Search the practice set
275 questions, 991 terms and 600 topics in 30 areas.
60 results for “Exactly-Once Semantics”
Idempotent Sink
A destination where writing the same record twice has the same effect as writing it once, which is what makes end-to-end exactly-once achievable at all.
At-Least-Once Delivery
The guarantee that a message will be delivered, possibly more than once — the practical default in every distributed messaging system.
Consumer Group
A set of consumers that cooperatively read one stream, with each partition assigned to exactly one member, so the group collectively processes every message once.
Decision Rights Over Data
Who is entitled to decide what a data element means, who may see it and how long it is kept — the substance of governance once the policy documents are set aside.
Normalisation
Organising a schema so each fact is stored exactly once, removing the update anomalies that duplication creates.
HTTP Method Semantics
The safety, idempotency and cacheability guarantees each HTTP method carries, which clients and intermediaries rely on.
Health Check Semantics
The distinction between liveness, readiness and startup checks, and the failure each is intended to address.
Radar Ring Semantics
The meaning attached to each ring of a technology radar, which determines whether it guides decisions or merely lists opinions.
Backend for Frontend
A separate, narrow backend per client experience, which aggregates and reshapes downstream services for exactly that client's needs.
Compensating Transaction
A business operation that semantically undoes a previously committed step — not a rollback, because the original effect was visible and may not be fully reversible.
Component Library Versioning
Treating a design system as a versioned product with a compatibility policy and a deprecation process, because its consumers cannot all upgrade at once.
Concurrency
The number of operations in progress at once — distinct from parallelism, which is how many are literally executing simultaneously.
Consumer Contract
A machine-readable record of exactly which parts of a provider's response one consumer relies on, generated from that consumer's own tests.
Data Contract
An explicit, versioned, enforced agreement between a data producer and its consumers about schema, semantics, quality and change policy.
Deadline Exceeded
The error returned when a request's overall budget expires — semantically distinct from a per-hop timeout, and a signal that must not be retried blindly.
Device Group Rollout
Applying configuration or firmware to a fleet in ordered cohorts with health checks between them, because a bad change applied to everything at once has no remedy.
Disagree and Commit
A convention in which a participant states their objection, accepts the decision once made, and supports its execution fully.
Embedding
A dense numeric vector representing a piece of content, positioned so that semantically similar content sits nearby.
Evidence Based Approval
Replacing a human judgement about whether a change is safe with a machine-produced record of the checks it passed, assessed once for the class rather than per instance.
Evolutionary Architecture
Designing for guided, incremental change rather than trying to get the structure right once, up front.
Forward Fix
Resolving an incident by shipping a correction rather than reverting, which is the right choice exactly when reverting is not actually available.
Grain Declaration
Stating exactly what one row of a fact table represents, before any column is chosen, because every later decision depends on it.
GraphQL
A query language and runtime where the client specifies exactly which fields it needs, against a typed schema, usually via a single endpoint.
Idempotency
The property that performing an operation many times has the same effect as performing it once.
Log Level Discipline
Consistent semantics for log severity so that levels can be used for routing, alerting and cost control.
Log Schema Consistency
Enforcing the same field names, types and semantics for structured log records across every service, so cross-service queries are possible.
Metric Definition Layer
Business metrics defined once in a versioned, tested place and served to every consuming tool, so the number cannot differ by which tool asked.
Offset Management
How a consumer records its position in a stream, and the decision that determines whether processing is at-least-once or at-most-once.
Outbox Relay
The process that reads pending messages from an outbox table and publishes them to a broker, providing at-least-once delivery with no distributed transaction.
Over-Fetching
Transferring fields the interface will not display, which costs bandwidth and parse time on exactly the devices least able to afford them.
A vendor claims their streaming platform provides exactly-once processing. How do you evaluate the claim?
Ask where the guarantee ends Nearly always at the platform's boundary. Within it, state and offsets commit together, so internal state reflects each input once.
A payments API must guarantee that a network retry never charges a customer twice. Design the mechanism end to end.
Why this is unavoidable A client that times out on a POST cannot know whether the request succeeded. Not retrying risks a lost payment; retrying risks a duplica
Design an order submission API that is safe when the client cannot tell whether its request succeeded. What exactly do you store, and when?
What the interviewer is testing Whether you know that "make it idempotent" is a design with specific failure modes, not a checkbox. The core design The client g
A list endpoint returning 2 million records times out on deep pages and returns 500 for validation failures. Fix both properly.
The pagination problem OFFSET 100000 requires the database to scan and discard 100,000 rows before returning the page, so each page is slower than the last. Off
Finance and Sales report different revenue for the same quarter. Both are convinced they are right. You are asked to fix the data. Where do you start?
Assume neither number is wrong The instinct is to look for a bug. In practice the overwhelming majority of these cases are two defensible definitions with no me
Exactly-Once Semantics
What the phrase really means, where it holds, and the idempotent sink underneath it.
Streaming Data
Windowing, watermarks, late arrivals and exactly-once semantics.
Semantic Layer
Metric definitions held once and served to every tool that asks.
Data Governance & Semantics
General material on ownership, meaning, quality and control of data at enterprise scale.
Accessibility Architecture
Semantics, focus management and announcements designed in rather than audited in.
BI Governance
Dashboard sprawl, certified reports, and the number the board is allowed to see.
Business Glossary
Agreeing what a term means before arguing about which number is right.
CRDTs
Data types that converge without coordination, and the semantics you must accept.
Continuous Controls Monitoring
Testing controls continuously instead of sampling them once a year.
Control Design vs Operation
A control that is well designed and never runs fails exactly like one that is absent.
Data Access Models
Role, attribute and purpose-based access over analytical data, and how they compose.
Data Catalog
Discovery, ownership and technical metadata, and why catalogues go stale.
Data Contracts
Producers committing to schema, semantics and freshness, and breaking builds when they do not.
Data Mesh
Domain ownership, data as a product, self-serve platform, and federated governance.
Data Observability
Freshness, volume, schema and distribution monitoring for pipelines that fail silently.
Data Products
A dataset with an owner, an interface, an SLO, and consumers who can rely on it.
Data Quality Dimensions
Completeness, accuracy, timeliness, consistency, validity and uniqueness as testable claims.
Data Sharing & Clean Rooms
Collaborating on data neither party may hand over, with computation as the interface.
Data Stewardship
The operating model that makes ownership a role rather than a slide.
Flaky Test Management
Quarantine, detection, and the trust a suite loses once red stops meaning broken.
Master Data Management
One authoritative record for a customer or product across systems that each have their own.
Messaging & Queues
Decoupling producer from consumer, and the semantics that come with it.
OpenTelemetry
Instrumenting once against an open standard rather than a vendor agent.