More in Cloud Platforms — page 6
Process running but load balancer says unhealthy
WHAT IT TESTS: debugging health checks systematically. OUTLINE: the probe path, port, protocol, or expected status may mismatch; a security group may block the probe; the app may be up but not ready.
Connecting two VPCs privately
WHAT IT TESTS: VPC connectivity options and their scaling tradeoffs. OUTLINE: VPC peering is simple but non-transitive and full-mesh; a Transit Gateway scales hub-and-spoke; PrivateLink exposes one service narrowly.
How cloud DNS resolves a URL to an IP
WHAT IT TESTS: end-to-end DNS resolution flow. OUTLINE: recursive resolver queries root, then TLD, then your authoritative cloud DNS zone; the matching record returns an IP, cached per TTL. RED FLAG: skipping the recursive resolver or caching.
Layer 4 vs Layer 7 load balancers
WHAT IT TESTS: OSI model and load balancer routing knowledge. OUTLINE: L4 routes on IP and TCP/UDP ports fast and protocol-blind; L7 inspects HTTP for host, path, and headers. RED FLAG: claiming L7 is always better.
Public and private subnet VPC design
WHAT IT TESTS: core VPC networking. OUTLINE: web server in a public subnet routed to an internet gateway, database in a private subnet with no inbound from the internet, NAT for outbound. RED FLAG: placing the database in a public subnet or relying only on.
Block storage availability across AZ failure
WHAT IT TESTS: HA design and RTO/RPO trade-offs. OUTLINE: block volumes are AZ-bound, so use snapshots, synchronous replication, or app-level replication, trading cost for lower RTO/RPO.
Object store vs NFS consistency models
WHAT IT TESTS: distributed consistency depth. OUTLINE: S3 gives strong read-after-write per object with no partial updates; NFS offers close-to-open with shared mutable files.
Maximizing object-store throughput for small files
WHAT IT TESTS: object-store scaling internals. OUTLINE: spread keys across prefixes for partition parallelism, batch small files, parallelize and use multipart uploads.
Object storage tiers and lifecycle automation
WHAT IT TESTS: storage cost optimization. OUTLINE: tiers trade storage price for retrieval cost and latency; automate with lifecycle rules or intelligent tiering.
Protecting object storage from accidental loss
WHAT IT TESTS: data durability and recovery design. OUTLINE: enable versioning, MFA delete or object lock, cross-region replication, and least-privilege policies.
Shared file system access across many VMs
WHAT IT TESTS: managed file storage trade-offs. OUTLINE: use a managed NFS service like EFS or Filestore; watch per-operation latency, throughput modes, and metadata-heavy small-file workloads.
Diagnosing high I/O wait on a database volume
WHAT IT TESTS: storage performance diagnosis. OUTLINE: check IOPS and throughput against the volume limit, look for burst-credit exhaustion, then move to provisioned IOPS or a larger volume.
Serving user images securely from object storage
WHAT IT TESTS: object storage plus secure delivery. OUTLINE: store images in a private bucket, serve via a CDN with origin access control, use signed URLs for private content.
Object vs block vs file storage
WHAT IT TESTS: storage fundamentals. OUTLINE: object stores flat keyed blobs over HTTP, block gives raw volumes for one VM, file offers a shared hierarchical mount.
Cost-effective fault-tolerant batch processing
WHAT IT TESTS: batch architecture and cost-fault trade-offs. OUTLINE: Spot fleets across types, a durable work queue, idempotent checkpointed tasks, retries.
Optimizing low-latency VM-to-VM networking
WHAT IT TESTS: networking performance depth. OUTLINE: cluster placement groups, enhanced networking and SR-IOV, larger instances for more bandwidth. RED FLAG: ignoring that tight placement concentrates instances in one AZ, trading availability for latency.
The instance metadata service explained
WHAT IT TESTS: cloud-native configuration and credential handling. OUTLINE: a local endpoint exposing instance identity, region, and temporary role credentials.
Managing state across ephemeral instances
WHAT IT TESTS: stateless design discipline. OUTLINE: keep instances stateless, externalize sessions to Redis, data to managed databases, files to object storage. RED FLAG: storing durable state on local instance disks that vanish when an instance is replaced.
On-Demand vs Reserved vs Spot pricing
WHAT IT TESTS: cost-optimization judgment. OUTLINE: On-Demand for unpredictable bursty work, Reserved or Savings Plans for steady baseline, Spot for interruptible fault-tolerant jobs.
Designing an auto-scaling web tier
WHAT IT TESTS: horizontal scaling design. OUTLINE: front with a load balancer, define a launch template, an auto-scaling group across AZs, and target-tracking policies. RED FLAG: scaling stateful instances or scaling without a load balancer and health checks.