GCP ACE Prep

Topic: Planning and Configuring Storage Options

The Two Main Storage Paradigms

The first step in choosing storage is to identify the type of data and access pattern. GCP storage options fall into two main categories.

Block Storage: Persistent Disk (PD)

Persistent Disk provides reliable, high-performance block storage for your Compute Engine instances. The key decision for the ACE exam is choosing between Zonal and Regional availability.

The core trade-off is Performance & Cost vs. High Availability.
            graph TD
                subgraph "Region: us-central1"
                    subgraph "Zone: us-central1-a"
                        VM_A(VM in Zone A)
                        ZonalPD(Zonal PD)
                        VM_A -- Attaches to --> ZonalPD
                    end
                    subgraph "Zone: us-central1-b"
                        Replica(Regional PD Replica)
                    end
                    RegionalPD(Regional PD
in Zone A) VM_A -- Attaches to --> RegionalPD RegionalPD -- Synchronously Replicates --> Replica end

1. Zonal Persistent Disk

2. Regional Persistent Disk

Object Storage: Cloud Storage Classes

Cloud Storage offers different "storage classes" for your data. The choice depends entirely on how frequently you plan to access the data.

The core trade-off is Storage Cost vs. Access Cost & Minimum Duration. Hotter tiers cost more to store but are free to access. Colder tiers are cheap to store but have retrieval fees and minimum storage durations.
Storage Class Access Frequency Min. Storage Duration Use Case
Standard High ("Hot" data)
Accessed frequently
None Website content, streaming video, active data for analytics.
Nearline Low ("Warm" data)
Accessed < 1x per month
30 days Monthly backups, data you don't need daily but might need quickly.
Coldline Very Low ("Cold" data)
Accessed < 1x per quarter
90 days Disaster recovery data, long-tail media content.
Archive Minimal ("Archive" data)
Accessed < 1x per year
365 days Long-term legal/regulatory archives, data under a legal hold.

Decision Flowchart for Storage Selection

graph TD
    A{Start: What kind of data?} --> B{Is it for a VM filesystem - Block Storage?};
    B -- Yes --> C{Does the disk itself need to survive a zonal failure?};
    C -- Yes --> D[Regional Persistent Disk];
    C -- No --> E[Zonal Persistent Disk];

    B -- No, it's files/objects --> F{How often will you access the data?};
    F -- Frequently --> G[Standard Storage];
    F -- About once a month --> H[Nearline Storage];
    F -- About once a quarter --> I[Coldline Storage];
    F -- About once a year or less --> J[Archive Storage];

    style D fill:#2E7D32, color:white
    style E fill:#2E7D32, color:white
    style G fill:#2E7D32, color:white
    style H fill:#2E7D32, color:white
    style I fill:#2E7D32, color:white
    style J fill:#2E7D32, color:white

            

ACE Exam Practice Questions

Question 1

You are running a critical, single-instance Oracle database on a Compute Engine VM. The business requires the database to remain available even if the entire GCP zone it is running in experiences an outage. Which storage option should you use for the database files?

Click to see options and answer

A. Zonal Persistent Disk

B. Regional Persistent Disk

C. Cloud Storage Standard class bucket

D. Local SSD

Correct Answer: B

Explanation: The key requirement is for the storage to survive a zonal outage for a single-instance workload. This is the exact use case for Regional Persistent Disk. A Zonal PD (A) would become unavailable with the zone. Cloud Storage (C) is object storage and not suitable for a running database filesystem. A Local SSD (D) is ephemeral and not persistent at all.

Question 2

Your company has a policy to back up all project source code at the end of every month. These backups are rarely accessed, but when they are, they need to be available immediately. You want to choose the most cost-effective storage solution for these monthly backups. Which Cloud Storage class should you use?

Click to see options and answer

A. Standard

B. Nearline

C. Coldline

D. Archive

Correct Answer: B

Explanation: The access pattern is "monthly." This directly maps to the intended use case for Nearline storage. Standard (A) would be too expensive for data accessed so infrequently. Coldline (C) is designed for quarterly access, and Archive (D) for yearly access; both would be cheaper to store but have higher retrieval costs and longer minimum storage durations than necessary for this use case, making Nearline the most cost-effective choice overall.