Objective 3.1High11 min

Architecture Considerations

Security architecture trade-offs including availability, resilience, cost, responsiveness, scalability, RTO/RPO, risk transference, and deployment ease. Understanding these factors is essential for designing secure and practical systems.

Understanding Architecture Considerations

Security architecture requires balancing multiple considerations—availability, cost, performance, and resilience. Understanding these trade-offs is essential for designing systems that are both secure and practical.

Key architecture considerations:Availability — System uptime requirements • Resilience — Ability to recover from failures • Cost — CapEx, OpEx, and total cost of ownership • Responsiveness — Performance and latency • Scalability — Handling growth and demand • Recovery — RTO and RPO objectives • Risk transference — Insurance and third-party risk

AWS's 2017 S3 outage took down large portions of the internet for hours because many companies relied on a single region without redundancy—proving that high availability requires intentional architecture, not assumptions.

Architecture decisions must balance security requirements with business constraints and operational realities.

Why This Matters for the Exam

Architecture considerations are heavily tested on SY0-701 as they connect security to business requirements. Questions cover availability metrics, recovery objectives, and cost trade-offs.

Understanding these trade-offs helps with security planning, disaster recovery, and risk management. Security architectures that ignore business realities fail.

The exam tests understanding of metrics (RTO, RPO, availability) and their implications for security design.

Deep Dive

What Do Availability "Nines" Mean?

Availability measures system uptime as a percentage.

Availability Levels:

AvailabilityDowntime/YearCommon Name
99%3.65 daysTwo nines
99.9%8.76 hoursThree nines
99.99%52.56 minutesFour nines
99.999%5.26 minutesFive nines
99.9999%31.5 secondsSix nines

Achieving High Availability:

  • Load balancing
  • Clustering
  • Redundant components
  • Geographic distribution
  • Automatic failover

Cost vs Availability:

Higher nines = Exponentially higher cost
99% → 99.9% = Significant investment
99.9% → 99.99% = Major investment
99.99% → 99.999% = Massive investment

What Is Resilience and How Do You Build It?

Resilience is the ability to withstand and recover from failures.

Resilience Components:

ComponentDescription
Fault toleranceContinue operating despite failures
RedundancyDuplicate critical components
RecoveryRestore after failure
AdaptabilityAdjust to changing conditions

Building Resilience:

  • Eliminate single points of failure
  • Implement redundancy at all layers
  • Regular testing and drills
  • Automated recovery procedures
  • Geographic distribution

What Is the Difference Between RTO and RPO?

Recovery Objectives:

MetricDefinitionQuestion Answered
RTORecovery Time ObjectiveHow long can we be down?
RPORecovery Point ObjectiveHow much data can we lose?

RTO Example:

Outage at 2:00 PM
RTO = 4 hours
System must be operational by 6:00 PM

RPO Example:

Backup at 6:00 AM
Failure at 2:00 PM
RPO = 4 hours
Maximum 4 hours of data loss acceptable

RTO and RPO Relationship:

  • Lower RTO = Faster recovery needed = Higher cost
  • Lower RPO = Less data loss acceptable = More frequent backups = Higher cost

What Are Hot, Warm, and Cold Sites?

Recovery Site Comparison:

Site TypeRTOCostDescription
Hot siteMinutesHighFully operational duplicate
Warm siteHoursMediumPartial equipment, needs config
Cold siteDaysLowEmpty facility, needs everything

Hot Site:

  • Fully equipped and running
  • Real-time data replication
  • Immediate failover capability
  • Highest cost, lowest RTO

Warm Site:

  • Equipment present but not running
  • Data restored from backups
  • Hours to become operational
  • Balance of cost and recovery time

Cold Site:

  • Empty facility, power, and network only
  • All equipment must be sourced
  • Days to become operational
  • Lowest cost, highest RTO

How Do Scalability Types Differ?

Scaling Types Comparison
Vertical (Scale Up)
Small Server
Bigger Server
More CPU, RAM, storage
Horizontal (Scale Out)
Server
S1
S2
S3
More instances
Vertical: Simple, limited • Horizontal: Complex, resilient

Scalability Comparison:

AspectVerticalHorizontal
ComplexitySimpleComplex
LimitHardware maxVirtually unlimited
ResilienceSingle pointDistributed
Cost patternExpensive incrementsLinear growth

What Cost Models Apply to Security Architecture?

Cost Types:

TypeDefinitionExamples
CapExCapital ExpenditureHardware, facilities
OpExOperational ExpenditureCloud services, subscriptions
TCOTotal Cost of OwnershipAll costs over lifecycle

Security Cost Trade-offs:

  • Higher availability = Higher cost
  • Better security = More investment
  • Redundancy = Duplicate expenses
  • Cloud shifts CapEx to OpEx

How Do You Transfer Security Risk?

Risk Transference Methods:

MethodDescription
InsuranceFinancial compensation for incidents
Cloud servicesProvider assumes some risks
OutsourcingThird party assumes operations
ContractsTransfer liability via agreements

What Risk Transference Does NOT Do:

  • Eliminate risk entirely
  • Transfer reputational damage
  • Remove regulatory responsibility
  • Cover all potential losses

How CompTIA Tests This

Example Analysis

Scenario: A financial services company needs to design a disaster recovery solution. Requirements: maximum 1 hour of downtime (RTO), maximum 15 minutes of data loss (RPO), and budget constraints require cost optimization.

Analysis - Recovery Architecture Design:

Requirements Analysis:

RequirementValueImplication
RTO1 hourHot or warm site needed
RPO15 minutesNear real-time replication
BudgetConstrainedCan't do unlimited spending

Site Selection:

Cold Site:

  • RTO: Days ❌ (doesn't meet 1 hour)
  • Cost: Low ✓
  • Verdict: Does not meet RTO requirement

Warm Site:

  • RTO: Hours ❌ (borderline, risky for 1 hour)
  • Cost: Medium ✓
  • Verdict: May not reliably meet RTO

Hot Site:

  • RTO: Minutes ✓ (easily meets 1 hour)
  • Cost: High (budget concern)
  • Verdict: Meets requirements but expensive

Recommended Solution:

DR Solution Architecture
Primary Site
On-Prem
replicate
←→
Real-time
Hot Site
Cloud
(scales when needed)
✓ RTO: <1 hour✓ RPO: 15 minCost optimized
Hybrid approach: Cloud standby scales up only during failover

Cost Optimization:

  • Use cloud for DR (pay for standby, not full)
  • Right-size standby resources
  • Scale up only during failover
  • Multi-region cloud deployment

Trade-offs:

  • Slightly higher RTO than full hot site
  • Lower cost than traditional hot site
  • Meets both RTO and RPO requirements
  • Acceptable budget impact

Key insight: RTO and RPO drive recovery architecture. The 15-minute RPO requires continuous or near-continuous replication. The 1-hour RTO requires hot or optimized warm site. Cloud enables cost-effective hot site alternatives.

Key Terms

architecture considerationsRTO RPOavailabilityresiliencescalabilityhot sitewarm sitecold sitedisaster recovery

Common Mistakes

Confusing RTO and RPO—RTO is recovery TIME (how long down), RPO is recovery POINT (how much data lost).
Assuming more nines is always better—each additional nine dramatically increases cost. Match availability to business needs.
Ignoring the cost of recovery sites—hot sites are expensive. Cold sites take days. Choose based on actual RTO requirements.
Not testing recovery—a recovery plan that hasn't been tested is just a theory. Regular DR drills are essential.

Exam Tips

RTO = Time (how long can you be down). RPO = Point/Data (how much data can you lose). Easy memory: RTO has "Time," RPO has "Point."
99.99% = "four nines" = approximately 52 minutes downtime per year. This is a common exam reference.
Hot site = minutes (running now). Warm site = hours (needs setup). Cold site = days (empty building).
Vertical scaling = bigger server. Horizontal scaling = more servers. Horizontal provides better resilience.
CapEx = buy it (capital). OpEx = rent it (operational). Cloud typically shifts CapEx to OpEx.
Risk transference (insurance, contracts) doesn't eliminate risk—it provides financial compensation or shifts some responsibility.

Memory Trick

RTO vs RPO - The Two Questions:

RTO = "Recovery Time" = "How long can we be Turned off?" RPO = "Recovery Point" = "How much data can we Permanently lose?"

Recovery Sites - Temperature Memory: Think of how much "heat" (activity) is at each site:

Hot 🔥 = Running hot right now (minutes to failover) Warm 🌡️ = Equipment there but cooling (hours to start) Cold ❄️ = Empty and cold (days to equip)

The "Nines" Pattern: "Every nine you add, the cost goes sky-high" 99% → 3.6 days down 99.9% → 8.7 hours down 99.99% → 52 minutes down (most common exam reference) 99.999% → 5 minutes down

Scaling Memory: Vertical = "Very big server" (scale UP ↑) Horizontal = "Herd of servers" (scale OUT →)

Cost Type Memory: CapEx = Capital = Buy a cap (one-time purchase) OpEx = Operations = Operating expenses (ongoing)

Test Your Knowledge

Q1.A company requires that systems be restored within 2 hours of failure and can tolerate losing up to 30 minutes of data. What are these requirements called?

Q2.Which disaster recovery site type provides the FASTEST recovery but at the HIGHEST cost?

Q3.What availability percentage allows approximately 52 minutes of downtime per year?

Want more practice with instant AI feedback?

Continue Learning

Ready for the Exam?

See exactly where you stand on this concept and 182 others.

99% pass rate · Pass guarantee