Cloud storage must be linked for external stage object no need for internal

stage

Snowflake data warehouse s not built on topp of ay existing data platform

Micro partitions – enable horizontal and verical query pruning and are
physical data files that compromise snowflake logical tables

Cloning objects external names stages are cloned tables and there
internal stages are lcones. Internal names stages and database and
cheme stages are not cloned

Table types – temporary transient and premanent and external (no time
travel)

True or false: You can point Snowflake at any S3 bucket to directly query
the files in that bucket as long as the files are in Parquet or ORC format.
They can be in in any file format

How often does Snowflake release new features? Weekly

The FLATTEN command will parse nested objects into separate rows. One
version of the FLATTEN command uses a join and the other uses an object
keyword. Select the two words that represent the options used with the
FLATTEN command TABLE , LATERAL

Cloud services layer does : User Authentication , Infrastructure

management, metadata management , query parsing and optimisation ,
access control, authentication

Standard snowflake edition does not have elastic multi cluster

warehouses.

Partner Connect a Partner Connect

 Includes automated role, user and staging database set up

 Can be connected from within the WebUI

 Includes a streamlined Partner Trial Account Signup

 Can be connected to Snowflake using a streamlined wizard

Scaling down and up in not a automated process

Different editions require different accounts

Here’s why:

 Cloning and Grants: When cloning a table in Snowflake, the COPY

GRANTS option is specifically used to transfer the access privileges
defined on the original table to the cloned object.
  Default Behavior: If COPY GRANTS is not specified during the
cloning process, the cloned table inherits no explicit access
privileges from the original table.

 Future Grants: Future grants defined for the table in the schema
will not automatically apply to the previously cloned object. You
would need to explicitly grant access to the cloned table for users to
interact with it.

When you clone a table in Snowflake without specifying the COPY

GRANTS option, the behavior is as follows:

 The new cloned table does inherit any explicit access privileges
granted on the original table.

 However, it does not inherit any future grants defined for the
table type in the schema.

Therefore, statement B is FALSE. Cloning a table without COPY

GRANTS creates a separate object with no inherited access privileges.

The clone object does inherit explicit access privileges but not future
grants. If you want the clone to inherit future grants as well, you need to
use the COPY GRANTS option during the clone operation

Each Snowflake account comes with two shared databases. One is a set of
sample data and the other contains Account Usage information. Check all
true statements about these shared databases.


SNOWFLAKE_SAMPLE_DATA contains several schemas from
TPC (tpc.org)

You do not specify availability zones

Providers cannot list data on the market place

Which of the following conditions are required for sharing data in

Snowflake?

Data providers with ACCOUNTADMIN role can set up shares

Consumer accounts must be in same Snowflake region as the

provider account

External stage = cloud storage credentials cloud storage l;ocation and

stage object in snowflake

What command (partial) is used to convert a SECURE VIEW to a “regular“

VIEW?
UNSET SECURE

When adding Consumers to Outbound Shares, what types of Consumer

Accounts can be selected?

Full

Reader

The Query Profiler view is only available for completed queries.

 FALSE

Which of the following terms are associated with the Cloud Services
Layer?

 Query Planning

 Query Compilation

 Query Optimization

Warehouse Size is a good indicartor that the query was executed

Snowflake caches are automatically invalidated if the underlying data

changes. TRUE

If a table is cloned with COPY GRANTS option specified, then the clone
object inherits any explicit access privileges granted on the original table,
but does not inherit any future grants defined for the table in the schema
– TRUE

Scaleout increase cluster increase server

PUT command cannot be executed from web ui – TRUE

Maximum number of consumer account per share object – Unlimited

Which of the following statements are true about Snowflake Data sharing?
Select all that apply.

 Data Sharing is integrated with role-based access controls

 Consumers can query shared tables in the same query as

their own tables

A shared data cannot be cloned by the consumer. It can only be

queried.

Data ui marketplace must be compliant to legal standards hippa must be

comliant to standards real data legally duistribuatbel

Cost – Size of warehouse, Amount of time warehouse has run.

Snowfalke cannot run on private on premisise or hosted infrastructure.

Only Table external tables secure views secure materialized views and
secure udf can be SHARED. REGUALR VIEWS cannot be SHARED

Unlimited shares can be created on a snowflake account and tou can add
unlimited amount of people to a share.

If data is shared with an existing snowflake customer the compute is

charged to the consumer

https://AB12345.snowflakecomputing.com is the ACCOUNT container

Query statement Encryption is done by Enterprise for sensitive data

Looker is business intelligence stuff.

Which statements about Data Integration Tech Partners are true?

 Data Integration Tech Partner software can be used to

extract data from other systems.

 Data Integration Tech Partner software can be used to carry

out transformations.

 Snowflake can carry out transformations after loading files

staged by partner software (ELT).

 Data Integration Tech Partner software should be used to

deliver data to stages, Snowflake is then used to load the
data
Snowflake-Hosted Accounts (on Amazon or AZURE cloud
infrastructure) olption for snowflake hosted accounts

The Snowflake Data Marketplace has two types of listings. These are
PERSONALISED AND DE-PERSONALISED

Account hosted on AWS can access a file staged in GCP or Azure

File formart for data loading is PARQUET CSV ORC AVRO XML JSON

USER STAGE AND TABLE STAGE are automatically created by snowflake

for the user

USER STAGE AND TABLE STAGE INTERNAL STAGE EXTENTAL STAGE

Recommended file size for parallel operation on data files is 100-250MB

COMPRESSED

VARIANT INDIVISUAL ROW MAXIMUM IS 16MB COMPRESSED

Snowflake instances in different regions require separate accounts

Snowflake view types – SECURE, MATERALISED , STANDARD

CONCAT(‘%‘,CONCAT(MY_COL,‘%‘)) %‘||MY_COL||‘%‘ in a column

named MY_COL being sandwiched between two percent signs

Snowpip charge 0.06 per 1000 files queued

External stages require a cloud storage provider

ODBC and JDBC drivers should be able to connect to a tool

Variant data – OPTIMIZED STORAGE BASED ON REPEATED ELEMENTS ,

CAN BE QUERIED using JSON path notation

FAIL SAFE cannot be disabled on a table

SNOWPIPE only works with external stages but SNOWPIP REST

method works with both stages

SNOWSQL IS THE COMMAND LINE INTERFACE

FEDERATED AUTHENTICATION FOR ALL SNOWFLAKE EDITIONS

COLUMN LEVEL SERCURTIY = DYNAMIC DATA MASKING ,

EXTENRAL TOKENIZATION

SHOW WAREHOUSES = SQL command to list all warehouses in an

account
SNOWFALKE contains an ACCOUNT USAGE it is a schema with many
secure views

Order from Lowest cardinality to highest cardinality.

Snowflake UI ribbon = WAREHOUSES(SECURITY ADMIN)

HISTROY(SECURITY ADMIN) ACCOUNT(SECURITY ADMIN)
NOTIFICATIONS (ACCOUNT ADMIN)

SHOW PIPES() COMMAND FOR to see pipes for which you have access
privileges for

Last in first out for servers

All files stored in stages ARE ENCRYPTED

HIPA FEDRAMP PCI DSS SOC1 SOC TYPE 2 NOT FOR Cloud GBDQ

VARIANT ROW DATA 16MB

The clustering depth for a populated table measures the average depth (1
or greater) of the overlapping micro-partitions for specified columns in a
table. The smaller the average depth, the better clustered the table is with
regards to the specified column

 It can be used to determine whether a large table would

benefit from explicitly defining a clustering key

 The depth of the overlapping micro-partitions

To change the warehouse that will be used to run a sql command is Run a
SQL Command like “USE WAREHOUSE LARGE_WH;“ or UPDATE
THE WAREHOUSE FIELD IN THE CONTEXT MENU LOCATED ABOVE
THE WORKSHEET

Micropartition uncompressed is between 50 MB to 500 MB

OLAP good online analytical processing not good for OLTP online
transition processing

Semi structured is nativels supported by snowflake

Results casche 24 hours unless the query executed again then 31 days

Metadata active for 64 days#

You can USE COMMAND for user role, warehouse, database ,schemea or
use xontext menu drop

UNDROP for timtravel

Increasing the size of a warehouse does not improve data loading

performance.file size and number of files does
6 minutes for the amount of queries that will stay backed up before
another cluster is started.

Snowflake minimizes the amount of storage required for historical data by

maintaining only the information required to restore the individual table
rows that were updated or deleted. As a result, storage usage is
calculated as a percentage of the table that changed. Full copies of tables
are only maintained when tables are dropped or truncated.