Intermediate everything in DevOps & Cloud, page 9
Object storage tiers and lifecycle automation
Tiers trade storage price for retrieval cost and latency; automate with lifecycle rules or intelligent tiering.
Protecting object storage from accidental loss
Enable versioning, MFA delete or object lock, cross-region replication, and least-privilege policies.
Shared file system access across many VMs
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
Check IOPS and throughput against the volume limit, look for burst-credit exhaustion, then move to provisioned IOPS or a larger volume.
The instance metadata service explained
A local endpoint exposing instance identity, region, and temporary role credentials.
Managing state across ephemeral instances
Keep instances stateless, externalize sessions to Redis, data to managed databases, files to object storage.
On-Demand vs Reserved vs Spot pricing
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
Front with a load balancer, define a launch template, an auto-scaling group across AZs, and target-tracking policies.
When hybrid cloud beats public or private
Give a scenario like regulated data plus bursty compute; keep sensitive data and legacy systems on-prem, run scalable or customer-facing workloads in public cloud.
Scalability vs elasticity in the cloud
Scalability is the ability to handle more load by adding capacity; elasticity is automatically adding AND removing capacity in real time to match demand.
IaaS vs PaaS for first cloud migration
IaaS (lift-and-shift) gives control with high ops overhead; PaaS lowers ops but may need refactoring; for a monolith with limited expertise, IaaS lift-and-shift is the lower-risk first step.
Publish/Subscribe Pattern
Publish/subscribe decouples senders from receivers: publishers emit messages to a topic without knowing who consumes them, and subscribers receive messages from topics they care about.
Persistent Volume and Persistent Volume Claim
Kubernetes separates storage provisioning from consumption. A PersistentVolume is a cluster storage resource an admin or driver provisions; a PersistentVolumeClaim is a pod's request for storage.
Mesh control plane vs data plane availability
The control plane configures and distributes policy; the data plane is the sidecars carrying traffic with cached config, so a brief control-plane outage keeps existing traffic flowing but…
How mounted ConfigMap updates propagate to pods
Kubelet refreshes mounted files within roughly a sync period via an atomic symlink swap, but the app must reload on its own; env-var injection never updates.
Image signing with Cosign vs trusting a digest
A digest proves content has not changed but not who produced it; Cosign cryptographically signs the digest so a verified key proves provenance, and policies enforce it at admission.
Docker layers and build cache efficiency
Each instruction makes a content-addressed read-only layer stacked by a union FS; shared layers are pushed/pulled once, and ordering the Dockerfile so volatile steps come last maximizes cache reuse.
Scaling on queue length with the HPA
Expose queue length through an external metrics adapter behind the metrics API, point the HPA at that external metric with a target per pod; KEDA packages this.
Impact of losing etcd quorum
Without quorum etcd goes read-only-ish and the API server cannot persist writes, so scheduling and changes stall, but kubelets keep running existing pods.
Container registries: public vs private
A registry stores and serves versioned image layers by digest; public registries are open and rate-limited, private ones add access control, scanning, signing and network isolation.
We are hiring for this. Every open role lists the topics its interview covers, so you can prepare for the real thing rather than guessing.
See open roles