ADA-C01 SnowPro Advanced Administrator Questions and Answers

This command will generate a sample of 20% of the table by using the SYSTEM (or BLOCK) sampling method, which selects each block of rows with a probability of 20/100. This method is suitable for taking data by blocks of rows, as the question requires. According to the Snowflake documentation, “SYSTEM (or BLOCK): Includes each block of rows with a probability of p/100. Similar to flipping a weighted coin for each block of rows. This method does not support fixed-size sampling.” The other options are either incorrect or do not meet the requirements of the question. Option A uses the BERNOULLI (or ROW) sampling method, which selects each row with a probability of 20/100, but does not take data by blocks of rows. Option C uses the BLOCK sampling method, but specifies a fixed number of rows (20) instead of a percentage (20%). Option D uses the SYSTEM sampling method, but specifies a seed value (1), which makes the sampling repeatable, contrary to the question.

According to the Snowflake documentation1, creating a user with a default role does not automatically grant that role to the user. The user must be explicitly granted the role by the role owner or a higher-level role. Therefore, the USERADMIN role, which created the DEVELOPER_ROLE, needs to explicitly grant the DEVELOPER_ROLE to the new user PTORRES using the GRANT ROLE command. Otherwise, the user PTORRES will not be able to use the DEVELOPER_ROLE and will see the default role of PUBLIC in the web UI. Option A is incorrect because the DEVELOPER_ROLE does not need to be granted to SYSADMIN before user PTORRES can use the role. Option B is incorrect because the new role can take effect immediately after it is created and granted to the user, and does not depend on the USERADMIN role logging out. Option D is incorrect because the new role will not be affected by the identity provider synchronization, as it is created and managed in Snowflake.

According to the Snowflake documentation1, the COPY GRANTS option can be used to copy all privileges, except OWNERSHIP, from the existing table to the new transient table. This option also preserves any future grants defined for the object type in the schema. Option A is incorrect because it does not copy any grants from the original table. Option C is incorrect because it does not copy the data from the original table, only the structure and grants. Option E is incorrect because transient tables are not session-based and do not have a Fail-safe period, but they do have a Time Travel retention period2.

1: CREATE TABLE | Snowflake Documentation 2: Working with Temporary and Transient Tables | Snowflake Documentation

According to the GRANT OWNERSHIP documentation, the ownership privilege grants full control over the table and can only be held by one role at a time. However, the current owner can also grant the ownership privilege to another role, which transfers the ownership to the new role. In this case, the SECURITYADMIN role granted the ownership privilege on future tables in the PROD.WORKING schema to the PROD_WORKING_OWNER role. This means that any table created in that schema after the grant statement will be owned by the PROD_WORKING_OWNER role. Therefore, the PROD_WORKING_OWNER role can alter or drop table XYZ, which was created by the ACCOUNTADMIN role in the PROD.WORKING schema. Additionally, the ACCOUNTADMIN role can also alter or drop table XYZ, because it is the top-level role that has all privileges on all objects in the account. Furthermore, the SYSADMIN role can also alter or drop table XYZ, because it was granted the PROD_WORKING_OWNER role by the SECURITYADMIN role. The SYSADMIN role can activate the PROD_WORKING_OWNER role and inherit its privileges, including the ownership privilege on table XYZ. The SECURITYADMIN role cannot alter or drop table XYZ, because it does not have the ownership privilege on the table, nor does it have the PROD_WORKING_OWNER role.

According to the Multi-cluster Warehouses documentation, auto-suspend is a feature that allows a warehouse to automatically suspend itself after a specified period of inactivity. For a multi-cluster warehouse, auto-suspend applies to the entire warehouse, not to individual clusters. Therefore, auto-suspend occurs when the minimum number of clusters is running and there is no activity for the specified period of time. The other options are incorrect because:

•A. Auto-suspend does not occur when there has been no activity on any cluster for the specified period of time. This would imply that each cluster has its own auto-suspend timer, which is not the case. The warehouse has a single auto-suspend timer that is reset by any activity on any cluster.

•B. Auto-suspend does not occur after a specified period of time when an additional cluster has started on the maximum number of clusters specified for a warehouse. This would imply that the auto-suspend timer is affected by the number of clusters running, which is not the case. The auto-suspend timer is only affected by the activity on the warehouse, regardless of the number of clusters running.

•D. Auto-suspend does apply for multi-cluster warehouses, as explained above. It is a feature that can be enabled or disabled for any warehouse, regardless of the number of clusters.

A masking policy is a SQL expression that transforms the data in a column based on the role that queries the column1. The is_role_in_session function returns true if the specified role is in the current session2. Therefore, the masking policy in option A will mask the column data with asterisks whenever it is queried through a view owned by the MASKED_VIEW_ROLE3. The other options use different functions that do not check the ownership of the view, but rather the current role, the invoker role, or the available roles in the session45. These functions may not return the desired result if the role that owns the view is different from the role that queries the view.

According to the Snowflake documentation1, resource monitors are a feature that helps you manage and control Snowflake costs by monitoring and setting limits on your compute resources. Resource monitors do not consume any credits or add any load to the virtual warehouses they monitor1. Resource monitors can also have multiple triggers that specify different actions (such as suspending or notifying) when certain percentages of the credit quota are reached2. Resource monitors can be applied to either the entire account or a specific set of individual warehouses1. The other options are not benefits of resource monitors. The cost of running a resource monitor is negligible, not 10% of a credit3. Multiple resource monitors cannot be applied to a single virtual warehouse; only one resource monitor can be assigned to a warehouse at a time2. Resource monitor governance is not tightly controlled; account administrators can enable users with other roles to view and modify resource monitors using SQL2.

According to the CREATE ACCOUNT documentation, the account name must be specified when the account is created, and it must be unique within an organization, regardless of which Snowflake Region the account is in. The other options are incorrect because:

•The account does not require at least one ORGADMIN role within one of the organization’s accounts. The account can be created by an organization administrator (i.e. a user with the ORGADMIN role) through the web interface or using SQL, but the new account does not inherit the ORGADMIN role from the existing account. The new account will have its own set of users, roles, databases, and warehouses.

•The account name is not immutable and can be changed. The account name can be modified by contacting Snowflake Support and requesting a name change. However, changing the account name may affect some features that depend on the account name, such as SSO or SCIM.

•The account name does not need to be unique among all Snowflake customers. The account name only needs to be unique within the organization, as the account URL also includes the region and cloud platform information. For example, two accounts with the same name can exist in different regions or cloud platforms, such as myaccount.us-east-1.snowflakecomputing.com and myaccount.eu-west-1.aws.snowflakecomputing.com.

According to the Snowflake documentation1, Snowflake uses a multi-version concurrency control (MVCC) model, which means that each transaction operates on a consistent snapshot of the database at a point in time. This allows queries and DML statements to run concurrently without blocking each other, as they do not modify the same data. Therefore, a deadlock, which occurs when concurrent transactions are waiting on resources that are locked by each other, cannot happen in Snowflake. Option B is incorrect because queries and DML statements do not block each other in Snowflake, unless they are explicitly started transactions and multiple statements in each transaction2. Option C is incorrect because transaction locking in Snowflake is enforced at the partition level, not the row or table level3. Option D is incorrect because queries executed within a given transaction do not see that transaction’s uncommitted changes, but only the committed changes that occurred before the transaction started1.

According to the Snowflake documentation1, when a role is dropped, ownership of all objects owned by the dropped role is transferred to the role that is directly above the dropped role in the role hierarchy. This is to ensure that there is always a single owner for each object in the system.

1: Drop Role | Snowflake Documentation

According to the AT | BEFORE documentation, the AT or BEFORE clause is used for Snowflake Time Travel, which allows you to query historical data from a table based on a specific point in the past. The clause can use one of the following parameters to pinpoint the exact historical data you wish to access:

•TIMESTAMP: Specifies an exact date and time to use for Time Travel.

•OFFSET: Specifies the difference in seconds from the current time to use for Time Travel.

•STATEMENT: Specifies the query ID of a statement to use as the reference point for Time Travel.

Therefore, the queries that will allow the user to view the historical data that was in the table before the query was executed are:

•B. SELECT * FROM my_table AT (TIMESTAMP => ‘2021-01-01 07:00:00’ :: timestamp); This query uses the TIMESTAMP parameter to specify a point in time that is before the query execution time of 07:01.

•D. SELECT * FROM my table PRIOR TO STATEMENT ‘8e5d0ca9-005e-44e6-b858-a8f5b37c5726’; This query uses the PRIOR TO STATEMENT keyword and the STATEMENT parameter to specify a point in time that is immediately preceding the query execution time of 07:01.

•F. SELECT * FROM my_table BEFORE (STATEMENT => ‘8e5d0ca9-005e-44e6-b858-a8f5b37c5726’); This query uses the BEFORE keyword and the STATEMENT parameter to specify a point in time that is immediately preceding the query execution time of 07:01.

The other queries are incorrect because:

•A. SELECT * FROM my table WITH TIME_TRAVEL (OFFSET => -60*30); This query uses the OFFSET parameter to specify a point in time that is 30 minutes before the current time, which is 07:30. This is after the query execution time of 07:01, so it will not show the historical data before the query was executed.

•C. SELECT * FROM TIME_TRAVEL (‘MY_TABLE’, 2021-01-01 07:00:00); This query is not valid syntax for Time Travel. The TIME_TRAVEL function does not exist in Snowflake. The correct syntax is to use the AT or BEFORE clause after the table name in the FROM clause.

•E. SELECT * FROM my_table AT (OFFSET => -60*30); This query uses the AT keyword and the OFFSET parameter to specify a point in time that is 30 minutes before the current time, which is 07:30. This is equal to the query execution time of 07:01, so it will not show the historical data before the query was executed. The AT keyword specifies that the request is inclusive of any changes made by a statement or transaction with timestamp equal to the specified parameter. To exclude the changes made by the query, the BEFORE keyword should be used instead.

Option B is the correct answer because it follows the steps described in the Snowflake documentation for sharing data from multiple databases using secure views. The data provider needs to grant the REFERENCE_USAGE privilege on each database that contains objects referenced by the secure view, and the USAGE privilege only on the database where the secure view is created. Option A is incorrect because it grants the USAGE privilege instead of the REFERENCE_USAGE privilege. Option C is incorrect because it grants the REFERENCE_USAGE privilege to a database role, which is not supported. Option D is incorrect because it grants the REFERENCE_USAGE privilege on the wrong database.

According to the Using Persisted Query Results documentation, the RESULT_SCAN function allows you to query the result set of a previous command as if it were a table. The LAST_QUERY_ID function returns the query ID of the most recent statement executed in the current session. Therefore, the combination of these two functions can be used to access the output of the SHOW WAREHOUSES command, which returns the configurations of all the virtual warehouses in the account. However, to persist the warehouse data in JSON format in the table VWH_META, the OBJECT_CONSTRUCT function is needed to convert the output of the SHOW WAREHOUSES command into a VARIANT column. The OBJECT_CONSTRUCT function takes a list of key-value pairs and returns a single JSON object. Therefore, the correct commands to execute are:

1.SHOW WAREHOUSES;

2.INSERT INTO VWH_META SELECT CURRENT_TIMESTAMP (), OBJECT_CONSTRUCT (*) FROM TABLE (RESULT_SCAN (LAST_QUERY_ID ()));

The other options are incorrect because:

•A. This option does not use the OBJECT_CONSTRUCT function, so it will not persist the warehouse data in JSON format. Also, it is missing the * symbol in the SELECT clause, so it will not select any columns from the result set of the SHOW WAREHOUSES command.

•B. This option does not use the OBJECT_CONSTRUCT function, so it will not persist the warehouse data in JSON format. It will also try to insert multiple columns into a single VARIANT column, which will cause a type mismatch error.

•D. This option does not use the OBJECT_CONSTRUCT function, so it will not persist the warehouse data in JSON format. It will also try to use the RESULT_SCAN function on a subquery, which is not supported. The RESULT_SCAN function can only be used on a query ID or a table name.

The inclusion of the CURRENT_TIMESTAMP() function in the second query prevents it from re-using the first query’s persisted result cache because this function makes each execution unique due to the constantly changing timestamp. According to the Snowflake documentation, “The query does not include non-reusable functions, which return different results for successive runs of the same query. UUID_STRING, RANDOM, and RANDSTR are good examples of non-reusable functions.” The CURRENT_TIMESTAMP() function is another example of a non-reusable function, as it returns the current date and time at the start of query execution, which varies for each run. Therefore, the second query is not identical to the first query, and the result cache is not reused. The other options are either incorrect or irrelevant to the question. Option B is incorrect, as the CURRENT_DATE() function is a reusable function, as it returns the same value for all queries executed within the same day. Option C is irrelevant, as the virtual warehouse used to execute the query does not affect the result cache reuse. Option D is also irrelevant, as the role used to execute the query does not affect the result cache reuse, as long as the role has the necessary access privileges for all the tables used in the query.

According to the Using Persisted Query Results documentation, the RESULT_SCAN function allows you to query the result set of a previous command as if it were a table. This can improve the performance of the second query by avoiding reading the same data from micro-partitions again. The other actions do not provide optimal performance for the second query because:

•Assigning additional clusters to the virtual warehouse does not affect the data access speed, but only the query execution speed. It also increases the cost of the warehouse.

•Increasing the STATEMENT_TIMEOUT_IN_SECONDS parameter in the session does not improve the performance of the query, but only allows it to run longer before timing out. It also increases the risk of resource contention and deadlock.

•Preventing the virtual warehouse from suspending between the running of the first and second queries does not guarantee that the data will be cached in memory, as Snowflake uses a least recently used (LRU) cache eviction policy. It also increases the cost of the warehouse.

https://docs.snowflake.com/en/user-guide/querying-persisted-results

Option B is the correct SQL command to create a resource monitor that meets the requirements. It sets the credit quota to 3000, the frequency to yearly, the start timestamp to January 1, 2022, and the triggers to notify and suspend the warehouse at the specified thresholds. Option A is incorrect because it does not specify the frequency. Option C is incorrect because it does not specify the frequency and it uses notify and suspend instead of suspend and suspend_immediate. Option D is incorrect because it does not specify the start timestamp. For more information about resource monitors, see Working with Resource Monitors and CREATE RESOURCE MONITOR.

This code snippet will create a row access policy that returns true if the current role is ALL_ACCESS_ROLE or if the current role matches the region value in the entitlement_table. This means that the ALL_ACCESS_ROLE can see all rows regardless of region, while the other roles can only see rows for their own regions. According to the Snowflake documentation, the CURRENT_ROLE context function returns the name of the current role for the session. The EXISTS function returns true if the subquery returns any rows. The OR operator returns true if either operand is true. Therefore, this code snippet satisfies the requirements of the question.