Deployment
130 bites tagged Deployment — interview questions with model answers, and 60-second explainers.
Graceful shutdown implementation and zero-downtime deployments?
Listen for SIGTERM, stop accepting new connections, drain in-flight requests, close resources, exit. process lifecycle, signal handling, and connection draining.
dependencies vs devDependencies in production?
Dependencies are needed at runtime, devDependencies only for development. npm install includes both, npm ci --only=production excludes dev. understanding package requirements and deployment optimization.
How does Node.js cluster module enable zero-downtime restarts?
Master-worker architecture, graceful shutdown of old workers, routing new requests to new workers. understanding graceful process replacement. forcing immediate termination of all workers instead of draining connections.
Limits of OTA Updates and Runtime Versions
OTA ships only JS and assets; native changes need a store build; runtime versions gate compatibility. OTA boundaries. assuming any change can ship OTA or ignoring runtime version mismatches.
CI/CD Pipeline for React Native to TestFlight and Play
A triggered workflow installs deps, builds signed artifacts on macOS and Linux runners, uploads via Fastlane to TestFlight and the Play internal track. mobile CI/CD design. storing signing secrets in the repo.
Over-the-Air Updates for JavaScript-Only Fixes
Ship a new JS bundle the app downloads at runtime and applies on next launch, with no store review. OTA update mechanics. thinking OTA can ship native code changes or bypass store policy.
Android App Bundles versus Universal APKs
An AAB is an upload format; Play generates per-device APKs via dynamic delivery, shrinking download size. Play distribution format. thinking an AAB installs directly on a device like an APK.
Managing Per-Environment Configuration in React Native
Use per-environment env files plus build flavors or schemes, inject at build time, keep secrets off the device. environment configuration strategy. hardcoding endpoints or branching on a runtime flag inside one bundle.
iOS Provisioning Profiles and Signing Certificates
The certificate proves developer identity, the App ID names the app, the profile binds certificate, App ID, and devices. iOS code-signing model. confusing the certificate with the profile or treating them as interchangeable.
Zero-downtime blue-green deploys on Kubernetes
Run blue and green deployments, switch a Service or ingress selector to the new color after readiness probes pass, drain old pods gracefully, and handle backward-compatible DB migrations. Safe deploys with no dropped requests.
Rolling update vs blue-green deployment
Rolling replaces instances gradually with minimal extra capacity but mixes versions; blue-green runs two full environments for instant switch and rollback at double the cost. Knowledge of deployment strategies and their costs.
Auto-rollback on failed blue-green cutover
Shift traffic gradually behind a smart router, use deep health checks plus real SLI monitoring, and auto-revert to blue on breach while blue stays warm. Safe cutover design.
Design automated canary analysis scoring
Track the golden-signal SLIs, compare canary to baseline statistically, weight and combine into a score with promote/rollback thresholds. Automated release gating.
Diagnose a degraded canary release
Check statistical significance versus baseline, confirm apples-to-apples comparison, isolate the cause via traces and logs, then weigh the regression against SLO budget. Canary decision-making.
Blue-green deploys with schema migrations
The shared database means both versions hit one schema, so breaking changes must be split into backward-compatible steps via expand-and-contract. Stateful deploy safety.
What is shadow deployment for ML models?
New model receives mirrored live traffic but its predictions are logged, not served; validates real-world behavior and latency without user risk. safe ML rollout strategy.
Automated rollback for a failed model deploy
Detect failure via health and metric checks, automatically route traffic back to the last known-good version (blue-green or canary), keep the registry entry but unpromote, and alert. safe deployment recovery.
CI/CD for microservice-based ML systems
Independent per-service pipelines, contract testing to protect interfaces and schemas, and incremental deploys (canary, blue-green); manage data and model contracts, not just code. ML CI/CD at service granularity.
Staging Environments for ML Pipelines
A staging environment mirrors production so models and pipelines are validated on production-like data and infrastructure before release. It catches drift, integration breaks, and serving regressions early, making promotion to production a safe, repeatable…
PTQ versus QAT for INT8 quantization
PTQ quantizes a trained model with light calibration, fast and cheap but more accuracy loss; QAT simulates quantization during training, higher accuracy but costly. practical quantization trade-offs.
Model quantization benefits and risks
Quantization stores weights and activations in lower precision like INT8 or INT4; benefits are smaller memory and faster, cheaper inference; risk is accuracy loss. deployment-efficiency fundamentals.
StatefulSet vs Deployment: what's the difference?
Deployments treat Pods as interchangeable; StatefulSets give stable ordinal names, stable per-Pod storage via volumeClaimTemplates, and ordered rollout. choosing the right workload controller.
Upgrading a stateful Flink job without losing state
Take a savepoint, stop with drain, deploy new jar, restore from savepoint with matching operator UIDs. stateful stream-job upgrades.
Roll out a design system color change
Update the token source, rebuild outputs, version and publish, bump the app's dependency, verify contrast and visual regressions. end-to-end change management. editing the hex in components or shipping unversioned.
Get Deployment bites daily.
Five a day, five minutes, offline. With quizzes so it sticks.
Open testing — you’ll join as an early tester.