ADA-C01 Question 393
Single answerCalculate data storage usage and credit consumptionA Snowflake administrator is reviewing last month's costs for a business unit. The account uses Enterprise Edition and has one XS warehouse for ETL and one M warehouse for BI. The following activity occurred during a 30-day month:
- 4 TB of average compressed data was actively stored in database tables for the entire month.
- Time Travel retention for these tables is 7 days, and on average 0.5 TB of changed data remained in Time Travel throughout the month.
- Fail-safe storage averaged 0.2 TB during the month.
- The XS warehouse ran 2 hours per day.
- The M warehouse ran 3.5 hours per day.
- No serverless features were used.
Which approach will produce the most accurate estimate of the month's Snowflake charges for data storage and compute credits?
- A
Calculate storage as 4.0 TB only, because Time Travel and Fail-safe are included at no additional charge; calculate compute using warehouse hourly rates multiplied by runtime.
- B
Calculate storage as 4.7 TB-months (4.0 + 0.5 + 0.2), then calculate compute credits separately as XS hours plus M hours using each warehouse size's credit rate and total runtime over the month.
- C
Calculate storage as 4.5 TB-months because Fail-safe is not billed; calculate compute by summing the two warehouse runtimes and multiplying by the M warehouse credit rate.
- D
Calculate storage as 4.7 TB-months, then convert that storage amount into credits using the warehouse size rates so storage and compute can be compared on the same billing basis.
Show answer and explanation
Correct answer: B
Explanation
To estimate Snowflake charges accurately, administrators must separate storage billing from compute billing. Storage is generally charged based on the average amount of data stored during the billing period, including active table storage as well as retained historical data for Time Travel and Fail-safe. In this scenario, the billed storage estimate is 4.7 TB-months, derived from 4.0 TB active data + 0.5 TB in Time Travel + 0.2 TB in Fail-safe.
Compute is billed separately in credits. Each virtual warehouse consumes credits according to its size while it is running. Because the environment uses both an XS warehouse and an M warehouse, the administrator must calculate runtime for each warehouse independently and apply the correct warehouse-size credit rate to each. The scenario explicitly states that no serverless features were used, so no additional serverless credit estimation is needed.
This aligns with Snowflake documentation and billing best practices: use ACCOUNT_USAGE/ORGANIZATION_USAGE views or Snowsight billing views to analyze storage categories and warehouse credit consumption separately, and do not assume Time Travel or Fail-safe are free from a storage-cost perspective.
- A. Incorrect.
Incorrect. Snowflake bills for storage consumed by active data, Time Travel, and Fail-safe. A common misconception is that Time Travel and Fail-safe are just logical features with no storage cost impact. While they are platform features, the bytes retained for those states contribute to billed storage. The compute portion of this option is directionally correct, but the storage assumption makes the overall approach inaccurate.
- B. Correct.
Correct. For storage estimation, the administrator should include all billed storage categories: active storage, Time Travel storage, and Fail-safe storage. In this scenario that is 4.0 + 0.5 + 0.2 = 4.7 TB-months on average for the month. Compute is billed separately from storage, so warehouse consumption should be calculated by warehouse size and runtime: XS warehouse hours over 30 days and M warehouse hours over 30 days, multiplied by each warehouse's credit consumption rate. This is the most accurate approach based on Snowflake billing principles.
- C. Incorrect.
Incorrect. Fail-safe storage is billed, so excluding the 0.2 TB understates storage usage. This option also incorrectly applies the M warehouse credit rate to all warehouse runtime. Snowflake warehouse credit consumption depends on warehouse size, so XS and M must be calculated separately. This distractor reflects two common cost-estimation mistakes: ignoring Fail-safe and using one warehouse rate for mixed workloads.
- D. Incorrect.
Incorrect. Storage charges are not converted into credits using warehouse rates. In Snowflake, storage and compute are separate billing dimensions: storage is billed based on average TB stored over time, while compute is billed in credits based on warehouse or serverless usage. This option mixes two different pricing models and would not reflect how Snowflake charges are actually calculated.