Why NoSQL on GCP matters - Performance Analysis

Time Complexity: Why NoSQL on GCP matters

O(n)

Understanding Time Complexity

When using NoSQL databases on GCP, it's important to understand how the time to perform operations changes as data grows.

We want to know how the number of database calls grows when we add more data or requests.

Scenario Under Consideration

Analyze the time complexity of the following operation sequence.

// Using Firestore NoSQL database on GCP
const db = Firestore();

async function addUsers(users) {
  for (const user of users) {
    await db.collection('users').add(user);
  }
}

// Adds each user document one by one to the 'users' collection

This code adds multiple user records to a Firestore collection, one at a time.

Identify Repeating Operations

Identify the API calls, resource provisioning, data transfers that repeat.

• Primary operation: Adding a user document to Firestore collection.
• How many times: Once per user in the input list.

How Execution Grows With Input

Each user added requires one separate API call to Firestore.

Input Size (n)	Approx. API Calls/Operations
10	10 calls
100	100 calls
1000	1000 calls

Pattern observation: The number of calls grows directly with the number of users added.

Final Time Complexity

Time Complexity: O(n)

This means the time to add users grows linearly with how many users you add.

Common Mistake

[X] Wrong: "Adding many users at once will take the same time as adding one user."

[OK] Correct: Each user requires a separate call, so more users mean more calls and more time.

Interview Connect

Understanding how NoSQL operations scale helps you design efficient cloud apps and answer questions about performance in real projects.

Self-Check

"What if we batch multiple user additions into a single API call? How would the time complexity change?"

Practice

1. Why is NoSQL often chosen on Google Cloud Platform (GCP) for certain applications?

easy

A. Because it handles flexible and rapidly changing data well

B. Because it requires complex server management

C. Because it only supports structured data

D. Because it is slower than traditional databases

Solution

1. Step 1: Understand NoSQL data handling

   NoSQL databases are designed to store flexible, semi-structured, or unstructured data that can change quickly.

2. Step 2: Relate to GCP use case

   On GCP, NoSQL services like Firestore and Bigtable support fast-changing data without rigid schemas.

3. Final Answer:

   Because it handles flexible and rapidly changing data well -> Option A

4. Quick Check:

   NoSQL = flexible, fast-changing data [OK]

Hint: NoSQL = flexible data, not rigid tables [OK]

Common Mistakes:

• Thinking NoSQL requires complex server management
• Assuming NoSQL only supports structured data
• Believing NoSQL is slower than SQL

2. Which GCP NoSQL service is best described as a document database?

easy

A. Cloud Spanner

B. Cloud Bigtable

C. Cloud SQL

D. Cloud Firestore

Solution

1. Step 1: Identify document database on GCP

   Cloud Firestore stores data as documents in collections, making it a document database.

2. Step 2: Differentiate from other services

   Cloud Bigtable is a wide-column store, Cloud SQL and Spanner are relational databases.

3. Final Answer:

   Cloud Firestore -> Option D

4. Quick Check:

   Document DB on GCP = Firestore [OK]

Hint: Firestore = document DB, Bigtable = wide-column [OK]

Common Mistakes:

• Confusing Bigtable as document DB
• Choosing relational DBs like Cloud SQL
• Mixing Spanner with NoSQL services

3. Given this code snippet using Firestore on GCP:

doc_ref = firestore_client.collection('users').document('user123')
doc_ref.set({'name': 'Alice', 'age': 30})
user = doc_ref.get().to_dict()
print(user['name'])

What will be printed?

medium

A. user123

B. Alice

C. 30

D. None

Solution

1. Step 1: Understand Firestore document creation

   The code sets a document with fields 'name' and 'age' under 'users/user123'.

2. Step 2: Retrieve and print the 'name' field

   The get() call fetches the document, and user['name'] accesses the 'name' value, which is 'Alice'.

3. Final Answer:

   Alice -> Option B

4. Quick Check:

   Print user['name'] = Alice [OK]

Hint: Set then get document fields to print values [OK]

Common Mistakes:

• Printing document ID instead of field
• Confusing field names
• Assuming None if document exists

4. You try to write data to Cloud Bigtable but get an error. Which mistake likely caused it?

table = bigtable_client.table('my-table')
row = table.row('row1')
row.set_cell('cf1', 'name', 'Bob')
row.commit()

medium

A. Missing column family 'cf1' creation before setting cell

B. Using wrong row key format

C. Calling commit() before set_cell()

D. Bigtable does not support string values

Solution

1. Step 1: Check Bigtable column family setup

   Bigtable requires column families to exist before writing cells; 'cf1' must be created first.

2. Step 2: Verify code sequence and data types

   Row key format is valid, commit() is correctly called after set_cell(), and Bigtable supports string values.

3. Final Answer:

   Missing column family 'cf1' creation before setting cell -> Option A

4. Quick Check:

   Column family must exist before writing [OK]

Hint: Create column family before writing cells [OK]

Common Mistakes:

• Forgetting to create column family
• Assuming commit() order is wrong
• Thinking Bigtable rejects strings

5. You want to build a scalable app on GCP that needs to store user profiles with flexible fields and also handle millions of fast writes per second. Which combination of NoSQL services should you choose?

hard

A. Use Cloud SQL for profiles and Cloud Spanner for writes

B. Use Cloud Bigtable for profiles and Cloud Firestore for writes

C. Use Cloud Firestore for flexible profiles and Cloud Bigtable for high-speed writes

D. Use Cloud Spanner for both profiles and writes

Solution

1. Step 1: Match data needs to NoSQL services

   Cloud Firestore handles flexible, document-based user profiles well.

2. Step 2: Match write speed needs

   Cloud Bigtable supports massive scale and very fast writes, ideal for high throughput.

3. Step 3: Evaluate other options

   Cloud SQL and Spanner are relational and less suited for flexible schema or extreme write speed.

4. Final Answer:

   Use Cloud Firestore for flexible profiles and Cloud Bigtable for high-speed writes -> Option C

5. Quick Check:

   Firestore + Bigtable = flexible + fast writes [OK]

Hint: Firestore for flexible data, Bigtable for speed [OK]

Common Mistakes:

• Choosing relational DBs for flexible schema
• Swapping Firestore and Bigtable roles
• Ignoring write speed requirements