Disaster recovery strategies in GCP - Time & Space Complexity

Time Complexity: Disaster recovery strategies

O(n)

Understanding Time Complexity

When planning disaster recovery in cloud systems, it's important to understand how the time to recover grows as the amount of data or resources increases.

We want to know how the recovery steps scale when more resources or data are involved.

Scenario Under Consideration

Analyze the time complexity of restoring multiple virtual machines from snapshots in GCP.

for vm in vm_list:
    snapshot = compute.snapshots().get(project=project_id, snapshot=vm.snapshot_name).execute()
    operation = compute.instances().insert(
        project=project_id,
        zone=vm.zone,
        body={
            'name': vm.name,
            'disks': [{'sourceSnapshot': snapshot['selfLink']}],
            'machineType': vm.machine_type
        }
    ).execute()
    wait_for_operation(operation)

This sequence restores each VM by retrieving its snapshot and creating a new instance from it.

Identify Repeating Operations

Look at what repeats for each VM being restored.

• Primary operation: Retrieving snapshot metadata and creating a VM instance from it.
• How many times: Once per VM in the list.

How Execution Grows With Input

Each VM requires its own snapshot retrieval and instance creation, so the total time grows as more VMs are restored.

Input Size (n)	Approx. Api Calls/Operations
10	About 10 snapshot gets and 10 instance creations
100	About 100 snapshot gets and 100 instance creations
1000	About 1000 snapshot gets and 1000 instance creations

Pattern observation: The number of operations grows directly with the number of VMs to restore.

Final Time Complexity

Time Complexity: O(n)

This means the recovery time increases in direct proportion to the number of virtual machines being restored.

Common Mistake

[X] Wrong: "Restoring multiple VMs can be done in constant time because snapshots are ready to use."

[OK] Correct: Each VM requires separate API calls and resource creation, so the total time grows with the number of VMs.

Interview Connect

Understanding how recovery steps scale helps you design better disaster recovery plans and shows you can think about system behavior as it grows.

Self-Check

"What if we restored all VMs in parallel instead of one by one? How would the time complexity change?"

Practice

1. What is the main goal of a disaster recovery strategy in cloud infrastructure?

easy

A. To improve user interface design

B. To keep services running during unexpected problems

C. To increase the speed of data processing

D. To reduce the cost of cloud services

Solution

1. Step 1: Understand disaster recovery purpose

   Disaster recovery aims to keep systems available and data safe during failures or disasters.

2. Step 2: Identify the main goal

   The main goal is to maintain service uptime and data integrity despite problems.

3. Final Answer:

   To keep services running during unexpected problems -> Option B

4. Quick Check:

   Disaster recovery = keep services safe [OK]

Hint: Disaster recovery means service stays up during trouble [OK]

Common Mistakes:

• Confusing disaster recovery with cost saving
• Thinking it improves speed or UI
• Ignoring service availability

2. Which GCP service is commonly used to store backups for disaster recovery?

easy

A. Cloud Pub/Sub

B. Google Compute Engine

C. Google Kubernetes Engine

D. Google Cloud Storage

Solution

1. Step 1: Identify backup storage service

   Backups need durable, scalable storage. Google Cloud Storage is designed for this.

2. Step 2: Exclude compute and messaging services

   Compute Engine and Kubernetes Engine run workloads, Pub/Sub handles messaging, not backups.

3. Final Answer:

   Google Cloud Storage -> Option D

4. Quick Check:

   Backup storage = Cloud Storage [OK]

Hint: Backups go to Cloud Storage, not compute or messaging [OK]

Common Mistakes:

• Choosing compute services for backup storage
• Confusing messaging service with storage
• Not knowing Cloud Storage purpose

3. Consider this GCP setup code snippet for disaster recovery:

resource "google_compute_instance" "primary" {
  name         = "primary-instance"
  zone         = "us-central1-a"
  machine_type = "e2-medium"
  boot_disk {
    initialize_params {
      image = "debian-cloud/debian-11"
    }
  }
}

resource "google_compute_instance" "backup" {
  name         = "backup-instance"
  zone         = "us-central1-b"
  machine_type = "e2-medium"
  boot_disk {
    initialize_params {
      image = "debian-cloud/debian-11"
    }
  }
}

What is the main benefit of this configuration for disaster recovery?

medium

A. It creates two instances in different zones to improve availability

B. It reduces the cost by using smaller machines

C. It backs up data automatically to Cloud Storage

D. It enables faster network speed between instances

Solution

1. Step 1: Analyze instance zones

   The two instances are in different zones (us-central1-a and us-central1-b), which protects against zone failure.

2. Step 2: Identify disaster recovery benefit

   Having instances in separate zones improves availability and fault tolerance during disasters.

3. Final Answer:

   It creates two instances in different zones to improve availability -> Option A

4. Quick Check:

   Multi-zone instances = better availability [OK]

Hint: Different zones mean better uptime during failures [OK]

Common Mistakes:

• Assuming cost reduction is main goal
• Thinking data backup happens automatically
• Confusing network speed with disaster recovery

4. You have a disaster recovery plan that uses snapshots for backups. After a disaster, the recovery fails because snapshots are missing. What is the most likely error?

medium

A. Snapshots were stored in the same zone as the primary data

B. Snapshots were not scheduled regularly

C. Snapshots were encrypted with the wrong key

D. Snapshots were created from the wrong disk

Solution

1. Step 1: Understand snapshot storage risk

   If snapshots are stored only in the same zone, a zone failure can destroy both primary data and backups.

2. Step 2: Identify cause of missing snapshots

   Storing snapshots in the same zone risks losing them during a disaster affecting that zone.

3. Final Answer:

   Snapshots were stored in the same zone as the primary data -> Option A

4. Quick Check:

   Backups must be in different zones [OK]

Hint: Backups in same zone vanish with disaster [OK]

Common Mistakes:

• Ignoring backup location importance
• Assuming encryption or disk choice caused failure
• Not scheduling snapshots regularly

5. You want to design a disaster recovery plan for a critical GCP application that must recover within minutes after a regional outage. Which combination of strategies is best?

hard

A. Use multi-zone deployment with weekly snapshots and manual recovery

B. Use single-zone deployment with daily manual backups to Cloud Storage

C. Use multi-region deployment with automated failover and continuous data replication

D. Use single-region deployment with monthly backups and no failover

Solution

1. Step 1: Identify recovery time objective (RTO)

   Recovery within minutes requires fast failover and minimal data loss.

2. Step 2: Evaluate strategies for fast recovery

   Multi-region deployment with automated failover and continuous replication ensures quick switch and up-to-date data.

3. Step 3: Compare with other options

   Single-zone or single-region with manual backups are slower and risk data loss; weekly or monthly backups are too infrequent.

4. Final Answer:

   Use multi-region deployment with automated failover and continuous data replication -> Option C

5. Quick Check:

   Fast recovery needs multi-region + automation [OK]

Hint: Fast recovery needs multi-region and automation [OK]

Common Mistakes:

• Choosing single-zone or single-region for critical apps
• Relying on manual backups for quick recovery
• Ignoring continuous data replication