Credit usage monitoring in Snowflake - Time & Space Complexity

Time Complexity: Credit usage monitoring

O(n)

Understanding Time Complexity

When monitoring credit usage in Snowflake, it's important to understand how the time to gather usage data changes as the amount of data grows.

We want to know how the execution time scales when querying credit usage information.

Scenario Under Consideration

Analyze the time complexity of the following Snowflake query that retrieves credit usage details.

SELECT 
  WAREHOUSE_NAME, 
  SUM(CREDITS_USED) AS TOTAL_CREDITS
FROM 
  SNOWFLAKE.ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY
WHERE 
  START_TIME >= DATEADD(day, -7, CURRENT_TIMESTAMP())
GROUP BY 
  WAREHOUSE_NAME;
    

This query sums the credits used by each warehouse over the last 7 days.

Identify Repeating Operations

Look for repeated actions in the query.

• Primary operation: Scanning rows in the WAREHOUSE_METERING_HISTORY table.
• How many times: Once for each row in the last 7 days.

How Execution Grows With Input

The number of rows scanned grows with the amount of usage data stored.

Input Size (rows)	Approx. Operations
10	10 scans and sums
100	100 scans and sums
1000	1000 scans and sums

Pattern observation: The work grows directly with the number of rows scanned.

Final Time Complexity

Time Complexity: O(n)

This means the time to get credit usage grows linearly with the number of usage records.

Common Mistake

[X] Wrong: "The query time stays the same no matter how much data there is."

[OK] Correct: The query must scan more rows as data grows, so it takes longer.

Interview Connect

Understanding how query time grows with data size helps you design efficient monitoring and alerting in real systems.

Self-Check

What if we added a clustering key on START_TIME? How would the time complexity change?

Practice

1. What is the main purpose of monitoring credit usage in Snowflake?

easy

A. To control and optimize Snowflake costs

B. To increase the number of warehouses automatically

C. To backup data to external storage

D. To improve query performance by caching

Solution

1. Step 1: Understand credit usage concept

   Credit usage monitoring tracks how many credits Snowflake consumes, which directly affects billing.

2. Step 2: Identify the goal of monitoring

   The goal is to control and optimize costs by understanding usage patterns.

3. Final Answer:

   To control and optimize Snowflake costs -> Option A

4. Quick Check:

   Credit monitoring = Cost control [OK]

Hint: Credit usage monitoring helps manage costs effectively [OK]

Common Mistakes:

• Confusing credit monitoring with performance tuning
• Thinking it automatically scales warehouses
• Assuming it backs up data

2. Which Snowflake view provides detailed credit usage information for warehouses?

easy

A. INFORMATION_SCHEMA.TABLES

B. ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY

C. ACCOUNT_USAGE.LOGIN_HISTORY

D. SNOWFLAKE.ACCOUNT_USAGE.QUERY_HISTORY

Solution

1. Step 1: Identify views related to credit usage

   ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY tracks credit consumption by warehouses.

2. Step 2: Exclude unrelated views

   INFORMATION_SCHEMA.TABLES shows table info, LOGIN_HISTORY tracks logins, QUERY_HISTORY tracks queries but not credit usage.

3. Final Answer:

   ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY -> Option B

4. Quick Check:

   Warehouse credit data = WAREHOUSE_METERING_HISTORY [OK]

Hint: Warehouse credit data is in WAREHOUSE_METERING_HISTORY view [OK]

Common Mistakes:

• Choosing QUERY_HISTORY which tracks queries, not credits
• Confusing LOGIN_HISTORY with credit usage
• Selecting INFORMATION_SCHEMA views unrelated to credits

3. What will the following query return?

SELECT WAREHOUSE_NAME, SUM(CREDITS_USED) AS TOTAL_CREDITS FROM SNOWFLAKE.ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY GROUP BY WAREHOUSE_NAME ORDER BY TOTAL_CREDITS DESC;

medium

A. List of warehouses with their current running queries

B. Total credits used by each user in the account

C. Total credits used by each warehouse, sorted from highest to lowest

D. A syntax error due to missing WHERE clause

Solution

1. Step 1: Analyze the SELECT and GROUP BY

   The query sums credits used grouped by warehouse name.

2. Step 2: Understand ORDER BY clause

   Ordering by TOTAL_CREDITS DESC sorts warehouses from highest to lowest credit usage.

3. Final Answer:

   Total credits used by each warehouse, sorted from highest to lowest -> Option C

4. Quick Check:

   SUM credits by warehouse, ordered DESC = Total credits used by each warehouse, sorted from highest to lowest [OK]

Hint: GROUP BY warehouse sums credits; ORDER BY DESC sorts high to low [OK]

Common Mistakes:

• Thinking it shows user credit usage
• Assuming it lists running queries
• Believing missing WHERE causes syntax error

4. You run this query but get no results:

SELECT WAREHOUSE_NAME, SUM(CREDITS_USED) FROM SNOWFLAKE.ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY WHERE START_TIME > CURRENT_DATE() GROUP BY WAREHOUSE_NAME;

What is the likely cause?

medium

A. Missing GROUP BY clause

B. Syntax error due to missing alias for SUM

C. Incorrect table name used

D. No credit usage recorded today yet

Solution

1. Step 1: Check the WHERE condition

   Filtering START_TIME > CURRENT_DATE() means only today's data is shown.

2. Step 2: Consider data availability

   If no credits used today yet, query returns empty result, not error.

3. Final Answer:

   No credit usage recorded today yet -> Option D

4. Quick Check:

   Empty result = no data matching date filter [OK]

Hint: Empty results often mean no data matches the filter [OK]

Common Mistakes:

• Assuming missing alias causes error
• Thinking table name is wrong
• Believing GROUP BY is missing

5. You want to monitor credit usage by service type (e.g., WAREHOUSE, CLOUD_SERVICES) for the last 7 days. Which query correctly achieves this?

hard

A. SELECT SERVICE_TYPE, SUM(CREDITS_USED) AS TOTAL_CREDITS FROM SNOWFLAKE.ACCOUNT_USAGE.METERING_HISTORY WHERE START_TIME >= DATEADD(day, -7, CURRENT_DATE()) GROUP BY SERVICE_TYPE ORDER BY TOTAL_CREDITS DESC;

B. SELECT SERVICE_TYPE, SUM(CREDITS_USED) FROM SNOWFLAKE.ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY WHERE START_TIME >= DATEADD(day, -7, CURRENT_DATE()) GROUP BY SERVICE_TYPE;

C. SELECT WAREHOUSE_NAME, SUM(CREDITS_USED) FROM SNOWFLAKE.ACCOUNT_USAGE.METERING_HISTORY WHERE START_TIME >= DATEADD(day, -7, CURRENT_DATE()) GROUP BY WAREHOUSE_NAME;

D. SELECT SERVICE_TYPE, CREDITS_USED FROM SNOWFLAKE.ACCOUNT_USAGE.METERING_HISTORY WHERE START_TIME >= DATEADD(day, -7, CURRENT_DATE());

Solution

1. Step 1: Identify correct view for service type credit data

   ACCOUNT_USAGE.METERING_HISTORY contains credit usage by service type; WAREHOUSE_METERING_HISTORY is only for warehouses.

2. Step 2: Check query filters and grouping

   Filtering last 7 days with DATEADD and grouping by SERVICE_TYPE is correct; ordering by total credits helps see highest usage.

3. Final Answer:

   SELECT SERVICE_TYPE, SUM(CREDITS_USED) AS TOTAL_CREDITS FROM SNOWFLAKE.ACCOUNT_USAGE.METERING_HISTORY WHERE START_TIME >= DATEADD(day, -7, CURRENT_DATE()) GROUP BY SERVICE_TYPE ORDER BY TOTAL_CREDITS DESC; -> Option A

4. Quick Check:

   Use METERING_HISTORY + SERVICE_TYPE + last 7 days = SELECT SERVICE_TYPE, SUM(CREDITS_USED) AS TOTAL_CREDITS FROM SNOWFLAKE.ACCOUNT_USAGE.METERING_HISTORY WHERE START_TIME >= DATEADD(day, -7, CURRENT_DATE()) GROUP BY SERVICE_TYPE ORDER BY TOTAL_CREDITS DESC; [OK]

Hint: Use METERING_HISTORY view and group by SERVICE_TYPE for service credit usage [OK]

Common Mistakes:

• Using WAREHOUSE_METERING_HISTORY for service types
• Not grouping by SERVICE_TYPE
• Selecting credits without aggregation