Backup and restore in GCP - Step-by-Step Execution

Process Flow - Backup and restore

Start Backup Process

↓

Identify Data to Backup

↓

Create Backup Snapshot

↓

Store Backup Securely

↓

Backup Complete

↓

Start Restore Process

↓

Select Backup Snapshot

↓

Restore Data from Backup

↓

Verify Restore Success

↓

Restore Complete

The flow shows starting a backup by selecting data, creating and storing a snapshot, then restoring by selecting that snapshot and verifying success.

Execution Sample

GCP

gcloud sql backups create --instance=my-instance
# Backup created

gcloud sql instances restore-backup 20230401 --instance=my-restored-instance
# Restore started

This code creates a backup of a Cloud SQL instance and then restores it from a specific backup ID to a new instance.

Process Table

Step	Action	Input/Parameter	Result/Output	Next Step
1	Start Backup	Instance: my-instance	Backup process initiated	Identify Data to Backup
2	Identify Data	Database data and config	Data ready for backup	Create Backup Snapshot
3	Create Snapshot	Data from instance	Backup snapshot created with ID 20230401	Store Backup Securely
4	Store Backup	Snapshot 20230401	Backup stored in GCP storage	Backup Complete
5	Backup Complete	-	Backup process finished successfully	Start Restore Process
6	Start Restore	Backup ID 20230401	Restore process initiated	Select Backup Snapshot
7	Select Snapshot	Backup ID 20230401	Backup snapshot selected	Restore Data from Backup
8	Restore Data	Backup snapshot data	Data restored to instance	Verify Restore Success
9	Verify Restore	Restored data	Restore verified successful	Restore Complete
10	Restore Complete	-	Restore process finished successfully	END

💡 Restore process ends after successful verification of restored data.

Status Tracker

Variable	Start	After Step 3	After Step 6	After Step 9	Final
Backup Snapshot ID	None	20230401	20230401	20230401	20230401
Backup Status	Not started	Created	Created	Restoring	Completed
Restore Status	Not started	Not started	Started	Verified	Completed

Key Moments - 3 Insights

Why do we need to verify the restore after restoring data?

Verification ensures the restored data matches the backup and the system works correctly, as shown in step 9 of the execution_table.

Can we restore from a backup before it is fully created?

No, the backup must be fully created and stored first (step 4) before starting the restore process (step 6).

What happens if the backup snapshot ID is incorrect during restore?

The restore will fail because the system cannot find the backup to restore, so selecting the correct snapshot (step 7) is crucial.

Visual Quiz - 3 Questions

Look at the execution_table, what is the Backup Status after step 3?

A Created

B Not started

C Restoring

D Completed

💡 Hint

Check the 'Backup Status' variable in variable_tracker after step 3.

At which step does the restore process start according to the execution_table?

A Step 4

B Step 6

C Step 5

D Step 7

💡 Hint

Look for 'Restore process initiated' in the Result/Output column.

If the backup snapshot ID was missing, which step would fail?

A Step 8

B Step 3

C Step 6

D Step 10

💡 Hint

Refer to step 6 where the restore process starts using the backup ID.

Concept Snapshot

Backup and restore in GCP involves creating a snapshot of your data and storing it securely.
Use 'gcloud sql backups create' to backup and 'gcloud sql instances restore-backup' to restore.
Always verify the restore to ensure data integrity.
Backups must complete before restore starts.
Keep track of backup IDs to select correct snapshots.

Full Transcript

This visual execution shows the backup and restore process in Google Cloud Platform. It starts by initiating a backup on a Cloud SQL instance, identifying the data to backup, creating a snapshot, and storing it securely. After the backup completes, the restore process begins by selecting the backup snapshot, restoring data to the instance, and verifying the restore success. Variables like backup snapshot ID and status change step-by-step. Key moments clarify why verification is needed, the order of operations, and the importance of correct snapshot IDs. The quiz tests understanding of these steps and states.

Practice

1. What is the main purpose of creating a backup in Google Cloud SQL?

easy

A. To save a copy of your database data safely for recovery

B. To increase the speed of your database queries

C. To permanently delete old data from your database

D. To automatically update your database schema

Solution

1. Step 1: Understand backup purpose

   Backups are copies of data saved to protect against data loss.

2. Step 2: Identify backup use in Cloud SQL

   In Cloud SQL, backups allow restoring data if something goes wrong.

3. Final Answer:

   To save a copy of your database data safely for recovery -> Option A

4. Quick Check:

   Backup = Safe data copy [OK]

Hint: Backup means saving data safely for later recovery [OK]

Common Mistakes:

• Confusing backup with performance improvement
• Thinking backup deletes data
• Assuming backup changes database structure

2. Which gcloud command correctly creates an on-demand backup for a Cloud SQL instance named my-instance?

easy

A. gcloud sql backup create --name=my-instance

B. gcloud sql instances backup create my-instance

C. gcloud sql backups create --instance=my-instance

D. gcloud sql create backup --instance=my-instance

Solution

1. Step 1: Recall correct gcloud backup syntax

   The correct command to create a backup is gcloud sql backups create --instance=INSTANCE_NAME.

2. Step 2: Match command to instance name

   Replacing INSTANCE_NAME with my-instance gives the correct command.

3. Final Answer:

   gcloud sql backups create --instance=my-instance -> Option C

4. Quick Check:

   Backup command = gcloud sql backups create [OK]

Hint: Use 'gcloud sql backups create --instance=NAME' to backup [OK]

Common Mistakes:

• Using 'instances backup' instead of 'backups create'
• Incorrect flag names like --name instead of --instance
• Wrong command order or missing 'create'

3. Given this command output snippet after restoring a Cloud SQL backup:

Restoring backup 2024-06-01_12-00 to instance my-instance... DONE

What does this output indicate?

medium

A. The backup was successfully restored to the instance

B. The backup creation failed

C. The instance was deleted

D. The backup is scheduled for later restoration

Solution

1. Step 1: Read the command output carefully

   The message says 'Restoring backup ... DONE', indicating completion.

2. Step 2: Understand the meaning of 'DONE'

   'DONE' means the restore process finished successfully.

3. Final Answer:

   The backup was successfully restored to the instance -> Option A

4. Quick Check:

   Restore output 'DONE' = success [OK]

Hint: Look for 'DONE' to confirm successful restore [OK]

Common Mistakes:

• Confusing restore success with backup failure
• Assuming instance deletion from restore message
• Thinking restore is scheduled, not completed

4. You run the command gcloud sql backups create --instance=my-instance but get an error: ERROR: (gcloud.sql.backups.create) The instance does not support on-demand backups.
What is the most likely cause?

medium

A. You typed the instance name incorrectly

B. The Cloud SQL instance is a First Generation instance which does not support on-demand backups

C. You need to enable billing on your Google Cloud project

D. The backup command requires a different flag for on-demand backups

Solution

1. Step 1: Understand error message about backup support

   The error says the instance does not support on-demand backups.

2. Step 2: Recall Cloud SQL instance generations

   First Generation instances do not support on-demand backups; only Second Generation and later do.

3. Final Answer:

   The Cloud SQL instance is a First Generation instance which does not support on-demand backups -> Option B

4. Quick Check:

   First Gen instances lack on-demand backup support [OK]

Hint: Check instance generation; First Gen lacks on-demand backups [OK]

Common Mistakes:

• Assuming typo in instance name causes this error
• Thinking billing is the cause
• Believing backup flag is incorrect

5. You want to restore a Cloud SQL instance prod-db to a previous state using a backup from June 1, 2024. Which sequence of steps correctly achieves this with minimal downtime?

hard

A. Restore the backup directly to the current instance while it is serving traffic

B. Delete the current instance, then restore the backup to the same instance name immediately

C. Export the backup to Cloud Storage, then import it back to the instance

D. Create a new instance from the backup, then switch your application to use the new instance

Solution

1. Step 1: Understand restore options for minimal downtime

   Restoring directly to a live instance causes downtime; better to create a new instance.

2. Step 2: Create new instance from backup and switch traffic

   Create a new instance restored from backup, then update your app to use it, minimizing downtime.

3. Final Answer:

   Create a new instance from the backup, then switch your application to use the new instance -> Option D

4. Quick Check:

   New instance restore + switch = minimal downtime [OK]

Hint: Restore to new instance, then switch app to avoid downtime [OK]

Common Mistakes:

• Deleting instance before restore causing data loss
• Restoring live instance causing service interruption
• Using export/import which is slower and more complex