Your one-stop platform to prepare, practice and ace your interviews.
Loading your questions…
18 hand-picked General interview questions with detailed answers. Open the interactive version above to search, filter by difficulty, run code, bookmark questions and track your progress.
GET read, POST create, PUT replace, PATCH partial update, DELETE remove. GET/PUT/DELETE are idempotent; POST is not.
Status codes: 200 OK, 201 Created, 204 No Content, 400 bad request, 401 unauthenticated, 403 forbidden, 404 not found, 500 server error.
User logs in → server returns a signed JWT (header.payload.signature). The client stores it and sends it as Authorization: Bearer <token>. The server verifies the signature and expiry on each request — no server-side session needed (stateless).
CORS (Cross-Origin Resource Sharing) is a browser rule that blocks requests to a different origin unless the server opts in with Access-Control-Allow-Origin headers. Fix it on the server; in dev you can use a proxy.
HTTPS is HTTP over TLS, which encrypts traffic, verifies the server's identity via certificates, and prevents tampering. It's required for PWAs, service workers, and modern browser features.
GET — params in the URL, cacheable, idempotent, no body; for reads. POST — data in the body, not cached, not idempotent; for creating/side-effecting operations. Never send sensitive data in a GET query string.
Authentication — proving who you are (login). Authorization — what you're allowed to do (roles/permissions). AuthN comes first, then AuthZ.
Servers send caching headers: Cache-Control (max-age, no-cache), ETag/Last-Modified for revalidation. The browser reuses cached responses until they expire, then revalidates. CDNs cache at the edge for speed.
Cache-Control: max-age=31536000, immutableBig-O describes how runtime/space grows with input size (worst case). Common: O(1) constant, O(log n) binary search, O(n) linear scan, O(n log n) good sorts, O(n²) nested loops.
Scrum works in fixed sprints (1–4 weeks). Ceremonies: sprint planning, daily stand-up, review (demo), and retrospective. Roles: Product Owner (backlog), Scrum Master (process), and the development team.
A reverse proxy sits in front of your servers, accepting client traffic and forwarding it inward (clients never see the backends). It centralizes: TLS termination, compression, caching, rate limiting, path routing (/api → backend, / → static Angular files).
A load balancer is a reverse proxy distributing across multiple instances — round robin, least-connections, IP-hash — with health checks ejecting dead backends. (Forward proxy = the mirror image: sits in front of clients.)
server {
listen 443 ssl;
location /api/ { proxy_pass http://backend:8080/; }
location / { root /usr/share/nginx/html; try_files $uri /index.html; }
}
upstream backend { server app1:8080; server app2:8080; } # least_conn; health checksA user story is a small requirement from the user's viewpoint: As a <role> I want <capability> so that <benefit> — a placeholder for a conversation, not a spec.
INVEST: Independent (schedulable alone), Negotiable, Valuable (user-visible benefit), Estimable, Small (fits a sprint), Testable.
Acceptance criteria — the concrete conditions that make it done, often Given/When/Then; they become the test cases and kill "that's not what I meant" at demo time.
Story: As a customer I want to filter orders by date so I can find a past order.
Acceptance criteria:
Given orders exist, When I pick a date range, Then only orders in it show
Given no orders match, Then an empty state with a clear-filters action shows
Date range max 1 year; invalid range shows inline validationStory points measure relative effort+complexity+uncertainty (Fibonacci 1,2,3,5,8...), not hours — a 5 is 'about like that other 5', which dodges human badness at absolute estimates.
Planning poker: everyone reveals an estimate simultaneously (no anchoring); big splits trigger discussion — the outliers usually know something. T-shirt sizing (S/M/L) for coarse roadmap-level sizing.
Velocity — points completed per sprint, averaged; it's the team's empirical capacity for forecasting, never a performance KPI. Burndown — remaining work vs time in the sprint; a flat line mid-sprint = blocked or bad breakdown.
DoR — a story may enter a sprint only when: clear acceptance criteria, dependencies identified, estimated, testable. Protects the sprint from half-baked work.
DoD — 'done' means: code + review + tests passing, acceptance criteria met, merged, deployed to the agreed environment, docs updated. One shared definition kills "done but not done-done".
Mid-sprint changes: the sprint goal is protected — genuine urgencies (prod incident) swap something out of equal size rather than piling on; everything else goes to the backlog for next planning. Constant mid-sprint churn is a signal to shorten sprints or move to Kanban.
Many teams run Scrumban — sprint cadence for planning/retros, WIP limits and pull for daily flow.
A repeatable pipeline:
Performance engineering — treating performance as a requirement through the lifecycle, not a post-hoc fix: set numeric targets (p95 latency, throughput), test against them (JMeter/Gatling, Lighthouse for frontend), profile bottlenecks (APM, EXPLAIN plans, bundle analysis), fix, and guard in CI (perf budgets) so regressions fail the build.