How to Make the Most of the Snowflake Free Trial

The Snowflake 30-day free trial gives you $400 in credits and full access to every Enterprise Edition feature — a powerful learning environment. For SnowPro exam prep, running things on a live system in addition to reading the official docs and practice questions dramatically improves your pass rate. This article walks through everything from sign-up to the exam-prep activities that pay off the most.

Free Trial Specs

How to Sign Up for the Trial

1. Go to the official Snowflake site (https://signup.snowflake.com/)
2. Enter your name, email address (work email recommended), and company name
3. Choose a Snowflake edition. For exam prep, pick Enterprise (the default)
4. Choose a cloud provider. AWS has the most regions and is generally recommended
5. Choose a region. If you live in Japan, AWS ap-northeast-1 (Tokyo) gives the lowest latency
6. Accept the terms of service and click GET STARTED
7. Follow the link in the confirmation email to set your password and finish account creation
8. Log in to Snowsight (the web UI) and you can run SQL right away

Managing the $400 in Credits Efficiently

Used carefully, $400 in credits is plenty for 30 days of exam prep. The biggest credit consumer is virtual warehouse runtime.

Credit Consumption by Warehouse Size

X-Small is plenty for exam-prep learning. Use Large or above only temporarily for performance comparison experiments, then drop back to X-Small immediately.

Configuring Auto-Suspend

Always configure auto-suspend on your warehouses. The default is 10 minutes, but for learning use you can save credits by shortening it to 1-2 minutes.

-- オートサスペンドを1分（60秒）に設定
ALTER WAREHOUSE compute_wh
SET AUTO_SUSPEND = 60
    AUTO_RESUME = TRUE;

-- 現在の設定を確認
SHOW WAREHOUSES LIKE 'COMPUTE_WH';

Hands-On Exercises That Map Directly to the Exam

Exercise 1: Confirm the 3-Layer Architecture (Core, ~20% of exam)

Get a hands-on feel for Snowflake's three-layer architecture (Storage / Compute / Cloud Services).

-- Storage Layer: テーブル作成とデータ格納
CREATE DATABASE exam_practice;
CREATE SCHEMA exam_practice.lab;

CREATE TABLE exam_practice.lab.sales (
  sale_id INT AUTOINCREMENT,
  product_name VARCHAR,
  quantity INT,
  sale_date DATE,
  amount DECIMAL(10,2)
);

INSERT INTO exam_practice.lab.sales (product_name, quantity, sale_date, amount)
SELECT 'Product_' || seq4()::VARCHAR,
       UNIFORM(1, 100, RANDOM()),
       DATEADD(day, -UNIFORM(1, 365, RANDOM()), CURRENT_DATE()),
       UNIFORM(100, 10000, RANDOM()) / 100.0
FROM TABLE(GENERATOR(ROWCOUNT => 10000));

-- Compute Layer: 異なるWarehouseでクエリを実行して分離を確認
CREATE WAREHOUSE analytics_wh WITH WAREHOUSE_SIZE = 'XSMALL';
USE WAREHOUSE analytics_wh;
SELECT product_name, SUM(amount) FROM exam_practice.lab.sales GROUP BY 1;

Exercise 2: Time Travel & Clone (Core, frequently tested)

-- Time Travel: データを誤って削除→復元
DELETE FROM exam_practice.lab.sales WHERE sale_date < '2025-06-01';

-- 削除前のデータを確認（AT句でTime Travel）
SELECT COUNT(*) FROM exam_practice.lab.sales
  AT(OFFSET => -60*5);  -- 5分前の状態

-- UNDROPでテーブルを復元
DROP TABLE exam_practice.lab.sales;
UNDROP TABLE exam_practice.lab.sales;

-- Zero-Copy Clone: ストレージコストなしでテーブルを複製
CREATE TABLE exam_practice.lab.sales_clone
  CLONE exam_practice.lab.sales;

-- クローンは独立したオブジェクト（変更が元テーブルに影響しない）
UPDATE exam_practice.lab.sales_clone SET amount = 0 WHERE sale_id = 1;

Exercise 3: Window Functions (most critical for Data Analyst)

-- サンプルデータでウィンドウ関数を練習
SELECT
  o_orderpriority,
  o_totalprice,
  ROW_NUMBER() OVER (PARTITION BY o_orderpriority ORDER BY o_totalprice DESC) AS rn,
  RANK() OVER (PARTITION BY o_orderpriority ORDER BY o_totalprice DESC) AS rnk,
  SUM(o_totalprice) OVER (
    PARTITION BY o_orderpriority
    ORDER BY o_orderdate
    ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW
  ) AS running_total
FROM SNOWFLAKE_SAMPLE_DATA.TPCH_SF1.ORDERS
QUALIFY rn <= 5;

Exercise 4: Cortex AI LLM Functions (Gen AI / Core C03)

-- Cortex COMPLETE: テキスト生成
SELECT SNOWFLAKE.CORTEX.COMPLETE(
  'mistral-large2',
  'Snowflakeのマイクロパーティションとは何かを簡潔に説明してください'
) AS response;

-- Cortex SENTIMENT: 感情分析
SELECT
  comment_text,
  SNOWFLAKE.CORTEX.SENTIMENT(comment_text) AS sentiment_score
FROM (
  SELECT 'このサービスは最高です！毎日使っています。' AS comment_text
  UNION ALL
  SELECT 'レスポンスが遅すぎて使い物にならない。' AS comment_text
);

Exercise 5: Security Settings (Core / Administrator)

-- ロールベースアクセス制御（RBAC）の実践
CREATE ROLE analyst_role;
CREATE ROLE engineer_role;

GRANT USAGE ON DATABASE exam_practice TO ROLE analyst_role;
GRANT USAGE ON SCHEMA exam_practice.lab TO ROLE analyst_role;
GRANT SELECT ON ALL TABLES IN SCHEMA exam_practice.lab TO ROLE analyst_role;

GRANT ALL ON DATABASE exam_practice TO ROLE engineer_role;
GRANT ROLE analyst_role TO ROLE engineer_role;  -- ロール階層

-- Dynamic Data Masking（Enterprise Edition）
CREATE MASKING POLICY email_mask AS (val STRING)
RETURNS STRING ->
  CASE
    WHEN CURRENT_ROLE() IN ('ACCOUNTADMIN', 'ENGINEER_ROLE')
    THEN val
    ELSE '***@***.com'
  END;

Exploring the Snowsight UI

Getting comfortable with the Snowsight web UI also helps on the exam. The Data Analyst exam in particular tests Snowsight Dashboard knowledge.

• Running SQL in worksheets: practice switching results to chart view and configuring filters
• Building dashboards: add tiles from multiple worksheets and verify filter linkage
• Query History: review past query executions and resource consumption
• Query Profile: visually inspect the execution plan and practice locating bottlenecks
• Activity > Usage: monitor credit consumption over time

Tips for Saving Credits

• Manually SUSPEND warehouses you are not using. Do not rely on AUTO_SUSPEND alone
• Default to X-Small for warehouse size; only size up temporarily for performance experiments
• Run SHOW WAREHOUSES regularly to check which warehouses are active
• Set up a Resource Monitor to alert when credit consumption crosses a threshold
• DROP unused databases, schemas, and tables frequently to save storage cost

-- Resource Monitorの設定（クレジット上限アラート）
CREATE RESOURCE MONITOR trial_monitor
  WITH CREDIT_QUOTA = 50
  FREQUENCY = DAILY
  START_TIMESTAMP = CURRENT_TIMESTAMP
  TRIGGERS
    ON 80 PERCENT DO NOTIFY
    ON 100 PERCENT DO SUSPEND;

ALTER WAREHOUSE compute_wh SET RESOURCE_MONITOR = trial_monitor;

Trial Limitations

• Duration: automatically stops after 30 days; cannot be extended
• Credits: usage above $400 is not allowed; additional credits cannot be purchased
• Data retention: after the trial ends, data is retained for 10 days and then deleted
• Organization features: the Organizations feature (multi-account management) is not available
• Support: Community Support only (Premier and Business Critical support are not available)

Check Your Understanding

Frequently Asked Questions