NCP-ADS study plan

Diagram the read and write I/O path from a host initiator through fabric, array controller, cache, and back-end media, labeling every hop where latency or data loss can occur.

Compare block, file, and object storage on access semantics, consistency model, and typical workload fit, and justify a choice for three given workload profiles.

Explain RAID 1, 5, 6, and 10 in terms of rebuild cost, write penalty, and usable capacity, and compute usable capacity and fault tolerance for a 12-drive configuration at two drive sizes.

Describe thin provisioning, deduplication, and compression, and state for each where it sits in the I/O path and what it does to capacity accounting.

Differentiate SAN, NAS, DAS, and hyperconverged architectures by protocol, scaling model, and failure domain, and map each to a scenario where it wins.

Trace an NVMe-oF and an iSCSI connection setup step by step, naming the discovery, session, and authentication stages.

Size a storage solution from a given IOPS, capacity, and growth requirement, showing the arithmetic for spindle and SSD tiers.

Identify the single points of failure in a two-fabric, dual-controller array design and describe how multipathing removes each.

Define RPO and RTO for three business scenarios and select a protection method (snapshot, replication, backup, archive) that meets each.

Explain snapshot mechanics (copy-on-write versus redirect-on-write) and describe how snapshot retention affects primary array capacity and performance.

Compare synchronous and asynchronous replication on latency impact, distance limits, and data-loss exposure during failover.

Describe the 3-2-1 rule and its modern extensions, and critique a given backup schedule against it.

Explain immutability, air-gapping, and anomaly detection as ransomware defenses, and state which layer each one protects.

Walk through a restore test from backup media to a running workload, listing verification points at each stage.

Interpret backup job success rates, change rates, and dedupe ratios to spot a failing protection policy.

Distinguish backup, replication, and archival copies by recovery use case and cost profile.

Explain Layer 2 and Layer 3 forwarding, VLAN segmentation, and routing between storage and compute networks, and identify where jumbo frames must be consistent end to end.

Describe spine-leaf topology and oversubscription ratios, and compute east-west bandwidth for a given leaf uplink count.

Compare Type 1 and Type 2 hypervisors, VM resource scheduling, and the effect of CPU overcommit and memory ballooning on performance.

Map a workload to on-premises, IaaS, PaaS, or SaaS using control, portability, and operational burden as criteria.

Write a declarative infrastructure definition (Terraform-style or equivalent) that provisions a network and a storage volume, and explain state, plan, and apply.

Identify idempotency, drift, and secret handling as the three failure modes of automated provisioning, and state a mitigation for each.

Describe a CI/CD pipeline for infrastructure changes including lint, plan review, and gated apply.

Explain how an API-driven storage array or hypervisor is automated, and outline error handling and retry logic with backoff.

Select the right metric type (counter, gauge, histogram) for latency, throughput, queue depth, and error rates, and explain why averages mislead.

Compute percentiles from a small latency sample and interpret tail latency in the context of a storage array.

Build a troubleshooting sequence for a slow-application report that moves from host to network to array to media, naming the check at each layer.

Correlate logs and metrics across layers to isolate whether a bottleneck is compute, fabric, or media.

Re-derive the RAID, oversubscription, and sizing calculations from memory without notes.

Answer a mixed set of practice questions across all six domains and log every wrong answer with its root cause.

Re-teach the three weakest concepts aloud or in writing in under five minutes each.

Finalize a one-page exam-day sheet of formulas, port numbers, and decision rules.