More in Cloud Platforms — page 3
Reducing cross-region data transfer costs
WHAT IT TESTS: data egress cost awareness. OUTLINE: identify cross-region, cross-AZ, and internet egress; co-locate chatty components; add VPC endpoints, CDN caching, and compression. RED FLAG: assuming all transfer is free or ignoring NAT gateway egress.
Lifecycle storage tiering for compliance logs
WHAT IT TESTS: object storage lifecycle design. OUTLINE: hot tier for 7-day query window, lifecycle rules transitioning to infrequent-access then archive, expiration at one year.
Rightsizing an underutilized VM fleet safely
WHAT IT TESTS: data-driven cost optimization. OUTLINE: gather multi-week percentile metrics across CPU, memory, network and disk; pick smaller or right-family types; roll out gradually with monitoring. RED FLAG: resizing on average CPU alone.
On-Demand vs Reserved vs Spot pricing models
WHAT IT TESTS: cloud cost fluency. OUTLINE: On-Demand is flexible but priciest, Reserved trades a 1-3 year commitment for discounts, Spot is cheapest but interruptible. RED FLAG: putting stateful or latency-critical work on Spot.
Attribute cloud costs to teams
WHAT IT TESTS: cost allocation fundamentals. OUTLINE: tag resources with team and project metadata, activate them as cost-allocation tags, group the cost report by that tag, and enforce tagging with policy. RED FLAG: attribution with no tagging.
Design automated cloud cost optimization
WHAT IT TESTS: FinOps automation with guardrails. OUTLINE: target idle resources, oversized instances, orphaned storage, and commitment gaps; act via rightsizing and cleanup; safeguard with tagging, scoping, and approvals.
Monitoring with SLOs and error budgets
WHAT IT TESTS: SRE reliability targets. OUTLINE: define SLIs from the user's view, set SLO targets, derive an error budget, and alert on burn rate rather than raw thresholds. RED FLAG: paging on every CPU blip with no link to user impact.
Strategy for large multi-team IaC projects
WHAT IT TESTS: scaling IaC organizationally. OUTLINE: versioned reusable modules, state split per environment and component, promotion of identical code via variables, and externalized secrets. RED FLAG: one giant shared state or secrets in code.
Diagnose 100% CPU on a managed database
WHAT IT TESTS: structured DB triage. OUTLINE: correlate the spike with deploys and traffic, find top queries via the engine's views, inspect plans for missing indexes, then tune before scaling. RED FLAG: scaling up without finding the offending query.
Enforce a cloud resource compliance policy
WHAT IT TESTS: policy-as-code guardrails. OUTLINE: prevent at creation with org policies or admission checks, detect violations via continuous config scanning, and auto-remediate by stripping the IP or alerting owners. RED FLAG: manual periodic audits only.
State drift in Terraform
WHAT IT TESTS: IaC state versus reality. OUTLINE: drift is when live infrastructure diverges from recorded state, usually via manual console changes; detect with plan or refresh, remediate by re-applying or importing. RED FLAG: hand-editing state.
Centralized logging across microservices
WHAT IT TESTS: aggregating and correlating logs. OUTLINE: ship structured logs from every service into a central searchable store, then propagate a correlation ID through all hops to trace one request. RED FLAG: grepping per-host files with no shared ID.
Automate patching across a VM fleet
WHAT IT TESTS: fleet configuration management. OUTLINE: use a patch or config tool to target by tag, roll out in canaried waves with health checks, and prefer immutable golden images long term. RED FLAG: SSHing into each of 100 boxes by hand.
What is Infrastructure as Code?
WHAT IT TESTS: the case for declarative infrastructure. OUTLINE: IaC defines infra in version-controlled files, giving repeatability, peer review, and drift-free consistency, unlike error-prone console clicks. RED FLAG: calling it mere scripting.
High-throughput serverless stream processing
WHAT IT TESTS: stream design at scale. OUTLINE: partition by key for per-user ordering, use a sharded log with batched consumers for backpressure, and tune batch size and shards for cost. RED FLAG: a global FIFO queue or per-event invocation at 100k/sec.
Serverless functions with a relational database
WHAT IT TESTS: the connection-storm problem. OUTLINE: concurrent function instances each open connections and exhaust the database's bounded pool; fix with a connection proxy or pooler, init-phase reuse, or capped concurrency. RED FLAG: a connection per call.
Distributed tracing for serverless apps
WHAT IT TESTS: end-to-end observability. OUTLINE: propagate a trace context through every hop, instrument with OpenTelemetry or X-Ray, capture spans including the database, and correlate traces with structured logs. RED FLAG: per-service logs, no correlation.
Idempotency in event-driven systems
WHAT IT TESTS: handling at-least-once delivery. OUTLINE: idempotency means repeated processing yields the same end state; it matters because messages get redelivered; achieve it with idempotency keys and conditional writes. RED FLAG: assuming exactly-once.
State machines versus long-running functions
WHAT IT TESTS: orchestration for long workflows. OUTLINE: state machines externalize state, pause for human input without running compute, give built-in retries and audit history, and bill per transition. RED FLAG: a function blocking for days on input.
Purpose and setup of a Dead-Letter Queue
WHAT IT TESTS: handling unprocessable messages. OUTLINE: a DLQ captures messages that repeatedly fail so they neither block the queue nor get lost; configure a redrive policy with a max receive count and alarm on it. RED FLAG: silently dropping failures.