Terminology
2240 terms, tools, patterns and metrics an architect is expected to use precisely. Each one gets a short explanation of what it is, and — where it matters — what it is commonly confused with. Search filters as you type; the column headers sort.
All areas2240
Architecture Fundamentals77
Distributed Systems107
Data Architecture110
Cloud Architecture94
Networking88
API & Integration Architecture82
Reliability & Resilience80
Observability75
Performance & Capacity Engineering72
Security Architecture86
Cost Architecture & FinOps71
Business Architecture67
Architecture Communication67
Enterprise Architecture66
Legacy Modernization71
AI-Era Architecture74
Software Architecture & Engineering71
Architecture Patterns71
Architecture Decision-Making63
The Architect's Meta-Skills61
Delivery & Release Engineering66
Platform Engineering & Developer Experience70
Testing & Quality Architecture71
Data Platform Architecture69
Streaming & Real-Time Data69
Data Governance & Semantics74
Frontend & Experience Architecture66
Edge, Mobile & IoT68
Regulatory & Data Protection Architecture65
Assurance, Audit & Model Risk69
2240 terms shown.
| Term | Kind | Topic | What it is |
|---|---|---|---|
| Training-Serving Skew Online-Offline Skew, Feature Skew | concept | ML Platform | A divergence between the features a model was trained on and the features computed at serving time - producing a model that performs well offline and worse in production, with nothing erroring. |
| Transaction Log Stream | concept | CDC to Stream | Change events derived from a database's write-ahead log — faithful to the table's mutations and to its internal model rather than to the business's. |
| Transactional Audit Emission Audit in the Same Transaction, Complete Audit Guarantee | practice | Auditability | Writing the audit record in the same transaction as the state change it describes, so that a crash cannot produce a change with no record - the property that distinguishes an audit trail from application logging. |
| Transactional Outbox Outbox Table, Reliable Event Publication, Store-and-Forward Events | pattern | Outbox | Writing an event into a table within the same transaction as the business change, then publishing it separately - so the event exists if and only if the change committed, which no ordering of two independent w… |
| Transactional Outbox Outbox Pattern | pattern | Outbox | Writing an event into the same database transaction as the state change, then relaying it, to eliminate the dual-write problem. |
| Transactions and Isolation ACID Isolation Levels | concept | Transactions & Isolation | What a database guarantees when concurrent transactions touch the same data — and the anomalies each level permits. |
| Transfer Mechanism | concept | Cross-Border Transfer | The specific legal instrument permitting personal data to leave a jurisdiction, which must exist per flow and which architecture must make identifiable. |
| Transformation Framework ELT Framework, SQL Transformation Tool | tool | Transformation Frameworks | A tool that turns warehouse transformations into version-controlled, tested, dependency-aware code rather than a collection of scheduled scripts. |
| Trigger Condition Revisit Trigger, Decision Tripwire, Monitorable Assumption | practice | Deciding Under Uncertainty | A specific, observable condition recorded alongside a decision that indicates it should be reconsidered - converting a point-in-time judgement into a monitorable position. |
| Trough-Based Commitment Baseline Reservation, Commit to the Floor | practice | Reserved & Committed Capacity | Sizing reserved capacity to the quietest period rather than the average, because only the trough is capacity you are certain to use for the whole commitment term. |
| Trunk-Based Development TBD | practice | Branching Models | All developers integrate small changes into a single shared branch at least daily, keeping it always releasable. |
| Trunk-Based Development in Practice | practice | Trunk-Based Development | Everyone integrates to one shared branch at least daily, with incomplete work hidden behind flags rather than isolated in branches. |
| Trust Boundary Classification Where mTLS Is Required, Encryption Scope | practice | TLS & Certificates | Writing down which network segments count as trust boundaries, so that mutual TLS is applied where it changes the security posture rather than uniformly or by intuition. |
| Trust Boundary Diagram | practice | Communicating Threat Models | A diagram marking where data crosses between zones of differing trust, which is the structure a threat model is built on and the form security findings are best communicated in. |
| Try-Confirm-Cancel TCC, Reservation Pattern | pattern | Distributed Transactions | A three-phase distributed transaction where each participant first reserves resources, and a coordinator then confirms or cancels all reservations. |
| TSB 2018: A Big-Bang Migration That Was Not Reversible TSB IT Migration | case-study | Cutover Planning | A UK bank moved millions of customer accounts to a new platform in a single weekend cutover, and the failure locked customers out for weeks. |
| Twitter Timelines: Fan-Out on Write, Except for Celebrities Timeline Fan-Out, Hybrid Fan-Out | case-study | Materialized Views | Precomputing every user's timeline at write time is fast to read and impossible for accounts with millions of followers, so the two approaches are combined. |
| Twitter's Timeline Fan-Out | case-study | Performance & Capacity | Twitter precomputes each user's timeline at write time but handles very-high-follower accounts at read time, because neither strategy alone survives both ends of the distribution. |
| Two-Phase Commit 2PC, XA | protocol | Distributed Systems | A blocking protocol for atomic commit across several resources: a coordinator asks all participants to prepare, then tells them all to commit or abort. |
| Two-Phase Locking 2PL | protocol | Transactions & Isolation | The concurrency control protocol behind serializable isolation — acquire locks in a growing phase, release only in a shrinking phase, never interleaving the two. |
| Two-Pizza Team | concept | Business Architecture | Amazon's heuristic that a team should be small enough to be fed by two pizzas, and — the substantive part — should own its service end to end. |
| Two-Stage Retrieval | pattern | Reranking | Retrieving a broad candidate set cheaply and then reordering it with an expensive, more accurate model. |
| Two-Way Door | concept | Reversibility | A decision that can be undone cheaply, and which therefore warrants a fast decision by the people closest to the work rather than extensive analysis. |
| Uber DOMA: Structure Above Microservices Domain-Oriented Microservice Architecture | case-study | Microservices | Having reached thousands of microservices, Uber introduced domains and layers above them because independent services had recreated the coordination problem they were meant to solve. |
| Uber H3: Hexagonal Spatial Indexing H3, Hexagonal Hierarchical Index | case-study | Indexing | Uber built and open-sourced a hexagonal grid index because uniform neighbour distance matters when you are analysing supply and demand across space. |
| Uber's Domain-Oriented Microservice Architecture DOMA | case-study | Software Architecture | After growing to roughly 2,200 microservices, Uber grouped them into domains behind gateways with strict dependency layering, to recover the comprehensibility that fine-grained decomposition had cost. |
| Uber's H3 Spatial Index | case-study | Data Architecture | Uber indexes the world with hexagons rather than squares, because uniform neighbour distance makes supply, demand and pricing computations correct as well as fast. |
| Uber: Choosing a Database for Write Amplification Uber Postgres to MySQL | case-study | Technology Selection | Uber publicly documented moving from Postgres to MySQL for their workload, and the reasoning is a useful lesson in evaluating a datastore against your actual access pattern. |
| Ubiquitous Language | practice | Domain-Driven Design | A shared vocabulary used identically by domain experts and in the code, so that translation between business and implementation is unnecessary. |
| UI Monorepo Strategy | practice | UI Monorepo Strategy | Whether frontend code lives in one repository or many — a trade between atomic change and independent release. |
| Unbounded Dataset | concept | Streaming & Real-Time Data | Input with no known end, which removes the option of waiting for completeness and forces every aggregate to be provisional. |
| Unbundled Event Store Disaggregated Telemetry Store, Object-Storage Column Store | pattern | Real-Time Analytical Stores | A telemetry store that separates writing, compaction and querying into independently scaled services over shared object storage and a metadata store, so retention cost stops being tied to the size of the query… |
| Unenforced Standard Paper Standard, Decorative Governance | concept | Influence Without Authority | A written rule with no mechanism behind it - an anti-pattern whose cost is paid per decision in argument and divergence work while compliance stays partial and uniformity is never achieved. |
| Unfunded Technical Consequence Business-Model Debt, Structural Liability, Implied Requirement | concept | Business Capabilities | A computational or operational difficulty implied by a business-model decision but never costed, because the decision was evaluated on commercial merits in a conversation the technical consequence was not part of. |
| Unintentional Transfer Incidental Transfer, Shadow Data Flow | concept | Cross-Border Transfer | Personal data crossing a jurisdictional boundary through a path nobody classified as a transfer - telemetry, support access, backups, or a processor's staff - which is what audits actually find. |
| Unit Cost Ceiling Cost per Transaction Limit, Infrastructure Cost Budget per Order | metric | Business Understanding | The infrastructure cost per transaction that the business model can bear, derived from contribution margin and fixed before design, which decides how much work a request is allowed to do. |
| Unit Economics | metric | Unit Economics | Infrastructure cost expressed per unit of business value delivered, which reveals efficiency trends that absolute spend cannot. |
| Unit Economics of a System | metric | Unit Economics | The cost to serve one unit of business value — a customer, an order, a request — which determines whether a system's economics improve or worsen with scale. |
| Universal Scalability Law USL, Gunther's Law | concept | Horizontal vs Vertical Scaling | A model showing that throughput rises with concurrency, flattens due to contention, and then falls due to coherency costs. |
| Unmanaged Change Surface Out-of-Band Change Path, Ungated Change | concept | Deployment Gates | Every mechanism that can alter production without passing through the deployment pipeline - so the organisation's rollout controls apply to a fraction of its changes and the incidents come from the rest. |
| Unmerge Capability Merge Reversal, Split-Back | pattern | Master Data Management | The designed ability to reverse an incorrect entity merge and restore both original records, which requires that source keys and pre-merge attribute values were never destroyed. |
| Unreachable Node | concept | Edge, Mobile & IoT | A component you cannot log into, restart or inspect, which removes every operational technique that assumes access to the running system. |
| Untrusted Build Step Third-Party CI Step, Curl-Pipe-Shell Step | concept | Pipeline Secrets | Any pipeline step whose code you do not control and did not review - a downloaded script, a third-party action, a plugin - which runs with the whole job's credentials and network access. |
| Untrusted Client Timestamp Device Time Is Not Authoritative, Client Clock Assumption | concept | Mobile App Architecture | The recognition that a timestamp set by a device is user-editable input rather than a fact - so reporting periods, ordering and de-duplication must not depend on it. |
| Update Cadence Commitment Next-Update Promise, Incident Comms Rhythm | practice | Presenting to Executives | Promising when the next update will arrive rather than when the problem will be fixed, so the one commitment made during an incident is one the team can actually keep. |
| Update Long Tail Version Distribution Tail, Fleet Stratification | concept | OTA Updates | The persistent stratum of devices running old firmware because they were offline during rollouts - which grows across releases and is measured by version distribution rather than by rollout percentage. |
| Usage Telemetry Who Calls This, Consumer Visibility | practice | Platform Telemetry | Per-capability, per-consumer usage data about a platform's own surfaces - the precondition for every deprecation, roadmap and change-impact decision a platform team makes. |
| Usage-Based Access Review Last-Used Entitlement Review, Evidence-Based Access Recertification | practice | Identity & Access Management | Recertifying entitlements against observed use rather than against manager approval, so unused permissions are removed by default instead of being re-approved by someone who cannot evaluate them. |
| Use Case Risk Classification | practice | AI Risk Tiering | Assigning an AI application to a risk tier based on the consequence of it being wrong, which then determines the obligations that apply. |
| USE Method | practice | Observability | For every resource, track Utilisation, Saturation and Errors — the resource-centric complement to request-centric monitoring. |
| Use-Time Enforcement Check at the Point of Use, Consent as a Service | concept | Consent Architecture | Checking permission at the moment data is used rather than at the moment it is collected, because a copied permission flag is stale the moment it is made. |
| User Journey SLO Critical User Journey, CUJ | practice | SLI, SLO & SLA | Defining reliability targets around what a user is trying to accomplish, measured where the user is, rather than around per-request success rates at the server. |
| Utilisation and Queueing Delay | concept | Queueing Theory | The non-linear relationship by which waiting time grows as utilisation approaches one, explaining why systems degrade suddenly rather than gradually. |
| Utilisation Break-Even Break-Even Utilisation, Execution Model Crossover | metric | Serverless vs Containers | The sustained busy fraction at which always-on compute becomes cheaper than per-invocation billing, which turns the serverless-or-containers argument into arithmetic instead of preference. |
| Utilisation Floor | metric | Rightsizing | The minimum utilisation a resource should sustain before it is considered over-provisioned, set with enough margin for failure and peak. |
| Utilisation Target | practice | Queueing Theory | The operating point chosen from a latency requirement rather than from cost efficiency, because queueing delay rises non-linearly as utilisation approaches saturation. |
| Utility Tree | practice | Quality Attributes | A structured decomposition of quality attributes into concrete, prioritised scenarios, used to focus architectural analysis on what actually matters. |
| Vacuous Pass Assertion-Free Test, Empty-Set Pass, Silent Test | concept | Testing & Quality Architecture | A test that passes because it verified nothing rather than because the code is correct - an assertion that never ran, a loop over an empty collection, a mock that could not fail - which is worse than having no… |
| Validated Envelope Model Use Envelope, Validation Boundary | concept | Model Risk Management | The explicit conditions a model's validation actually covers - input schema, population, feature sources, thresholds, upstream versions - outside which the sign-off does not hold and the serving path must refu… |
| Value Stream | concept | Business Architecture | The end-to-end sequence of activities that delivers a result to a customer, viewed across whatever departments and systems it happens to cross. |
Nothing on this page matches. Search the whole glossary.