Resilience & recovery for SY0-701

Exam objectiveExplain the importance of resilience and recovery in security architecture

The short answer

What objective 3.4 asks you to know

Objective 3.4 covers resilience and recovery: staying up when parts fail and coming back quickly when something fails badly, through redundancy, alternate sites, diversity, planning, testing, backups, and power.

Staying up. High availability uses load balancing, which spreads requests across independent servers, or clustering, which makes several servers behave as one system that survives the loss of a member. Platform diversity and multi-cloud systems keep one flaw or outage from becoming your entire outage: redundancy is a spare, and diversity is a spare of a different kind. Power resilience pairs a UPS for immediate, short-term power with a generator for slower-starting, long-term power.

Where and how to recover. Hot, warm, and cold sites differ in equipment and data currency: more readiness means faster recovery and higher cost. Geographic dispersion guards against regional loss. Continuity of operations decides which functions must keep running during a crisis, and capacity planning covers the people, technology, and infrastructure to make that true. Recovery plans are proven by tabletop exercises, simulations, parallel processing, and fail over tests, which trade realism for risk and cost.

Backups. Backups are tested, protected copies kept separate from the original. Choices include onsite or offsite storage, frequency, encryption, snapshots, replication, and journaling. A snapshot on the same storage shares its failures, and replication mirrors damage as well as good data, so neither replaces an independent backup.

Exam scenarios give a recovery requirement and ask which site type, backup method, or test fits; faster recovery generally costs more, so the answer is usually a trade-off.

Written by Knotch · reviewed Nov 15, 2026 · 7 study articles behind this page

Practice the rest

This page is the map. Knotch is where you drill it until it sticks.

  • 7step-by-step lessons with predict-and-reveal and look-alike duels
  • 60exam-style questions with a rationale for every wrong answer
  • 1review queue that brings back what you missed
Start practicing freeFree to start. No card needed.
Tell them apart

Pairs people mix up

  • Load balancing vs. Clustering

    Load balancing distributes work among independent nodes. Clustering coordinates nodes into one system with shared state.

  • Replication vs. Backup

    Replication mirrors data, including damage such as ransomware encryption. A backup preserves history, so earlier versions can be restored.

  • Parallel test vs. Fail over test

    A parallel test runs the standby alongside production and compares results, with no real users on it. A fail over test puts real users on the standby, which carries the highest risk.

  • Redundancy vs. Diversity

    Redundancy adds spare resources so one failure need not stop the service. Diversity builds those resources on different hardware or software, which limits common-mode failure. A design can have both.

The full set of look-alike pairs for this objective is inside Knotch.

Know these cold

Key terms

Hot site
A fully equipped, current recovery site that can take over in minutes, at the highest cost.
Warm site
A partly equipped recovery site that needs data restored and configuration finished, so recovery takes hours to days.
Differential backup
A backup of everything changed since the last full backup, so a restore needs only the full backup and the latest differential.
Incremental backup
A backup of changes since the previous full or incremental backup, so it is small but a restore may need a chain of several sets.
Tabletop exercise
A recovery test in which participants discuss a scenario rather than act it out, as they would in a simulation.
Uninterruptible power supply (UPS)
A battery system that takes over instantly when utility power fails. Runtime depends on battery size and load; most are sized for minutes, enough to shut systems down cleanly or bridge until a generator starts.
Try one

Sample question

Question 3.4-002 · medium

A company takes a full backup every Sunday. To keep restores simple, it wants a restore to need only the Sunday full backup plus ONE more backup set, even if a failure occurs late in the week.

Which method should it use for the weekday backups?

  1. A. Incremental backups
  2. B. Snapshots stored on the same array
  3. C. Continuous replication
  4. D. Differential backups
Show the answer and why
Answer: D, Differential backups

A differential stores everything changed since the last full backup, so a restore needs the full plus the latest differential.

Incremental backups are quick and small but a restore may chain several runs. Differentials tend to grow through the week as more data changes, but a restore needs only the full backup and the latest differential.

Why not A: An incremental holds only what changed since the previous full or incremental run, so a restore may need a chain of several sets.

Why not B: A snapshot depends on the same storage and is not an independent backup.

Why not C: Replication produces a live copy, not point-in-time sets to restore from.

Quick answers

Common questions on this objective

What is the difference between a hot, warm, and cold site?

They differ in equipment and data currency. A hot site is fully equipped and current, a warm site is partly equipped and needs a restore, and a cold site has space, power, and connectivity but little or no equipment; more readiness means faster recovery and higher cost.

What is the difference between an incremental and a differential backup?

An incremental copies changes since the previous full or incremental backup, so restores may need a chain. A differential copies changes since the last full backup, so restores need only two sets.

Is replication a backup?

No. Replication keeps a current standby copy but mirrors damage, including ransomware encryption, and on its own does not give a protected history. A backup preserves earlier versions.

Know it. Now spot it in a scenario.

Start practicing free