Circuit Breaker
A resilience pattern that temporarily stops calls to a failing dependency so the failure does not keep propagating.
Glossary
Technical glossary
This page collects the English technical terms that appear often across the blog and explains them in a short, practical way so readers can pick up the context quickly before diving deeper into each article.
Architecture
Architecture
CQRS
A pattern that separates commands from queries, allowing write and read models to evolve independently.
Event-Driven
An architecture style in which systems react to published events that represent state changes or business actions.
Idempotency
The property that repeated execution of the same request produces the same result as a single execution.
Kafka
A distributed messaging platform used to publish and consume large event streams reliably in asynchronous and event-driven systems.
Latency
The delay between sending a request and receiving a response, directly affecting perceived performance.
Microservice
An architectural approach that splits functionality into small independently deployable services, trading flexibility for operational complexity.
P95 / P99
Latency percentiles that represent the 95th or 99th percentile of response times, often revealing user experience better than averages.
Retry
A strategy that retries the same request or operation after a transient failure, usually designed together with timeout, backoff, and idempotency.
SLA
Short for Service Level Agreement, a contractual commitment that defines expected service levels such as availability or response time.
SLI
Short for Service Level Indicator, a measured metric that reflects actual system behavior such as success rate or P95 latency.
SLO
Short for Service Level Objective, an internal reliability target for metrics such as availability, latency, or error rate.
Snapshot
A captured state at a specific point in time, often stored to reduce recovery or comparison cost in event sourcing and testing.
Timeout
A time limit after which a request or operation is treated as failed if no response arrives, helping contain cascading failures.
Backend
Backend
API
An interface contract that defines how applications exchange data, including request format, response shape, and authentication rules.
Authentication
The process of verifying who a user is, such as login, token validation, or session checks.
Authorization
The process of deciding what an authenticated user is allowed to do, based on roles and access policies.
Cache
A temporary storage layer used to reuse frequently accessed data quickly. It improves speed but requires careful expiration and invalidation strategies.
GraphQL
An API query model that lets clients ask for the exact data shape they need, with strong flexibility but important schema and execution-cost concerns.
JWT
A signed token format used for authentication by carrying identity and claims, with token lifetime and revocation strategy being critical.
N+1 Query
An inefficient query pattern where one list query is followed by an extra query for each item, often seen in ORM-based applications.
Rate Limit
A policy that limits how many requests can be made within a short period to protect system stability.
Resolver
A function in GraphQL that fetches or computes the value for a specific field from the underlying data source.
TTL
Short for Time To Live, the amount of time cached data is considered valid before it expires or needs to be recalculated.
Frontend
Frontend
Hook
A React API for composing state and side effects in functions, often used to create reusable UI logic boundaries.
Hydration
The process of attaching client-side JavaScript to server-rendered HTML so the page becomes interactive in the browser.
Module Federation
A technique that lets frontend applications share build artifacts at runtime, often used to implement micro-frontends.
Refetch
The act of requesting data again to refresh it with a newer server state.
Server State
Data whose source of truth lives on the server rather than only in the client, usually requiring caching and synchronization strategies.
SSR
Server-Side Rendering, a technique that renders HTML on the server before sending it to the browser.
Stale Time
A server-state setting that defines how long fetched data should be treated as fresh before it becomes eligible for refetching.
DevOps
DevOps
Blue-Green Deployment
A deployment strategy that keeps the old and new environments side by side and switches traffic all at once, enabling fast rollback.
Canary Deployment
A deployment strategy that exposes a new version to a small portion of traffic first in order to reduce rollout risk.
CI/CD
A delivery pipeline that automatically validates and deploys code changes through testing, build, and release steps.
Container
A packaging unit that bundles an application and its runtime environment so it behaves consistently across environments.
Deployment
The process of releasing an application into an environment, including rollout, rollback, versioning, and availability strategies.
DevOps
An operating approach that connects development and operations through shared ownership, automation, and delivery reliability.
HPA
Short for Horizontal Pod Autoscaler, a Kubernetes feature that automatically scales the number of Pods based on CPU or custom metrics.
IaC
Infrastructure as Code, a way to manage infrastructure through source-controlled code instead of manual clicking.
Kubernetes
A container orchestration platform used to deploy and operate applications through a declarative model.
Liveness Probe
A Kubernetes health check that determines whether a container is still alive and should be restarted when repeated failures occur.
Observability
The ability to infer internal system state through logs, metrics, and traces so issues can be diagnosed quickly.
Proxy
An intermediary layer that forwards requests between clients and servers, often used for routing, security, caching, and logging.
Readiness Probe
A Kubernetes health check that determines whether a container is ready to receive traffic and should be included behind a Service.
Replica
The number of identical application instances running the same workload, important for availability and scalability.
Tracing
An observability technique that follows a single request across multiple services to identify bottlenecks and failure points.
Testing
Testing
Test Double
A replacement object used in tests instead of a real dependency, including mocks, stubs, spies, and fakes.