1

Q1

Is re-clustering in Snowflake only triggered if the table would benefit from the
operation?

A) True.

B) False.

Solution:

A) True.

Explanation:

DML operations (INSERT, UPDATE, DELETE, MERGE, COPY) can make the data in the table
become less clustered. To solve that, Snowflake provides periodic & automatic re-clustering
to maintain optimal clustering. It only reclusters a clustered table if it benefits from the
operation.
2

Q2

What are the usual data loading steps in Snowflake?

Solution:

Explanation:

A source system (like a web application) loads the data into a Snowflake Stage (for example,
an external stage like Amazon S3). Then we can copy the data into a Snowflake table. You
can see this behavior in the following image:
3

Q3

How will you store JSON data in Snowflake?

A) Using a column of the JSON type.

B) Using a column of the VARCHAR type.

C) Using a column of the VARIANT type.

D) Using a column of the NULL type.

Solution:

C) Using a column of the VARIANT type.

Explanation:

Semi-structured data is saved as Variant type in Snowflake tables, with a maximum limit
size of 16MB, and it can be queried using JSON notation. You can store arrays, objects, etc.
4

Q4

Do tables with Fail-Safe turned on incurs additional storage costs compared to

tables where Fail-Safe is turned off?

A) True.

B) False.

Solution:

A) True.

Explanation:

Fail-safe ensures historical data is protected in the event of a system failure or other
catastrophic event, providing a (NON-CONFIGURABLE) 7-day period during which
Snowflake support may recover historical data. It requires additional storage (as other
functionalities like Time Travel), that's why it incurs additional storage costs. You can see
an example of how Fail-Safe works in the following image:
5

Q5

In which of the below scenarios is Snowpipe recommended to load data?

A) We have a small volume of frequent data.

B) We have a huge volume of data generated as part of a batch schedule.

C) In both of the previous scenarios.

Solution:

A) We have a small volume of frequent data.

Explanation:

Snowpipe is a serverless service that enables loading data when the files are available in
any (internal/external) stage. You use it when you have a small volume of frequent data,
and you load it continuously (micro-batches).

Q6

What technique does Snowflake use to limit the number of micro-partitions

retrieved as part of a query?

A) Pruning.

B) Clustering.

C) Indexing.

D) Computing.

Solution:

A) Pruning.

Explanation:

Query pruning consists of analyzing the smallest number of micro-partitions to solve a

query. This technique retrieves all the necessary data to give a solution without looking at
all the micro-partitions, saving a lot of time to return for the result. You can find a real
example at the following link.
6

Q7

Which command will we use to download the files from the stage/location
loaded through the COPY INTO <LOCATION> command?

A) GET.

B) PUT.

C) UNLOAD.

D) INSERT INTO.

Solution:

A) GET.

Explanation:

We will use the GET command to DOWNLOAD files from a Snowflake internal stage (named
internal stage, user stage, or table stage) into a directory/folder on a client machine. You
need to use SnowSQL to use this command.
7

Q8

Can virtual warehouses be resized while they are running?

A) True.

B) False.

Solution:

A) True.

Explanation:

A warehouse can be resized up or down (through the web interface or using SQL) at any
time, including while it is running and processing statements. In the following image, you
can see how to resize them using the web interface:
8

Q9

For which activities does Snowflake have administration settings to help with
resource consumption?

A) Help control costs associated with unexpected warehouse credit usage.

B) Manage access to Snowflake for specific users.

C) Manage the availability of the product.

Solution:

A) Help control costs associated with unexpected warehouse credit usage.

Explanation:

Snowflake provides resource monitors to help control costs and avoid unexpected credit
usage caused by running warehouses. You can impose limits on the number of credits that
warehouses consume.

Q10

Does GET support downloading files from external stages?

A) True.

B) False.

Solution:

B) False.

Explanation:

If you want to download data from external stages, you need to use the cloud provider; for
example, download them directly from AWS S3.
9

Q11

A query executed a couple of hours ago, which spent more than 5 minutes to
run, is executed again, and it returned the results in less than a second. What
might have happened?

A) Snowflake used the persisted query results from the metadata cache.

B) Snowflake used the persisted query results from the query result cache.

C) Snowflake used the persisted query results from the warehouse cache.

D) A new Snowflake version has been released in the last two hours, improving the speed of
the service.

Solution:

B) Snowflake used the persisted query results from the query result cache.

Explanation:

The query result cache stores the results of our queries for 24 hours, so as long as we

same result without using the warehouse and without consuming credits.

Q12

Which table function allows you to convert semi-structured data to a

relational representation?

A) FLATTEN.

B) CHECK_JSON.

C) PARSE_JSON.

Solution:

A) FLATTEN.

Explanation:

FLATTEN is a table function that allows us to convert semi-structured data to a relational

representation, taking a VARIANT, OBJECT, or ARRAY column and producing a lateral view.
10

Q13

What are the different stages available in Snowflake?

A) User.

B) Table.

C) Named internal.

D) Named external.

E) Account.

Solution:

A) User.

B) Table.

C) Named internal.

D) Named external.

Explanation:

Snowflake stages are a big topic in this exam. External stages reference data files stored
outside Snowflake, and internal ones Stores data files internally within Snowflake. The User
user, whereas the table stage is
the one that each table has by default. You can see the different stages in the following
image:
11

Q14

A) The query execution time.

B) The query execution time and the waiting query time.

C) The warehouse size and for how long it runs.

Solution:

C) The warehouse size and for how long it runs.

Explanation:

The cost does not depend on how many queries you run in the warehouse. It depends on
which warehouse size and how long the warehouse runs.

Q15

Can you resize the warehouse once you have selected the size?

A) True.

B) False.

Solution:

A) True.

Explanation:

You can always change the warehouse size depending on your needs, even when it's
running.
12

Q16

Do UDFs support both SQL & JavaScript in Snowflake?

A) True.

B) False.

Solution:

A) True.

Explanation:

User-defined functions (UDFs) let you extend the system to perform operations that are not
-in, system-defined functions. They support SQL,
JavaScript, Java, and Python (these last two are new features).
13

Q17

You have two virtual warehouses in your Snowflake account. If one of them
updates the data in the storage layer, when will the other one see it?

A) Immediately.

B) After an average time of 5 seconds.

C) After the sync process.

Solution:

A) Immediately.

Explanation:

All the warehouses of your account share the storage layer, so if the data is updated, all the
warehouses will be able to see it. You can see this behavior in the following image:
14

Q18

Select the term that is associated with the compute layer:

A) Query optimization.

B) Query planning.

C) Query processing.

Solution:

C) Query processing.

Explanation:

You can find the name of the layers in different ways, like Query Processing for the Compute
Layer.
15

Q19

What property from the Resource Monitors lets you specify whether you want
to control the credit usage of your entire account or a specific set of
warehouses?

A) Credit Quota.

B) Monitor Level.

C) Schedule.

D) Notification.

Solution:

B) Monitor Level.

Explanation:

The monitor level is a property that specifies whether the resource monitor is used to
monitor the credit usage for your entire account or individual warehouses.
16

Q20

You have two types of named stages, one is an external stage, and the other
one is an internal stage. Will external stages always require a cloud storage
provider?

A) True.

B) False.

Solution:

A) True.

Explanation:

External stages reference data files stored in a location outside of Snowflake. Amazon S3
buckets, Google Cloud Storage buckets, and Microsoft Azure containers are the currently
supported cloud storage services. You can see an example of it in the following diagram:
17

Q21

Is clustering generally more cost-effective for tables that are queried

frequently and do not change often?

A) True.

B) False.

Solution:

A) True.

Explanation:

Clustering is a feature that allows you to organize data in a table based on the values of one
or more columns. This can improve query performance by minimizing the amount of data
that needs to be scanned when querying the table. The more frequently a table changes, the
more expensive it will be to keep it clustered.

Q22

Which cloud provider is not supported by Snowflake?

A) AWS.

B) Azure.

C) IBM.

D) Google Cloud Platform.

Solution:

C) IBM.

Explanation:

A Snowflake account can only be hosted on Amazon Web Services, Google Cloud Platforms,
and Microsoft Azure for now.
18

Q23

Is shared data always available to Consumer Accounts?

A) True.

B) False.

Solution:

A) True.

Explanation:

New objects added to a share become immediately available to all consumers, providing
real-time access to shared data, which is always up-to-
storage, as the producer account already pays for it.

Q24

Which function returns the name of the warehouse of the current session?

A) ACTIVE_WAREHOUSE()

B) RUNNING_WAREHOUSE()

C) CURRENT_WAREHOUSE()

D) WAREHOUSE()

Solution:

C) CURRENT_WAREHOUSE()

Explanation:

learning commands by heart, and they are unlikely to appear on the

exam, but this one may be useful. You have other commands to show the current database
and schema, as you can see by executing the following command:

SELECT CURRENT_WAREHOUSE(), CURRENT_DATABASE(), CURRENT_SCHEMA();

19

Q25

How is query processing done in Snowflake?

A) AWS EMR with Spark.

B) AWS EC2 with Spark.

C) Virtual Warehouses.

Solution:

C) Virtual Warehouses.

Explanation:

A Virtual Warehouse is a cluster of computing resources in Snowflake. It provides the

resources necessary to perform queries and DML operations, such as CPU, memory, and
temporary storage. While a warehouse is running, it consumes Snowflake credits.

Q26

Can two different virtual warehouses from the same account access the same
data simultaneously without any contention issue?

A) True.

B) False.

Solution:

A) True.

Explanation:

All the warehouses of your account share the storage layer, so they can access the same data
simultaneously.
20

Q27

The three main S

A) Extraction, Ingestion, Load.

B) Centralized Storage, Compute & Cloud Services.

C) Database, Virtual Warehouse, Data Experience.

Solution:

B) Centralized Storage, Compute & Cloud Services.

Explanation:

You can see them in the following image (via docs.snowflake.com):

21

Q28

Which cache runs for 24 hours?

A) Metadata cache.

B) Results cache.

C) Warehouse cache.

Solution:

B) Results cache.

Explanation:

Query Result cache is also known as Results Cache, which holds the results of every query
executed in the past 24 hours. You can read about all the different types of caches at the
following link.

Q29

Which Snowflake edition (and above) allows until 90 days of Time Travel?

A) Standard.

B) Enterprise.

C) Business Critical.

Solution:

B) Enterprise.

Explanation:

By default, Time travel is enabled with a 1-day retention period. However, we can increase
it to 90 days if we have (at least) the Snowflake Enterprise Edition. It requires additional
storage, which will be reflected in your monthly storage charges.
22

Q30

A) SnowCLI.

B) SnowSQL.

C) SnowTerminal.

D) SnowCMD.

Solution:

B) SnowSQL.

Explanation:

SnowSQL is the command line client for connecting to Snowflake to execute SQL queries
and perform all DDL and DML operations, including loading data into and unloading data
out of database tables.

Q31

cost?

A) True.

B) False.

Solution:

B) False.

Explanation:

With temporary tables, you can optimize storage costs, as when the Snowflake session ends,
data stored in the table is entirely purged from the system. But they also require storage
costs while the session is active. A temporary table is purged once the session ends, so the
retention period is for 24 hours or the remainder of the session.
23

Q32

What should be the first option to restore data into a table?

A) Time-Travel.

B) Fail-Safe.

C) Zero-Copy Cloning.

Solution:

A) Time-Travel.

Explanation:

Time-Travel enables accessing historical data (i.e., data that has been changed or deleted) at
any point within a defined period. If we drop a table, we can restore it with time travel. You
can use it with Databases, Schemas & Tables. The following diagram explains how Time-
Travel works:
24

Q33

Does Snowflake support Nested Transactions?

A) True.

B) False.

Solution:

B) False.

Explanation:

A transaction is a sequence of SQL statements that are committed or rolled back as a unit.
All statements in the transaction are either applied (i.e., committed) or undone (i.e., rolled

Q34

Is Snowflake available on-premise?

A) True.

B) False.

Solution:

B) False.

Explanation:

Snowflake is 100% SaaS.

25

Q35

Which of the below APIs are Snowpipe REST APIs?

A) insertFiles.

B) insertReport.

C) insertHistoryScan.

D) loadFiles.

E) loadHistoryScan.

Solution:

A) insertFiles.

B) insertReport.

E) loadHistoryScan.

Explanation:

You can make calls to REST endpoints to get information. For example, by calling the
following insertReport endpoint, you can get a report of files submitted via insertFiles:

GET
https://<account_id>.snowflakecomputing.com/v1/data/pipes/<pipeNa
me>/insertReport
26

Q36

Can a Warehouse be assigned to more than one Resource Monitor?

A) True.

B) False.

Solution:

B) False.

Explanation:

A monitor can be assigned to one or more warehouses. However, a warehouse can be

assigned to only a single resource monitor below the account level.
27

Q37

What are the two types of data consumer accounts available in Snowflake?

A) Shared Account.

B) Reader Account.

C) Public Account.

D) Full Account.

Solution:

B) Reader Account.

D) Full Account.

Explanation:

There are two types of data consumers. The first one is the Full Accounts, the consumers
with existing Snowflake accounts. In this case, the consumer account pays for the queries
they make. We also have the Reader Accounts, the consumers without Snowflake accounts.
In this last case, the producer account pays all the compute credits that their warehouses
use. You can see this behavior in the following diagram:
28

Q38

Do you need to contact Snowflake Support to retrieve data from Fail-Safe?

A) True.

B) False.

Solution:

A) True.

Explanation:

Fail-Safe provides a (NON-CONFIGURABLE) 7-day period during which Snowflake support

may recover historical data, but you cannot recover this data by yourself; you have to ask
the Snowflake support.
29

Q39

What types of views are available in Snowflake?

A) Regular.

B) Secure View.

C) Table View.

D) Materialized View.

E) External View.

Solution:

A) Regular.

B) Secure View.

D) Materialized View.

Explanation:

You can see the differences between them in the following image:
30

Q40

When staging uncompressed files in a Snowflake stage, are the files

automatically compressed using gzip unless compression is explicitly
disabled?

A) True.

B) False.

Solution:

A) True.

Explanation:

If you have already-compressed files, Snowflake can automatically detect any of these
compression methods (gzip, bzip2, deflate, and raw_deflate) or you can explicitly specify the
method that was used to compress the files.

Q41

A medium (M) warehouse has auto-suspend configured after 15 minutes. You

have noticed that all of the queries that run on this warehouse finish within a
minute. What will you do to optimize compute costs?

A) Delete the warehouse after a minute.

B) Reduce the auto-suspend time to 1 minute.

C) Use another data-warehouse.

Solution:

B) Reduce the auto-suspend time to 1 minute.

31

Q42

Is Snowpipe Serverless?

A) True.

B) False.

Solution:

A) True.

Explanation:

Snowpipe enables loading data when the files are available in any (internal/external) stage.
You use it when you have a small volume of frequent data, and you load it continuously
(micro-
Warehouses. You can see how Snowpipe works in the following diagram:
32

Q43

Are Snowflake caches automatically invalidated if the underlying data

changes?

A) True.

B) False.

Solution:

A) True.

Explanation:

If the data in the Storage Layer changes, the caches are automatically invalidated.
33

Q44

Which of the following roles are System-Defined Roles in Snowflake?

A) ACCOUNTADMIN.

B) SECURITYADMIN.

C) VIEWER.

D) USERADMIN.

E) SYSADMIN.

Solution:

A) ACCOUNTADMIN.

B) SECURITYADMIN.

D) USERADMIN.

E) SYSADMIN.

Explanation:

The PUBLIC role is also a System-Defined role. You can see the differences between them in
the following table:
34
35

Q45

What are the different caching mechanisms available in Snowflake?

A) Metadata cache.

B) Query result cache.

C) Index cache.

D) Table cache.

E) Warehouse cache.

Solution:

A) Metadata cache.

B) Query result cache.

E) Warehouse cache.

Explanation:

The Metadata cache is maintained in Global Service Layer and contains Objects Information
& Statistics. For example, when you execute the command COUNT() on a table, the result
comes from this cache; that's why it returns the information quickly.
The Query Result cache holds the results of every query executed in the past 24 hours. If
you repeat a statement and the underlying data hasn't changed, it will use this cache.
The last one is the warehouse cache, attached to the SSD of each warehouse. In this case, the
information is lost when the warehouse is suspended. You can read more information about
the different caching mechanisms at the following link.
36

Q46

Which Snowflake object stores a generated identity and access management

(IAM) entity for your external cloud storage, along with an optional set of
allowed or blocked storage locations (Amazon S3, Google Cloud Storage, or
Microsoft Azure)?

A) Storage Schema.

B) Storage Integration.

C) User Stage.

Solution:

B) Storage Integration.

Explanation:

A storage integration is a Snowflake object that stores a generated identity and access
management (IAM) entity for your external cloud storage. This option will enable users to
avoid supplying credentials when creating stages or when loading or unloading data.
37

Q47

What statements are true about streams?

A) A Stream itself does NOT contain any table data.

B) A stream only stores the offset for the source table.

C) The hidden columns used by a stream consume storage.

D) All of them.

Solution:

D) All of them.

Explanation:

Streams are Snowflake objects that record data manipulation language (DML) changes
made to tables and views, including INSERTS, UPDATES, and DELETES, as well as metadata
about each change. All of the previous options are correct about them.

Q48

Can you load data using the PUT command through worksheets in the
Snowflake UI?

A) True.

B) False.

Solution:

B) False.

Explanation:

We can use the PUT command to UPLOAD files from a local directory/folder on a client
machine into INTERNAL STAGES. It does NOT work with external stages, and we cannot use
it from the Snowflake Web UI.
38

Q49

What types of tables are available in Snowflake?

A) Permanent.

B) Temporary.

C) Transient.

D) External.

E) Internal.

Solution:

A) Permanent.

B) Temporary.

C) Transient.

D) External.

Explanation:

You can see the differences between these tables in the following image:
39

Q50

In which layer of Snowflake architecture is stored all security-related

information?

A) Storage.

B) Compute.

C) Cloud Services.

D) All of the above.

Solution:

C) Cloud Services.

Explanation:

The Cloud Services layer is a collection of services coordinating activities across Snowflake.
It's in charge of Authentication, Infrastructure management, Metadata management, Query
parsing and optimization, and Access control.

Q51

Can a Reader Account become a Provider of Listings in the Snowflake

Marketplace?

A) True.

B) False.

Solution:

B) False.

Explanation:

To become a provider to offer listings to consumers privately or on the Snowflake

Marketplace, you must have a full Snowflake account with the ability to provide or consume
datasets.
40

Q52

Select two limitations with materialized views:

A) Time Travel is not supported.

B) We cannot define Cluster Keys.

C) Streams cannot track changes.

Solution:

A) Time Travel is not supported.

C) Streams cannot track changes.

Explanation:

You can see the different types of views in the following image:
41

Q53

A) One Node Shared Data.

B) Multi-Cluster Shared Data.

C) One Node Private Data.

Solution:

B) Multi-Cluster Shared Data.

Explanation:

Snowflake's architecture is a hybrid of traditional shared-disk and shared-nothing database

architectures. Snowflake uses a central data repository for persisted data accessible from all
compute nodes in the platform. At the same time, it processes queries using virtual
warehouses where each node in the cluster stores a portion of the entire data set locally.

Q54

Which certifications are compliant with Snowflake?

A) HIPAA.

B) PCI-DSS.

C) FedRAMP.

Solution:

A) HIPAA.

B) PCI-DSS.

C) FedRAMP.

Explanation:

They won't ask you in-depth questions about this topic in the exam, but it's important to
remember some of the most important ones. You can see other certifications at the
following link.
42

Q55

Which statements are correct about micro-partitions in Snowflake?

A) Contiguous units of storage.

B) Non-contiguous units of storage.

C) 50 and 500MB of compressed data.

D) 50 and 500MB of uncompressed data.

E) Organized in a columnar way.

Solution:

A) Contiguous units of storage.

D) 50 and 500MB of uncompressed data.

E) Organized in a columnar way.

Explanation:

tables are
automatically divided into micro-partitions, which are contiguous units of storage between
43

Q56

What happens to the incoming queries when a warehouse does not have
enough resources to process them?

A) Queries are aborted.

B) Queries are queued and executed when the warehouse has resources.

C) Snowflake resizes the warehouse.

Solution:

B) Queries are queued and executed when the warehouse has resources.

Explanation:

If the warehouse does not have enough remaining resources to process a query, the query is
queued, pending resources that become available as other running queries complete.

Q57

Which roles can create shares and resource monitors?

A) ACCOUNTADMIN.

B) SECURITYADMIN.

C) SYSADMIN.

D) USERADMIN.

Solution:

A) ACCOUNTADMIN.

Explanation:

ACCOUNTADMIN is the only role that is able to create Shares and Resource Monitors by
default. However, account administrators can choose to enable users with other roles to
view and modify resource monitors using SQL.
44

Q58

What actions can the resource monitor associated with a Warehouse take
when it reaches (or is about to) hit the limit?

A) Suspend the Warehouse.

B) Send a notification alert.

C) Kill the query that is running.

D) Delete the Snowflake account.

Solution:

A) Suspend the Warehouse.

B) Send a notification alert.

C) Kill the query that is running.

Explanation:

A resource monitor can Notify, Notify & Suspend, and Notify & Suspend Immediately. You
can see these three actions in the following image:
45

Q59

Which are the additional columns that the streams create?

A) METADATA$ACTION.

B) METADATA$ISREAD.

C) METADATA$ISUPDATE.

D) METADATA$ROW_ID.

E) METADATA$COLUMN_ID.

Solution:

A) METADATA$ACTION.

C) METADATA$ISUPDATE.

D) METADATA$ROW_ID.

Explanation:

METADATA$ACTION Indicates the DML operation (INSERT, DELETE) recorded.

METADATA$ISUPDATE indicates whether the operation was part of an UPDATE statement.
METADATA$ROW_ID is a unique and immutable ID for the row.
You can see real examples of how these columns are modified depending on the incoming
data at the following link.
46

Q60

Which character identifies a table stage?

Solution:

Explanation:

The following example shows how we can upload files from our local disk to a Table stage.
As you can see, you need to specify the previous character while indicating the table:

PUT file://c:/myData/myCSV.csv @%myTable

47

Q61

Which of the following services are provided by the Cloud Services Layer?

A) Metadata Management.

B) Authentication.

C) Storage.

D) Infrastructure Management.

E) Query Execution.

Solution:

A) Metadata Management.

B) Authentication.

D) Infrastructure Management.

Explanation:

The Cloud Services layer is a collection of services coordinating activities across Snowflake.
It's in charge of Authentication, Infrastructure management, Metadata management, Query
parsing and optimization, and Access control.
48

Q62

What can you easily check to see if a large table will benefit from explicitly
defining a clustering key?

A) Clustering depth.

B) Clustering ratio.

C) Values in a table.

Solution:

A) Clustering depth.

Explanation:

The clustering depth measures the average depth of the overlapping micro-partitions for
specified columns in a table (1 or greater). The smaller the cluster depth is, the better
clustered the table is. You can get the clustering depth of a Snowflake table using this
command:

SELECT SYSTEM$CLUSTERING_DEPTH('MY_TABLE', '(col1, col2)');

You can also see a real example of how it works at the following link.
49

Q63

Select the statements that are true about Fail-Safe:

A) There is no difference between Time-Travel and Fail-Safe.

B) The data retention period for a permanent table with 30 days of Time-Travel is 37 days.

C) Fail-Safe provides a non-configurable 7-day period.

D) Fail-Safe provides a non-configurable 90-day period.

E) Fail-Safe ensures that historical data is protected in the event of a system failure or other
catastrophic events.

Solution:

B) The data retention period for a permanent table with 30 days of Time-Travel is 37 days.

C) Fail-Safe provides a non-configurable 7-day period.

E) Fail-Safe ensures that historical data is protected in the event of a system failure or other
catastrophic events.

Explanation:

Regarding the second option, we have 30 days of Time-Travel, apart from 7 days of Fail-
Safe. 37 days in total. You can see how Fail-Safe works in the following diagram:
50

Q64

What different types of streams exist in Snowflake?

A) Standard.

B) Append-only.

C) Update-only.

D) Insert-only.

Solution:

A) Standard.

B) Append-only.

D) Insert-only.

Explanation:

Standard and Append-only streams are supported on tables, directory tables, and views.
The Standard one tracks all DML changes to the source table, including inserts, updates, and
deletes, whereas the Append-only one Tracks row inserts only.
The Insert-only stream also tracks row inserts only. The difference with the previous one is
that this one is only supported on EXTERNAL TABLES.
51

Q65

Which data-loading method requires a user-specified warehouse to execute

COPY statements?

A) Bulk Data Load.

B) Snowpipe.

C) Both.

Solution:

A) Bulk Data Load.

Explanation:

Snowpipe is serverless, meaning it doesn't need a running warehouse to load data into
Snowflake.
52

Q66

Which type of data incurs in Snowflake storage costs?

A) Data Stored in permanent tables.

B) Data Stored in temporal tables.

C) Cache results.

D) Data retained for Fail-Safe & Time-Travel.

Solution:

A) Data Stored in permanent tables.

B) Data Stored in temporal tables.

D) Data retained for Fail-Safe & Time-Travel.

Explanation:

Storage fees are incurred for maintaining historical data during both the Time Travel and
Fail-safe periods. To help manage the storage costs, Snowflake provides temporary and
transient tables, which do not incur the same fees as permanent tables but also incur
charges. We can also include data stored in Snowflake locations (i.e., user and table stages
or internal named stages).

Q67

Does Snowflake automatically stores data in encrypted form in all editions?

A) True.

B) False.

Solution:

A) True.

Explanation:

Snowflake encrypts all customer data by default at no additional cost.

53

Q68

What will happen to the child task if you remove its predecessor?

A) The child task is removed from the system.

B) The child task may become the root task.

C) The child task may become a standalone task.

Solution:

B) The child task may become the root task.

C) The child task may become a standalone task.

Explanation:

Also, if the owner role of a task is deleted, the Task Ownership is reassigned to the role that
dropped this role. This is also a typical exam question.
54

Q69

Which properties of a Resource Monitor can be modified?

A) Credit Quota.

B) Monitor Level.

C) Schedule.

D) Actions.

Solution:

A) Credit Quota.

B) Monitor Level.

C) Schedule.

D) Actions.

Explanation:

The Credit Quota specifies the number of Snowflake credits allocated to the monitor for the
specified frequency interval.
The Monitor Level specifies whether the resource monitor is used to monitor the credit
usage for your entire account or individual warehouses.
The Schedule indicates when the monitor will start monitoring and when the credits will
reset to 0.
Each action specifies a threshold and the action to perform when the threshold is reached
within the specified interval.
55

Q70

Queries in Snowflake are getting queued on the warehouses and delaying the
ETL processes of the company. What are the possible solution options you can
think of, considering we have the Snowflake Enterprise addition?

A) Resize the warehouse.

B) Use a multi-cluster warehouse.

C) Set auto-resize parameter to TRUE.

D) Contact Snowflake support to increase the size of the warehouse.

Solution:

A) Resize the warehouse.

B) Use a multi-cluster warehouse.

Explanation:

By resizing the warehouse, your company will scale up, reducing the time to execute big
queries. Using multi-cluster warehouses, you will have more queries running
simultaneously and a high concurrency when they execute, and this is the definition of
scaling out. You can see the differences between the different ways to scale in the following
picture:
56

Q71

Which factors influence the unit cost of snowflake credits and data storage?

A) Snowflake Edition.

B) Region of the Snowflake account.

C) On-Demand or Pre-Paid account.

D) Users on Snowflake.

Solution:

A) Snowflake Edition.

B) Region of the Snowflake account.

C) On-Demand or Pre-Paid account.

Explanation:

You can create as many users as you want without additional cost; that's why that option is
incorrect. You can see a guide about pricing at the following link.
57

Q72

Which are the two metadata columns for staged files?

A) METADATA$FILENAME.

B) METADATA$FILEFORMAT.

C) METADATA$FILE_ROW_NUMBER.

Solution:

A) METADATA$FILENAME.

C) METADATA$FILE_ROW_NUMBER.

Explanation:

The METADATA$FILENAME column is the name of the staged data file that the current row
belongs to. The METADATA$FILE_ROW_NUMBER is the row number for each record in the
container staged data file. This is a way of query the stage metadata:

SELECT metadata$filename, metadata$file_row_number from @MY_STAGE

You can see another example (via docs.snowflake.com) in the following image:
58

Q73

If you want a multi-cluster warehouse, which is the lowest Snowflake edition

that you should opt for?

A) Standard.

B) Enterprise.

C) Business Critical.

D) Virtual Private Snowflake.

Solution:

B) Enterprise.

Explanation:

You can see some differences between the Snowflake editions in the following image:
59

Q74

Which Snowflake object returns a set of rows instead of a single, scalar value,
and can be accessed in the FROM clause of a query?

A) UDF.

B) UDTF.

C) Stored procedure.

Solution:

B) UDTF.

Explanation:

User-defined functions (UDFs) let you extend the system to perform operations that are not
-in, system-defined functions. UDTFs can return multiple
e only difference with UDFs.

Q75

If you want a dedicated virtual warehouse, which is the lowest Snowflake

edition you should opt for?

A) Standard.

B) Enterprise.

C) Business Critical.

D) Virtual Private Snowflake.

Solution:

A) Standard.

Explanation:

In Snowflake, all the Virtual Warehouses are dedicated to the users. If you create a virtual
warehouse, you will only be the one using it.
60

Q76

Can we query data stored in AWS S3 without loading it on a Snowflake table?

A) True.

B) False.

Solution:

A) True.

Explanation:

Imagine you have data on the cloud providers, but the data cannot be copied or moved to
any other location due to compliance regulations. This is a use case of External tables, which
allow us to query data stored in a Cloud Location, for example, AWS S3.

Q77

Which system functions are available in Snowflake to view/monitor the

clustering metadata for a table?

A) SYSTEM$CLUSTERING_DEPTH

B) SYSTEM$CLUSTERING_INFORMATION

C) SYSTEM$CLUSTERING_METADATA

Solution:
A) SYSTEM$CLUSTERING_DEPTH

B) SYSTEM$CLUSTERING_INFORMATION

Explanation:

The clustering depth measures the average depth of the overlapping micro-partitions for
specified columns in a table (1 or greater). The smaller the cluster depth is, the better
clustered the table is. You can use any previous commands to get the Cluster Depth of a
table.
61

Q78

After how many days does the load history of Snowpipe expire?

A) 1 day.

B) 14 days.

C) 90 days.

D) 180 days.

Solution:

B) 14 days.

Explanation:

The load history is stored in the metadata of the pipe for 14 days. Must be requested from
Snowflake via a REST endpoint, SQL table function, or ACCOUNT_USAGE view.

Q79

What influences Snowflake pricing? (SELECT ONLY ONE)

A) Amount of data queried from Snowflake.

B) Amount of data scanned from Snowflake.

C) Snowflake pricing is based on usage & storage.

Solution:

C) Snowflake pricing is based on usage & storage.

Explanation:

-based metrics: compute

and data storage, and you can pay for them independently.
The computing cost is based on the number of credits used to run queries or perform a
service. The storage charge is based on the number of bytes stored per month and the cost
of moving data.
62

Q80

Select the true statements about Snowflake tasks:

A) A task can execute a single SQL Statement.

B) A task can execute multiple SQL Statements.

C) A task can execute a call to a Stored Procedure.

D) A task can execute a function.

Solution:

A) A task can execute a single SQL Statement.

C) A task can execute a call to a Stored Procedure.

Explanation:

Only one SQL statement is allowed to be executed through a task. If you need to execute
multiple statements, build a procedure. You can read more information about Snowflake
tasks at the following link.
63

Q81

Which database objects can be shared using Snowflake Secure Data Sharing?

A) Tables.

B) External tables.

C) Secure views.

D) Secure materialized views.

E) Secure UDFs.

Solution:

A) Tables.

B) External tables.

C) Secure views.

D) Secure materialized views.

E) Secure UDFs.

Explanation:

Secure Data Sharing lets you share selected objects in a database in your account with other
Snowflake accounts. You can share all the previous database objects.
64

Q82

After how many days does the COPY INTO load metadata expire?

A) 1 day.

B) 14 days.

C) 64 days.

D) 180 days.

Solution:

C) 64 days.

Explanation:

The information about the loaded files is stored in Snowflake metadata. It means that you
cannot COPY the same file again in the next 64 days unless you specify it (with the
"FORCE=True" option in the COPY command). You can see this behavior in the following
image:
65

Q83

Does Snowpipe guarantee that files are loaded in the same order they are
staged?

A) True.

B) False.

Solution:

B) False.

Explanation:

Snowpipe generally loads older files first, but there is no guarantee that files are loaded in
the same order they are staged.

Q84

Which of the following commands cannot be executed from the Snowflake UI?

A) SHOW.

B) LIST <stages>

C) GET.

D) COPY INTO.

E) PUT.

Solution:

C) GET.

E) PUT.

Explanation:

These two commands cannot be executed from the Snowflake web interface; instead, you
should use the SnowSQL client to GET or PUT data files.
66

Q85

Which character identifies a user stage?

Solution:

Explanation:

Each user has a Snowflake personal stage allocated to them by default for storing files, and
no one can access them except the user it belongs to. It's represented with the "@~"
character. In the following example, we are uploading the file "myfile.csv" to the stage from
the current user:

PUT file://C:\data\myfile.csv @~
67

Q86

To improve the performance, which of the below techniques can be used in

Snowflake?

A) Cluster Keys.

B) Multi-Warehouses.

C) Maximize the cache use.

D) Increasing the Warehouse Size.

E) Dedicated Warehouses.

Solution:

A) Cluster Keys.

B) Multi-Warehouses.

C) Maximize the cache use.

D) Increasing the Warehouse Size.

E) Dedicated Warehouses.
68

Q87

While loading data through the COPY command, you can transform the data.
Which of the below transformations are allowed?

A) Truncate columns.

B) Omit columns.

C) Filters.

D) Reorder columns.

E) Cast.

Solution:

A) Truncate columns.

B) Omit columns.

D) Reorder columns.

E) Cast.

Explanation:

You can see other transformations at the following link.

69

Q88

We need to temporarily store intermediate data, which an ETL process will

only use. We don't need the data outside the ETL process. If you want to
optimize storage cost, what type of table will you create to store this data?

A) Permanent.

B) Temporary.

C) Transient.

D) External.

Solution:

B) Temporary.

Explanation:

With temporary tables, you can optimize storage costs, as when the Snowflake session ends,
data stored in the table is entirely purged from the system. But they also require storage
costs while the session is active. A temporary table is purged once the session ends, so the
retention period is for 24 hours or the remainder of the session.
70

Q89

Which service does Snowflake use to provide the Zero-Copy cloning

functionality?

A) SSD Cache of the Virtual Warehouses.

B) Cache.

C) Metadata from the service layer.

Solution:

C) Metadata from the service layer.

Explanation:

Zero-Copy cloning does NOT duplicate data; it duplicates the metadata of the micro-
partitions. For this reason, Zero- When you modify
some cloned data, it will consume storage because Snowflake has to recreate the micro-
partitions. You can see this behavior in the following image:
71

Q90

What actions can a Resource Monitor perform when it hits the limit?

A) Notify & Suspend.

B) Notify & Suspend Immediately.

C) Notify.

D) Notify & Increase the limit.

Solution:

A) Notify & Suspend.

B) Notify & Suspend Immediately.

C) Notify.

Explanation:

- Notify --> It performs no action but sends an alert notification (email/web UI).
- Notify & Suspend --> It sends a notification and suspends all assigned warehouses after all
statements being executed by the warehouse (s) have been completed.
- Notify & Suspend Immediately --> It sends a notification and suspends all assigned
warehouses immediately.
72

Q91

Which of the following options are true about roles in Snowflake?

A) Snowflake users have a limit on the number of roles that they can assume.

B) Snowflake users can have one or more roles.

C) Only a single primary role can be active at a time in a user session.

D) Privileges can be directly assigned to users.

Solution:

B) Snowflake users can have one or more roles.

C) Only a single primary role can be active at a time in a user session.

Explanation:

Each user can be assigned multiple roles (and vice versa) but can assume only one role
simultaneously. Privileges are assigned to roles; that's why the last option is false.
73

Q92

What option will you specify to delete the stage files after a successful load
into a Snowflake table with the COPY INTO command?

A) DELETE = TRUE

B) REMOVE = TRUE

C) PURGE = TRUE

D) TRUNCATE = TRUE

Solution:

C) PURGE = TRUE

Explanation:

If the PURGE option is set to TRUE, Snowflake will try its best to remove successfully loaded
data files from stages. If the purge operation fails for any reason, it won't return any error
for now.

COPY INTO mytable PURGE = TRUE;

74

Q93

What techniques would you consider to improve the performance of a query

that takes a lot of time to return any result?

A) Define partition keys.

B) Create cluster keys & turn auto clustering on the table.

C) Create an index on the search result.

Solution:

B) Create cluster keys & turn auto clustering on the table.

Explanation:

Clustering keys are a subset of columns or expressions on a table designated to co-locate the
data in the same micro-partitions. Data might become less clustered when we perform
many DML operations on a table. To solve that, Snowflake also provides periodic &
automatic re-clustering to maintain optimal clustering. These techniques improve the
performance of queries that takes a lot of time, as Snowflake will analyze fewer micro-
partitions.

Q94

How is the data storage cost computed for Snowflake?

A) Based on the average daily amount of uncompressed data stored.

B) Based on the average daily amount of compressed data stored.

C) Based on the amount of uncompressed data stored on the last day of the month.

D) Based on the amount of compressed data stored on the last day of the month.

Solution:

B) Based on the average daily amount of compressed data stored.

Explanation:

Storage costs benefit from the automatic compression of all data stored, and the total
compressed file size is used to calculate the storage bill for an account.
75

Q95

In which of the below scenarios will you use an external table?

A) You have data on the cloud providers, but the data cannot be copied or moved to any
other location due to compliance regulations.

B) You have a high volume of data on the cloud providers, but we only need a part of the
data in Snowflake.

C) You have data on the cloud providers that need to be updated by Snowflake.

D) You have XML data on the cloud provider.

Solution:

A) You have data on the cloud providers, but the data cannot be copied or moved to any
other location due to compliance regulations.

B) You have a high volume of data on the cloud providers, but we only need a part of the
data in Snowflake.

Explanation:

The third answer is incorrect, as external tables can only read data. The fourth option is also
76

Q96

Which command will you run to list all users and roles to which a role has
been granted?

A) SHOW GRANTS TO ROLE <ROLE>

B) SHOW GRANTS OF ROLE <ROLE>

C) SHOW GRANTS IN ROLE <ROLE>

Solution:

B) SHOW GRANTS OF ROLE <ROLE>

Explanation:

the privileges to which this role has access. Here you can see an example of running the
command in my Snowflake account:
77

Q97

A warehouse ran for 62 seconds, and it was suspended. After some time, it ran
for another 20 seconds. For how many seconds will you be billed?

A) 20 seconds.

B) 62 seconds.

C) 92 seconds.

D) 122 seconds.

Solution:

D) 122 seconds.

Explanation:

You will be billed for 122 seconds (62 + 60 seconds) because warehouses are billed for a
minimum of one minute. The price would be different if the warehouse wasn't suspended
before executing the second query.
For example, if we had only run a query, and it had only run for 62 seconds, you would be
billed for these 62 seconds. If it had only run for 20 seconds, you would've been billed for 60
seconds.
78

Q98

Which of the following clustering metadata for the micro-partitions is

maintained by Snowflake in a table?

A) The number of micro-partitions that comprise the table.

B) The number of micro-partitions containing values that overlap with each other.

C) The depth of the overlapping micro-partitions.

D) None of the above.

Solution:

A) The number of micro-partitions that comprise the table.

B) The number of micro-partitions containing values that overlap with each other.

C) The depth of the overlapping micro-partitions.

Q99

If a transaction disconnects and goes into a detached state, which cannot be

committed or rolled back, how long will Snowflake take to abort the
transaction?

A) 15 minutes.

B) 60 minutes.

C) 4 hours.

D) 12 hours.

Solution:

C) 4 hours.

Explanation:

If the transaction is left open or not aborted by the user, Snowflake automatically rolls back
the transaction after being idle for four hours. You can still abort a running transaction with
the system function: SYSTEM$ABORT_TRANSACTION
79

Q100

Which Snowflake edition supports private communication between Snowflake

and your other VPCs through AWS PrivateLink?

A) Standard.

B) Enterprise.

C) Business Critical.

Solution:

C) Business Critical.

Explanation:

AWS PrivateLink is an AWS service for creating private VPC endpoints that allow direct,
secure connectivity between your AWS VPCs and the Snowflake VPC without traversing the
public Internet. This feature requires the Business Critical edition or higher. You can see the
differences between the Snowflake editions in the following image: