Analyzing and optimizing technical and business processes · Topic 4 of 4

Cost Optimization and Business Continuity

CapEx versus OpEx, committed use discounts, Spot VMs, BigQuery editions, budgets and recommendations, plus business continuity planning beyond technical recovery.

  • 4 min
In a nutshell
  • Cloud moves spending to OpEx; commitments add predictability.
  • Resource-based commitments tie to a series; flexible commitments follow spend.
  • Spot for interruptible work; batch pricing for work that can wait.

What it is

Cost optimization means paying only for what delivers value, with spend that the business can predict. Business continuity means the business keeps operating through disruptions, which includes but goes beyond technical disaster recovery. Both are business processes the architect shapes.

Visibilitybilling export and labelsOptimizerightsize, discountsGovernbudgets and policiesReview

How it works

CapEx and OpEx

Capital expense (CapEx)Operating expense (OpEx)
ModelBuy hardware up front, depreciatePay as you use
Cloud fitData center refresh, owned equipmentMost cloud services
EffectLarge, infrequent spending decisionsSpend follows usage; needs ongoing governance

Moving to the cloud shifts spending to OpEx. Commitments bring some predictability back without buying hardware.

Discounts and pricing models

LeverFits
Resource-based committed use discountsSteady use of a specific machine series in a region for one or three years
Flexible (spend-based) committed use discountsSteady spend that may change machine series or regions
Sustained use discountsApplied automatically to some VM families that run much of the month
Spot VMsInterruptible work at a large discount
BigQuery editions with reservations and autoscalingPredictable query spend; a baseline plus autoscaling covers peaks
BigQuery on-demandLight or irregular query volumes
Batch and lower-priority options (batch inference, batch queries)Work that can wait
Exam tip

"Will move to a newer machine series within a year" rules out a long resource-based commitment on the current series. A flexible committed use discount keeps the discount across series.

Resource optimization

  • Rightsize from Active Assist recommendations and real utilization.
  • Autoscale and scale to zero where the platform allows.
  • Storage lifecycle: Autoclass or lifecycle rules for ageing data; delete what is not needed.
  • Log costs: exclude noisy logs from the default sink and route archives to Cloud Storage.
  • Idle resources: unattached disks, idle VMs and unused IPs flagged by recommenders.

Cost governance

ToolUse
Labels and projects per teamAllocate cost
Billing export to BigQueryDetailed analysis and chargeback
Budgets and alertsWarn owners as spend approaches limits; can trigger automation through Pub/Sub
QuotasCap usage of expensive resources
FinOps reviewsRegular joint reviews by finance and engineering

Business continuity

A business continuity plan covers people, processes, suppliers and communications as well as systems:

ElementExample
Critical business functionsOrder placement, claims processing
Acceptable downtime and data lossRTO and RPO per function
Technical recoveryDR patterns from the technical process topic
Manual workaroundsTaking orders by phone if the portal fails
Communication plansWho informs customers and regulators
Supplier dependenciesWhat happens if an on-premises ERP or a partner is down
Regular exercisesTabletop and live tests

Decoupling helps continuity: queueing orders in Pub/Sub lets dealers keep ordering while a back-end ERP is unavailable, and the orders flow when it returns.

How the exam asks about this

  • "500 slots most days, 1,500 for six hours weekly, predictable spend, no queuing." A BigQuery edition reservation with a 500-slot baseline that autoscales to 1,500.
  • "Steady 400 vCPUs, moving to a newer machine series within a year." A one-year flexible committed use discount.
  • "Dealers must order while the ERP is down." Decouple with a queue and deliver when the ERP returns.

What to remember for the exam

  • Cloud moves spending to OpEx; commitments add predictability.
  • Resource-based commitments tie to a series; flexible commitments follow spend.
  • Spot for interruptible work; batch pricing for work that can wait.
  • Labels, billing export, budgets and recommendations govern cost.
  • Business continuity covers people, process, suppliers and communication, not just systems.
Watch out

Cutting cost by removing redundancy can break an availability requirement. Optimize within the requirements.

Common mistakes

Three-year commitments for capacity needed a few hours a week.

Spot VMs for critical stateful services.

Treating DR as the whole continuity plan. People and processes need plans too.

Check yourself

  • Which discount keeps applying when workloads move to a newer machine series?

    A flexible, spend-based committed use discount.

  • How can BigQuery spend become predictable while still covering a weekly peak?

    A BigQuery edition reservation with a baseline and autoscaling up to the peak.

  • What does a business continuity plan add beyond disaster recovery?

    Plans for people, processes, suppliers, manual workarounds and communications.

Topic 4 of 4 in Analyzing and optimizing technical and business processes

Test yourself on Analyzing and optimizing technical and business processes

The full Google Cloud Professional Cloud Architect question bank has 877 questions, with timed practice exams and an explanation for every answer.

This study guide is independent and is not affiliated with, endorsed by, or sponsored by Google Cloud. Vendor names, certification names and exam codes are trademarks of their respective owners and are used for identification only.