.org/wiki/Mean_time_to_repair (accessed Nov. 19, 2026).
- [13] “Non-functional requirement.” Wikipedia. https://en.wikipedia.org/wiki/Non-functional_requirement (accessed Nov. 19, 2026).
- [14] “Service-level agreement.” Wikipedia. https://en.wikipedia.org/wiki/Service-level_agreement (accessed Nov. 19, 2026).
- [15] “Service-level objective.” Wikipedia. https://en.wikipedia.org/wiki/Service-level_objective (accessed Nov. 19, 2026).
- [16] “Batch processing.” Wikipedia. https://en.wikipedia.org/wiki/Batch_processing (accessed Nov. 19, 2026).
- [17] “Software deployment.” Wikipedia. https://en.wikipedia.org/wiki/Software_deployment (accessed Nov. 19, 2026).
- [18] “Incident management.” Wikipedia. https://en.wikipedia.org/wiki/Incident_management (accessed Nov. 19, 2026).
- [19] “Load balancing (computing).” Wikipedia. https://en.wikipedia.org/wiki/Load_balancing_(computing) (accessed Nov. 19, 2026).
- [20] “Logging (computing).” Wikipedia. https://en.wikipedia.org/wiki/Logging_(computing) (accessed Nov. 19, 2026).
- [21] “Deployment environment.” Wikipedia. https://en.wikipedia.org/wiki/Deployment_environment (accessed Nov. 19, 2026).
- [22] “Project stakeholder.” Wikipedia. https://en.wikipedia.org/wiki/Project_stakeholder (accessed Nov. 19, 2026).
- [23] “Database transaction.” Wikipedia. https://en.wikipedia.org/wiki/Database_transaction (accessed Nov. 19, 2026).
- [24] “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 @paulius005; inline photo by @fyze, both on Unsplash.




